# AnxietyResearch.org full text > Anxiety studies explained clearly. AnxietyResearch.org is a research translation platform that turns peer-reviewed anxiety studies, clinical guidelines, and public-health data into clear, accurate, plain-language explanations people can understand and use. AnxietyResearch.org is published by shrinkMD Publishing, LLC. It's an editorial publication, not a clinical service, and it doesn't provide medical advice, diagnosis, or treatment. Content is reviewed by Shariq Refai, MD, MBA, FAPA, a board-certified psychiatrist. Every page draws on peer-reviewed literature, clinical guidelines, and primary public-health datasets, with sources cited on the page. The site doesn't sell products or accept advertising. Content last updated: 2026-07-03. This file inlines the full text of the site's core content pages. The curated index is at https://anxietyresearch.org/llms.txt. # Data reports ## The U.S. Anxiety Access Report URL: https://anxietyresearch.org/reports/us-anxiety-access Topic cluster: Access and systems Last updated: 2026-05-18 About one in five U.S. adults meets criteria for an anxiety disorder in any given year. A much smaller share receives treatment, and only a fraction of those treated receive what surveys classify as minimally adequate care. ### Key takeaways - About 19% of U.S. adults meet criteria for an anxiety disorder in any given year. - Only about 37% of adults with any anxiety disorder receive treatment in a given year, and of those treated, roughly one-third receive what surveys classify as minimally adequate care. - Access varies sharply by geography, insurance coverage, and provider density. ### What the data show - The clearest way to see access is as a cascade that narrows at each step. The starting point is prevalence. NIMH, drawing on the National Comorbidity Survey Replication, puts the past-year prevalence of any anxiety disorder at 19.1% of US adults, lifetime prevalence at 31.1%, and serious impairment among those affected at 22.8%. - The next step is treatment. Of adults with any anxiety disorder, NCS-R analysis found that 36.9% received any treatment in a given year. The step after that is adequacy. Of the people who received treatment, 33.9% received care that met the survey's definition of minimally adequate, meaning a threshold number of therapy visits or an adequate medication course with follow-up. Carried through the full cascade, only a minority of adults with an anxiety disorder receive care that meets that adequacy threshold. - A second, more recent anchor comes from the CDC Household Pulse Survey, which has reported recent symptoms of anxiety or depression in a substantial share of adults, around 27% in recent waves. Pulse measures symptoms with a short screener rather than diagnoses with a full interview, so its numbers run higher and answer a different question. The two sources are best read side by side, not merged into one figure. ### How to read this - NCS-R remains the canonical disorder-level anchor for US anxiety prevalence, even though its fieldwork dates to 2001 to 2003, because no later national survey has replicated its full diagnostic interview. 'Minimally adequate care' is a specific research definition, not a statement that the care was clinically sufficient for the patient. The Pulse figure and the NCS-R figure measure different things and shouldn't be compared directly. ### Limits - NCS-R fieldwork is more than two decades old. - The CDC Household Pulse Survey uses self-reported symptoms rather than diagnostic interviews. - National averages hide large variation by state, income, race, and age. - Survey designs differ in sampling, mode, and instruments, which limits direct comparison across sources. ### Sources - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - CDC Household Pulse Survey, Mental Health (https://www.cdc.gov/nchs/covid19/pulse/mental-health.htm) - Kessler et al., NCS-R, 2005 (https://pubmed.ncbi.nlm.nih.gov/15939837/) ## Psychiatric wait times in the United States URL: https://anxietyresearch.org/reports/psychiatric-wait-times Topic cluster: Access and systems Last updated: 2026-05-18 The AMN Healthcare/Merritt Hawkins physician appointment survey has put psychiatry among the harder specialties to reach for a new-patient visit for two decades. Secret-shopper studies find similar delays plus high rates of directories that don't reach real appointments. A long delay is itself an access barrier: some patients disengage before the first visit ever happens. ### Key takeaways - AMN Healthcare/Merritt Hawkins' 2022 survey reported an average new-patient wait for family medicine of about 26 days in large metros, with psychiatry historically among the specialties with the longest waits. - Malowney and colleagues' 2015 secret-shopper study of Boston, Chicago, and Houston found that a first appointment took an average of 25 days when a patient could reach a clinician at all, and that many listings were unreachable. - Wait times are longer in rural areas, longer for in-network psychiatrists, and longer for public insurance than for private. - Roughly a third of the US population lives in a Mental Health Professional Shortage Area (HRSA). - The psychiatric workforce is aging; roughly 60 percent of psychiatrists are over age 55 per AAMC data. ### What the surveys measure Two kinds of studies dominate the psychiatric wait-time literature. Appointment-availability surveys, like the recurring AMN Healthcare/Merritt Hawkins survey, sample a set of metro areas and call clinics posing as patients to book new-patient appointments. Secret-shopper studies, like Malowney and colleagues' 2015 work in Boston, Chicago, and Houston, do the same thing more intensively in a smaller geographic footprint and report both the wait when an appointment is available and the fraction of listed clinicians who cannot be reached at all. ### What the surveys find Averages of roughly 20 to 30 days for a first psychiatric appointment are common in the metro-area studies; rural averages run considerably longer. What the numeric averages hide is that a large fraction of listed clinicians are unreachable, not accepting new patients, or not accepting the patient's insurance. Malowney reported that when a patient could reach a clinician, the wait was about 25 days; a substantial fraction of listings were unreachable, which is functionally equivalent to unavailable care. ### The workforce side AAMC workforce data show that roughly 60 percent of US psychiatrists are over age 55, and a large share is within a decade of typical retirement. Psychiatry residency slots have not expanded as fast as the demand curve. Bishop and colleagues' 2016 Health Affairs paper documented an outright decline in the total number of practicing US psychiatrists from 2003 to 2013. Whether that trend has fully reversed depends on the year and dataset. ### What actually shortens the wait Three things move the number, per the literature. The first is telepsychiatry, which expands the reach of the existing workforce (see [telepsychiatry report](/reports/telepsychiatry-and-anxiety-care)). The second is expanded scope of practice for psychiatric nurse practitioners, which adds prescribers. The third is collaborative care and stepped-care models that keep milder cases in primary care and reserve psychiatry for more complex ones (see [anxiety in primary care](/reports/anxiety-and-primary-care)). None of these fix the underlying workforce gap; they redistribute how the existing workforce is used. ### What this doesn't prove Wait-time averages are misleading if read in isolation. The distribution matters: a metro-area average of 25 days masks both same-week availability at one type of practice (typically cash-pay or academic centers with capacity) and 8-week or longer waits at another. Insurance status changes the picture completely. A wait-time number is a useful proxy for access but not a substitute for a specific patient's experience. ### What the data show - Wait times for a new outpatient psychiatric appointment are long across much of the United States. Surveys of appointment availability, including the periodic physician appointment wait-time survey conducted by AMN Healthcare and Merritt Hawkins, have repeatedly placed psychiatry among the harder specialties to access on a new-patient basis, with waits commonly measured in weeks. Secret-shopper studies, in which researchers pose as patients and call to book, have documented real delays. Malowney and colleagues' 2015 Psychiatric Services paper reported that in Boston, Chicago, and Houston, a first outpatient appointment took an average of 25 days when a patient could reach a clinician at all, and that a substantial fraction of listed clinicians could not be reached, were not accepting new patients, or would not accept the patient's insurance. - Two structural factors drive the wait. The first is workforce size. The supply of psychiatrists hasn't kept pace with demand, and the psychiatric workforce is aging, with roughly 60 percent of psychiatrists over age 55 per AAMC workforce data and a large share nearing retirement. The second is distribution. Psychiatrists cluster in metropolitan areas and around academic medical centers, which lengthens waits sharply in rural and lower-income areas. - Wait time isn't only an inconvenience. A long delay between seeking help and receiving it is itself an access barrier, because some people disengage during the wait and never start care. This is a documented piece of the treatment cascade described in the [U.S. Anxiety Access Report](/reports/us-anxiety-access). See also the reports on [insurance acceptance trends](/reports/insurance-acceptance-trends) and [rural and urban access](/reports/rural-and-urban-access) for the specific factors that lengthen the wait. ### How to read this - Wait-time surveys differ in method and in which cities they sample, so any single headline figure should be read as illustrative rather than exact. The consistent finding across methods is direction: psychiatry waits are long relative to other specialties. - Secret-shopper studies measure appointment availability at the time of the call; they do not measure whether patients eventually got care. ### Limits - No single national registry tracks psychiatric wait times. - Estimates come from periodic surveys and audit studies with differing methods. - Waits vary widely by region, by insurance type, and between new-patient and established-patient appointments. - Secret-shopper studies are typically limited to a few metro areas. ### Sources - AMN Healthcare / Merritt Hawkins, Survey of Physician Appointment Wait Times (https://www.amnhealthcare.com/) - HRSA Bureau of Health Workforce, mental health workforce reports (https://bhw.hrsa.gov/data-research) - Malowney M, et al. Availability of outpatient care from psychiatrists: a secret-shopper survey. Psychiatric Services, 2015. PMID: 25686810 (https://pubmed.ncbi.nlm.nih.gov/25686810/) - Malowney M, et al. Availability of outpatient care from psychiatrists: a secret-shopper survey. Psychiatric Services, 2015. PMID: 25686810 (https://pubmed.ncbi.nlm.nih.gov/25686810/) - Bishop TF, et al. Population of US Practicing Psychiatrists Declined, 2003-13. Health Affairs, 2016. PMID: 27385238 (https://pubmed.ncbi.nlm.nih.gov/27385238/) - AAMC, Physician Specialty Data Report (https://www.aamc.org/data-reports/workforce) - AMN Healthcare / Merritt Hawkins, Survey of Physician Appointment Wait Times (https://www.amnhealthcare.com/) ## Anxiety treatment access by state URL: https://anxietyresearch.org/reports/anxiety-treatment-access-by-state Topic cluster: Access and systems Last updated: 2026-05-18 State-level anxiety-care access varies by roughly two-fold across the US. HRSA shortage-area designations, KFF workforce data, and Mental Health America's annual state rankings each pick up different pieces of the same picture: rural and lower-income states consistently sit at the bottom. National averages hide most of what matters for a given resident. ### Key takeaways - As of recent HRSA data, roughly a third of the US population lives in a Mental Health Professional Shortage Area, and shortage-area designations meet only about 27 percent of estimated need on average. - Psychiatrist density varies from over 20 per 100,000 in Massachusetts to under 5 per 100,000 in several rural states (HRSA workforce data). - Mental Health America's annual state rankings combine prevalence and access measures into comparative scores; rankings shift year to year but relative positions are stable. - State rankings correlate with rural population share and with Medicaid expansion status. - See the [state data directory](/statistics/states) for the state-by-state detail. ### What varies by state Three things vary the most: workforce density, insurance acceptance, and Medicaid coverage rules. Massachusetts, Connecticut, and Vermont sit at the top on most rankings; Nevada, Alabama, Mississippi, and several plains states sit at the bottom. The gap between top and bottom is roughly four-fold in psychiatrist density. On MHA's composite rankings the spread is narrower because prevalence and coverage measures also enter, but the direction is the same. ### HRSA's shortage-area framework HRSA designates Mental Health Professional Shortage Areas at the county or sub-county level based on population-to-provider ratios and other factors. About a third of the US population lives in a designated area, and designated areas meet on average only about 27 percent of estimated need per KFF's analysis. The designations are used to allocate federal resources including National Health Service Corps placements and loan repayment. ### Where state rankings help and where they mislead State rankings are useful for seeing where a state sits relative to others and for tracking change over time. They mislead when they're treated as measures of individual access: a state can rank high on average while a specific rural county in that state ranks very low. The Maine access report is a case study of exactly this: [Maine ranks 2nd nationally on access](/reports/maine-mental-health-access) while meeting only about 14 percent of estimated mental-health-practitioner need in its designated shortage areas. ### Levers that move the needle The state-policy levers most studied are Medicaid expansion (which improves insurance coverage but doesn't create providers), telehealth-friendly licensure and payment rules (which expand cross-state access), and workforce investments including psychiatric residency slots, loan repayment, and integration of behavioral health into primary care. Collaborative care payment codes, expansion of scope of practice for advanced-practice providers, and support for community mental health centers are the other levers on the table. See the report on [telepsychiatry](/reports/telepsychiatry-and-anxiety-care) for the biggest structural shift of the past decade. ### What the data show - Access to anxiety care isn't distributed evenly across the country. Two state-level factors do most of the work. The first is workforce density, the number of psychiatrists and other mental health prescribers per resident. The federal government tracks the shortfall through Mental Health Professional Shortage Area designations, maintained by HRSA. As of recent HRSA data, roughly a third of the US population lives in a designated shortage area, and those areas meet on average only about 27 percent of estimated need. State-level psychiatrist density varies from over 20 per 100,000 in states like Massachusetts to under 5 per 100,000 in several rural states. - The second factor is insurance acceptance. Where psychiatrists are scarce, the ones who do practice are more able to operate outside insurance networks, which shifts cost onto patients and narrows real access even where a provider technically exists (see the report on [insurance acceptance trends](/reports/insurance-acceptance-trends)). - State mental health rankings, such as those published annually by Mental Health America, combine prevalence and access measures into a single comparative score. They're useful for seeing the spread between states, although the underlying measures change from year to year. The full state-by-state detail is collected in the [state data directory](/statistics/states) on this site. State-specific access reports for [Maine](/reports/maine-mental-health-access) and [Nebraska](/reports/nebraska-mental-health-access) show how the national numbers land in specific places. ### How to read this - State rankings combine several measures and shift year to year, so they're best read as a comparative snapshot rather than a fixed grade. A county-level shortage designation flags undersupply but not its severity. - "Share of need met" in HRSA data is an estimate, not a direct count of unmet appointments. ### Limits - County-level shortage designations are a blunt measure. - State rankings depend on which measures are included and how they're weighted. - National and state averages both hide variation at the county and neighborhood level. - Comparing across years requires the same methodology; HRSA and MHA both revise their approaches periodically. ### Sources - HRSA Health Professional Shortage Areas (https://data.hrsa.gov/topics/health-workforce/shortage-areas) - Mental Health America, state rankings (https://mhanational.org/issues/state-mental-health-america) - Kaiser Family Foundation, Mental Health (https://www.kff.org/mental-health/) - HRSA Bureau of Health Workforce, mental health workforce reports (https://bhw.hrsa.gov/data-research) - Mental Health America, State of Mental Health in America (annual state rankings) (https://mhanational.org/issues/state-mental-health-america) - KFF State Health Facts, Mental Health Care HPSAs (https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-areas-hpsas/) ## Insurance acceptance trends in psychiatry URL: https://anxietyresearch.org/reports/insurance-acceptance-trends Topic cluster: Access and systems Last updated: 2026-05-18 Psychiatrists accept insurance at lower rates than physicians in almost every other medical specialty. That gap has been stable across two decades of data, and it's the reason so much US psychiatric care is delivered out-of-network or paid for out-of-pocket. Parity law raised coverage on paper without closing the network-adequacy gap on the ground. ### Key takeaways - Psychiatrists accepted private insurance for new patients at 55.3 percent versus 88.7 percent for physicians in other specialties in the most-cited national study (Bishop et al., JAMA Psychiatry, 2014). - Medicare acceptance among psychiatrists was 54.8 percent versus 86.1 percent for other specialties in the same study. - The gap widens further for mental-health-specific carve-out networks: patients frequently encounter narrow or inaccurate directories, sometimes called ghost networks. - Mental health parity law (MHPAEA 2008, plus 2020 and 2024 amendments) requires comparable coverage but does not directly change network size. - Out-of-pocket spending is a larger share of psychiatric care than of other outpatient medical care. ### What the acceptance-rate gap looks like Bishop and colleagues' 2014 JAMA Psychiatry paper analyzed a national sample from the CDC's National Ambulatory Medical Care Survey. They compared office-based psychiatrists with office-based physicians in every other specialty on a single question: whether the practice accepted new patients with a given payer. The results were consistent across every payer they looked at. Psychiatrists accepted private non-capitated insurance at 55.3 percent versus 88.7 percent for other specialties, Medicare at 54.8 percent versus 86.1 percent, and Medicaid at 43.1 percent versus 73.0 percent. Almost a decade of secret-shopper studies since then have found similar gaps in real-world appointment availability. See the [Malowney psychiatric secret-shopper study (Psychiatric Services, 2015)](https://pubmed.ncbi.nlm.nih.gov/25686810/) for one widely-cited replication. ### How the gap opened, and why it persists Three factors show up repeatedly in the workforce literature. First, reimbursement per visit for evaluation and management codes has not kept pace with the time psychiatric visits take, especially initial evaluations and complex follow-ups. Second, administrative burden, including pre-authorization for medications and network paperwork, adds unpaid overhead. Third, demand exceeds supply in most metropolitan areas, so a psychiatrist can fill a schedule without accepting any insurance at all. The result is a market in which cash-pay and out-of-network practice is more sustainable in psychiatry than in most of medicine. Bishop and colleagues' follow-up work on network adequacy documented that many psychiatrists listed as in-network are unreachable, not accepting new patients, or unwilling to accept the plan they are listed under, patterns often described as ghost networks. ### Where cash-pay and out-of-network care fit in For patients, the practical consequence is higher out-of-pocket spending. AHRQ Medical Expenditure Panel Survey data show that mental health services have a higher out-of-pocket share than most outpatient medical categories. Some patients pay cash and submit for out-of-network reimbursement; some pay cash and don't submit; and some don't start care because the cost gap is prohibitive. This is a piece of the treatment cascade described in [the U.S. Anxiety Access Report](/reports/us-anxiety-access), and it interacts with wait times covered in the report on [psychiatric wait times](/reports/psychiatric-wait-times). ### What parity law does and doesn't fix The Mental Health Parity and Addiction Equity Act of 2008 requires group health plans that offer mental health and substance use benefits to cover them at parity with medical and surgical benefits. The Consolidated Appropriations Act of 2021 added a requirement that plans conduct and document a comparative analysis of non-quantitative treatment limitations, and the 2024 amendments strengthened enforcement. See the US Department of Labor's [MHPAEA overview](https://www.dol.gov/agencies/ebsa/laws-and-regulations/laws/mental-health-and-substance-use-disorder-parity) for the current framework. Parity law changes what benefits look like on paper. It does not directly change the number of in-network psychiatrists, and enforcement is uneven. The Department of Labor's annual reports to Congress have repeatedly identified compliance gaps, and state insurance regulators have brought actions against insurers whose behavioral-health networks fell short of adequacy tests. ### The network-adequacy problem Even when a plan lists a robust behavioral-health network, patients often can't reach the people in it. Secret-shopper studies calling insurer-published directories have found high rates of wrong numbers, providers not accepting the plan, providers not accepting new patients, or providers no longer in practice. These findings have been documented enough times to have a name: ghost networks. The result is that on-paper access, which is what parity law regulates, can meaningfully overstate real access. ### What this doesn't prove Acceptance-rate surveys measure the practice, not the patient. They don't say why a particular patient couldn't find in-network care, and they don't identify which policy lever moves the number fastest. The Bishop numbers are national averages; the reality varies by state, insurer, and whether behavioral health is carved out to a managed behavioral health organization. Any local number should be read against the local plan and network. ### What the data show - Psychiatry accepts insurance at lower rates than most other medical specialties. The most-cited evidence is a 2014 study by Bishop and colleagues in JAMA Psychiatry, which found that psychiatrists accepted new patients with private non-capitated insurance at 55.3 percent, compared with 88.7 percent for physicians in other specialties. Medicare acceptance among psychiatrists was 54.8 percent versus 86.1 percent for other specialties. Medicaid acceptance among psychiatrists was 43.1 percent versus 73.0 percent. The gap was large enough to be a structural feature of the field rather than a rounding difference. - The reasons are debated but include reimbursement that has not kept pace with the time an evaluation and management visit takes, administrative burden, and demand high enough that psychiatrists can fill their schedules without contracting with insurers. The result is a larger cash-pay and out-of-network segment in psychiatry than in most of medicine. AHRQ Medical Expenditure Panel Survey data on out-of-pocket spending consistently show mental health as a higher-cost category per user than most outpatient care. - Mental health parity law, principally the Mental Health Parity and Addiction Equity Act of 2008 and the Consolidated Appropriations Act of 2021's amendments to it, requires insurers to cover mental health benefits comparably to medical and surgical benefits. Parity has improved coverage on paper, but enforcement and network adequacy remain ongoing concerns. A covered benefit is of little practical use if few local psychiatrists are in network. The US Department of Labor's annual MHPAEA reports to Congress have repeatedly found compliance gaps, and states have taken enforcement action against insurers whose behavioral-health networks failed adequacy tests. ### How to read this - The Bishop study is from 2014, so the exact percentages will have shifted. The direction of the gap, psychiatry below other specialties, has been stable across the available evidence, including newer secret-shopper studies that reach the same conclusion. - "In network" and "accepting new patients" are different questions, and both matter separately. ### Limits - Insurance-acceptance data is collected periodically, not continuously. - Acceptance varies by region, by insurer, and by whether behavioral health is carved out to a managed behavioral health organization. - Accepting insurance and having capacity for new patients aren't the same thing, and a survey that measures one doesn't measure the other. - Directories and network lists are known to overstate available in-network capacity. ### Sources - Bishop et al., JAMA Psychiatry, 2014 (https://pubmed.ncbi.nlm.nih.gov/24337499/) - Mental Health Parity and Addiction Equity Act of 2008 (U.S. Department of Labor) (https://www.dol.gov/agencies/ebsa/laws-and-regulations/laws/mental-health-and-substance-use-disorder-parity) - Kaiser Family Foundation, Mental Health (https://www.kff.org/mental-health/) - Malowney M, et al. Availability of outpatient care from psychiatrists: a secret-shopper survey. Psychiatric Services, 2015. PMID: 25686810 (https://pubmed.ncbi.nlm.nih.gov/25686810/) - AHRQ Medical Expenditure Panel Survey (https://www.ahrq.gov/data/meps.html) - Bishop TF, et al. Population of US Practicing Psychiatrists Declined, 2003-13, Which May Help Explain Poor Access to Mental Health Care. Health Affairs, 2016. PMID: 27385238 (https://pubmed.ncbi.nlm.nih.gov/27385238/) - US Department of Labor, MHPAEA Report to Congress (https://www.dol.gov/agencies/ebsa/laws-and-regulations/laws/mental-health-and-substance-use-disorder-parity) - Consolidated Appropriations Act, 2021 (P.L. 116-260), MHPAEA amendments (https://www.congress.gov/bill/116th-congress/house-bill/133) ## Telepsychiatry and anxiety care URL: https://anxietyresearch.org/reports/telepsychiatry-and-anxiety-care Topic cluster: Access and systems Last updated: 2026-05-18 Telepsychiatry moved from a niche service to the dominant mode of outpatient psychiatric care between 2020 and 2024. Systematic reviews and randomized trials find outcomes broadly comparable to in-person care for assessment, medication management, and most psychotherapies, and use has stayed high long after the initial pandemic surge. The open questions now are regulatory: controlled-substance prescribing, cross-state licensure, and payer parity. ### Key takeaways - Telehealth accounted for roughly one in three behavioral-health visits in the years after 2020, the highest sustained share of any specialty (KFF analysis of Medicare claims). - Systematic reviews find telepsychiatry produces outcomes broadly comparable to in-person care for assessment, medication management, and cognitive behavioral therapy (Hilty et al., 2013; Bashshur et al., 2016). - Access gains have been largest in rural and provider-shortage areas, where telepsychiatry can bridge geographic gaps. - The regulatory picture, especially controlled-substance prescribing under the Ryan Haight Act, is still being worked out through DEA rulemaking. - Broadband access, patient preference, and state-by-state licensure still limit reach. ### What the utilization data show Before 2020, telehealth was a small share of outpatient care and a smaller share of psychiatric care. During the pandemic, Medicare and most commercial payers loosened the rules, and psychiatric telehealth claims rose more than five-fold in a matter of months. Behavioral health has since held the highest telehealth share of any specialty. KFF's analysis of Medicare claims and Mehrotra and colleagues' analyses of commercial claims both find that behavioral-health visits remain about one-third telehealth years after the initial policy relaxation, a level no other specialty has come close to. See the network's [evidence summary on telepsychiatry for anxiety](/evidence-summaries/telepsychiatry-for-anxiety) for the clinical-outcome side of this same question. ### The evidence on outcomes The outcomes literature is now large enough to draw stable conclusions for the common cases. Hilty and colleagues' 2013 review, and Bashshur and colleagues' 2016 update, both concluded that telepsychiatry is generally comparable to in-person care for assessment, diagnosis, medication management, and CBT, with satisfaction levels among patients that match or exceed those of in-person care. More recent meta-analyses of telehealth-delivered CBT for anxiety and depression report effect sizes similar to those of face-to-face delivery. The evidence is strongest for depression and anxiety, weaker for severe mental illness, and mixed for children, where family and school context can matter to the visit. Group therapies and modalities that involve physical interaction are the least well-studied in a telehealth format. ### Where telepsychiatry helps most Telepsychiatry has produced the largest access gains where the in-person option was thinnest. HRSA-designated mental health professional shortage areas cover most rural US counties; telepsychiatry allows a patient in a shortage area to see a licensed psychiatrist who lives elsewhere in the same state. For a national picture of that geography, see the report on [rural and urban access](/reports/rural-and-urban-access). Correctional systems, community mental health centers, and federally qualified health centers have used tele-consultation for the same reason. ### The regulatory picture Two federal frameworks shape what telepsychiatrists can and can't do. The first is the [Ryan Haight Online Pharmacy Consumer Protection Act of 2008](https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title21-section829), which requires an in-person medical evaluation before a controlled substance is prescribed by telemedicine, unless an exception applies. Pandemic-era waivers made most controlled-substance telehealth prescribing possible without an in-person visit, and the DEA has extended those flexibilities through a series of temporary rules while it works on a permanent framework. The rules for stimulants, benzodiazepines, and buprenorphine remain the most closely watched. The second is state medical licensure. Physicians must be licensed in the state where the patient is located at the time of the visit. The Interstate Medical Licensure Compact has made it faster to hold multiple state licenses, but it doesn't waive the requirement. Some states have telehealth-specific registration or waivers; most don't. Payer coverage of audio-only visits, cross-state care, and originating-site rules varies by insurer and by state. ### What telepsychiatry doesn't fix Telepsychiatry expands access; it doesn't create new prescribers. In a shortage area, tele-visits still draw on the national workforce of licensed psychiatrists, which itself is limited and unevenly distributed. Broadband access, device access, and comfort with technology are not universal, and they are correlated with income, age, and geography, which means telepsychiatry can widen access without closing all of the same gaps. For patients whose situation includes psychiatric emergency, imminent risk of harm, medication-assisted treatment for opioid use disorder, or long-acting injectable medications, in-person care remains necessary at some point in the care plan. ### What this doesn't prove The utilization data measure claims; they don't tell you whether the patient got well. The outcomes data measure averages across trials; they don't predict any one person's course. "Comparable outcomes" is a comparison, not a claim that telepsychiatry is preferable in every situation. Long-term outcomes and safety data for controlled-substance prescribing by telemedicine are still accumulating. ### What the data show - Telepsychiatry moved from a niche service to a mainstream mode of care between 2020 and the years that followed. Pandemic-era policy changes temporarily relaxed restrictions on telehealth, and use rose sharply. Mental health was the specialty where telehealth use stayed highest after the initial surge. Utilization data from CMS and SAMHSA show that a substantial share of outpatient psychiatric visits continued to be delivered virtually well after in-person care resumed, and behavioral health has consistently led all specialties in telehealth's share of visits. - The research on outcomes is reassuring. Systematic reviews report that telepsychiatry produces results broadly comparable to in-person care for assessment, medication management, and several psychotherapies, with high patient satisfaction. The access gains have been largest in rural and shortage areas, where telepsychiatry can connect a patient to a clinician who isn't physically nearby. Randomized trials of telehealth-delivered CBT for anxiety report effect sizes similar to face-to-face delivery. - The open questions are regulatory rather than clinical. The prescribing of controlled substances by telemedicine, governed by the federal Ryan Haight Act and by DEA rulemaking, has moved through a series of temporary extensions, and the long-term rules are still being worked out. State licensure remains a constraint, since a clinician must be licensed where the patient is located at the time of the visit. Payer coverage of audio-only visits, cross-state waivers, and originating-site rules vary by insurer and year. ### How to read this - Telehealth utilization shifts with policy. Figures from the pandemic peak aren't a steady-state baseline, and any current share should be read against the policy environment of its year. - "Comparable outcomes" in the outcomes literature means the average result is similar, not that telepsychiatry is right for every patient or every clinical situation. ### Limits - Utilization data lags and varies by payer. - Outcome research covers some conditions and some treatment types better than others; evidence is strongest for depression and anxiety, thinner for severe mental illness and children. - The regulatory picture, especially for controlled-substance prescribing, is still changing. - Access gains presuppose access to broadband and to a device, which are unevenly distributed. ### Sources - SAMHSA (https://www.samhsa.gov/) - CMS telehealth utilization (https://www.cms.gov/) - Hilty et al., Telemedicine and e-Health, 2013 (https://pubmed.ncbi.nlm.nih.gov/23697504/) - Ryan Haight Online Pharmacy Consumer Protection Act of 2008, codified at 21 U.S.C. ยง829 (https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title21-section829) - Bashshur RL, et al. The empirical evidence for telemedicine interventions in mental disorders. Telemedicine and e-Health, 2016. PMID: 26624248 (https://pubmed.ncbi.nlm.nih.gov/26624248/) - Berryhill MB, et al. Videoconferencing psychotherapy and depression: a systematic review. Telemedicine and e-Health, 2019. PMID: 30352016 (https://pubmed.ncbi.nlm.nih.gov/30352016/) - KFF Mental Health issue-brief archive (telehealth utilization and access analyses) (https://www.kff.org/mental-health/) - Mehrotra A, et al. The impact of the COVID-19 pandemic on outpatient care. Commonwealth Fund analyses of visit data (https://www.commonwealthfund.org/publications/2020/apr/impact-covid-19-outpatient-visits) - Interstate Medical Licensure Compact Commission (https://www.imlcc.org/) ## Rural and urban access disparities URL: https://anxietyresearch.org/reports/rural-and-urban-access Topic cluster: Access and systems Last updated: 2026-05-18 Andrilla and colleagues' analyses of the US psychiatric workforce find that roughly 65 percent of nonmetropolitan counties have no practicing psychiatrist. Non-psychiatric prescribers, principally nurse practitioners and primary care physicians, fill much of the gap. Telepsychiatry has expanded reach but has not created new prescribers. The rural-urban gap remains one of the most consistent findings in mental-health workforce research. ### Key takeaways - Roughly 65 percent of US nonmetropolitan counties have no practicing psychiatrist (Andrilla et al., WWAMI Rural Health Research Center analyses). - Nurse practitioners and primary care physicians write the majority of psychiatric medications in rural counties. - HRSA Mental Health Professional Shortage Areas cover a disproportionate share of rural US population. - Telepsychiatry expands the reach of the existing workforce without adding to it (see [telepsychiatry report](/reports/telepsychiatry-and-anxiety-care)). - Broadband access, licensure rules, and workforce distribution together shape real access more than any single number captures. ### The Andrilla workforce analyses Andrilla and colleagues at the University of Washington's WWAMI Rural Health Research Center have produced the most-cited analyses of the US rural mental-health workforce. Their 2018 and follow-up papers document that roughly 65 percent of nonmetropolitan counties have no practicing psychiatrist. The gap is larger for child and adolescent psychiatry, addiction psychiatry, and geriatric psychiatry, all of which concentrate even more heavily in metropolitan areas. ### Who fills the rural prescribing gap Nurse practitioners and primary care physicians write the majority of psychiatric medications in rural counties. In states with independent-practice scope-of-practice rules for nurse practitioners, they can practice without a supervising physician, which expands rural reach. Physician assistants and clinical pharmacists also contribute in some settings. The consequence is that most rural patients with anxiety are treated by clinicians who are not psychiatrists, which puts more weight on primary-care-based tools like the GAD-7 and collaborative-care models (see [anxiety in primary care](/reports/anxiety-and-primary-care)). ### Therapist shortages, too The rural shortage is not limited to prescribers. Licensed clinical social workers, psychologists, and licensed professional counselors are also unevenly distributed. Data on rural therapist supply is less complete than on prescribers, but HRSA workforce reports and state licensure data consistently show under-supply. The consequence is that even when a rural patient can get a prescription, the CBT component of first-line treatment (see [CBT for anxiety](/evidence-summaries/cbt-for-anxiety)) is often not available locally. ### What telepsychiatry does and doesn't do Telepsychiatry allows a licensed clinician in one part of the state to see a patient in another. It expands the reach of the existing workforce, and it has meaningfully improved rural access since 2020. What it cannot do is create new clinicians. The national psychiatric workforce is limited and unevenly distributed, and telehealth alone doesn't change that. Broadband access, device access, and comfort with telehealth are also unevenly distributed and correlate with the same factors that produce the underlying access gap. ### What the data show - The rural and urban gap in anxiety care is one of the most consistent findings in mental health workforce research. Andrilla and colleagues at the WWAMI Rural Health Research Center have documented the pattern in multiple analyses of federal workforce data. Roughly 65 percent of nonmetropolitan counties have no practicing psychiatrist, while psychiatrists concentrate in metropolitan areas at density several times higher than rural areas. The federal HRSA shortage-area designations show the same pattern, with rural counties heavily represented among the areas marked as undersupplied. - The consequences compound. Where there's no psychiatrist, primary care clinicians and psychiatric nurse practitioners manage most anxiety presentations, often without nearby specialty backup for complex cases. Andrilla's work shows that nurse practitioners are increasingly filling the rural mental-health-prescribing gap, especially in states with independent-practice scope-of-practice rules. Patients travel long distances, wait longer, or go without care. Rural areas also tend to have fewer therapists, so the shortage isn't limited to prescribers. - Telepsychiatry has narrowed the gap but hasn't closed it. It depends on broadband access, which is itself unevenly distributed, and on a clinician licensed in the patient's state. Telepsychiatry extends the reach of the existing workforce. It doesn't, on its own, increase the number of clinicians. See the reports on [telepsychiatry](/reports/telepsychiatry-and-anxiety-care) and on [Nebraska's psychiatrist desert](/reports/nebraska-mental-health-access) for adjacent detail. ### How to read this - A county-level count of whether a psychiatrist practices there is a blunt measure. One psychiatrist in a large rural county doesn't mean access is adequate. 'Rural' is also defined differently across studies (nonmetropolitan, Rural-Urban Commuting Area codes, Census-Bureau rural), which affects the headline figures. - Workforce counts capture licensed and practicing providers; whether they are accepting new patients is a separate question. ### Limits - Workforce counts come from periodic analyses with differing methods. - Broadband access and licensure rules shape real access but are hard to capture in a single number. - Shortage-area designations flag undersupply without grading its severity. - Rural mental-health data on therapists is thinner than the data on prescribers. ### Sources - HRSA Health Professional Shortage Areas (https://data.hrsa.gov/topics/health-workforce/shortage-areas) - HRSA Bureau of Health Workforce, mental health workforce reports (https://bhw.hrsa.gov/data-research) - CDC rural health data resources (https://www.cdc.gov/ruralhealth/) - Andrilla CHA, et al. Geographic variation in the supply of selected behavioral health providers. American Journal of Preventive Medicine, 2018. PMID: 29866262 (https://pubmed.ncbi.nlm.nih.gov/29866262/) - Andrilla CHA and Patterson DG. Tracking the Diminishing Supply of Rural Primary Care Physicians in the United States. Journal of Rural Health, 2022. PMID: 34595776 (https://pubmed.ncbi.nlm.nih.gov/34595776/) - HRSA Bureau of Health Workforce, National Center for Health Workforce Analysis (https://bhw.hrsa.gov/data-research) ## Anxiety symptoms by age group URL: https://anxietyresearch.org/reports/anxiety-symptoms-by-age Topic cluster: Epidemiology Last updated: 2026-05-18 Past-year anxiety-disorder prevalence peaks in young adulthood and declines with age in NCS-R data. Adolescent anxiety trends have moved upward over the past decade, a pattern the Surgeon General flagged as a national priority in 2021. Late-life anxiety is real and typically under-recognized; the low survey numbers in older age groups reflect measurement gaps more than absence of the condition. ### Key takeaways - NCS-R past-year prevalence of any anxiety disorder is roughly 22 percent in adults 18-29, roughly 23 percent in 30-44, roughly 21 percent in 45-59, and roughly 10 percent in adults 60+ (Kessler et al., 2005). - CDC Youth Risk Behavior Surveillance System data show adolescent reports of persistent sadness or hopelessness rose from about 30 percent in 2013 to about 42 percent in 2021. - The US Surgeon General's 2021 Advisory on Protecting Youth Mental Health identified anxiety and depression among young people as national priorities. - Late-life anxiety is under-recognized; older adults often present with somatic complaints rather than naming anxiety. - USPSTF's 2023 anxiety-screening recommendation is Grade B for adults 19-64 and an I (insufficient evidence) statement for adults 65+. ### The age gradient in NCS-R The classic NCS-R age curve shows past-year prevalence of any anxiety disorder highest in young adulthood (18-29), roughly stable through middle age (30-59), and lowest in adults 60 and older. The gradient is present in every specific anxiety disorder studied, though the size of the drop-off varies: it's largest for specific phobia and social anxiety and smaller for GAD. ### The adolescent trend and the Surgeon General's advisory CDC YRBSS data have tracked a clear rise in adolescent distress over the past decade. The share of high-school students reporting persistent feelings of sadness or hopelessness rose from about 30 percent in 2013 to about 42 percent in 2021, with the largest increases among girls and among LGBTQ+ students. The Surgeon General's 2021 advisory identified youth mental health as a national priority and framed the trend as a public-health problem rather than a clinical one alone. See the reports on [anxiety and college students](/reports/anxiety-and-college-students) and [anxiety and school absenteeism](/reports/anxiety-and-school-absenteeism) for related pieces. ### Older adults and the measurement problem Late-life anxiety is under-recognized for at least three reasons. First, older adults are more likely to present with somatic complaints (chest pain, GI distress, fatigue) than to name anxiety directly. Second, anxiety in older adults often overlaps with medical illness and depression, complicating diagnosis. Third, most anxiety screeners were validated in younger samples. Wolitzky-Taylor and colleagues' geriatric anxiety reviews consistently argue that the low survey numbers reflect measurement gaps rather than true absence. ### Age effect vs cohort effect Cross-sectional data show today's older adults with lower anxiety rates than today's younger adults. That could mean anxiety fades with age (an age effect) or it could mean today's older adults grew up in an era with lower baseline anxiety exposure (a cohort effect). Longitudinal cohort studies suggest both contribute: individuals do show some decline in symptom levels with age on average, and generations do differ. Neither pattern fully explains the other, and both are consistent with the survey data. ### What the data show - Anxiety isn't spread evenly across the lifespan. In national survey data, anxiety symptoms and anxiety disorders are most common in younger adults and tend to decline with age. Kessler and colleagues' NCS-R analyses show past-year prevalence of any anxiety disorder at roughly 22-23 percent among adults 18-44, roughly 21 percent among adults 45-59, and roughly 10 percent among adults 60 and older. The CDC Household Pulse Survey, which tracks recent symptoms, shows the same age gradient, with the highest rates in younger adults. - Adolescents are a particular focus. The CDC Youth Risk Behavior Surveillance System has documented a rise over the past decade in adolescents reporting persistent feelings of sadness or hopelessness, from about 30 percent in 2013 to about 42 percent in 2021. Surveys of high-school and college students (see the report on [anxiety and college students](/reports/anxiety-and-college-students)) have tracked rising anxiety-related distress. The [US Surgeon General's 2021 Advisory on Protecting Youth Mental Health](https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/index.html) drew this evidence together at the national level. - Older adults present a measurement challenge. Late-life anxiety is real and often goes under-recognized, partly because older adults may report physical symptoms rather than naming anxiety, and partly because anxiety in older adults frequently occurs alongside medical illness and depression. The USPSTF's 2023 anxiety-screening recommendation is Grade B for adults 19-64 and an I (insufficient evidence) statement for adults 65 and older, reflecting the thinner evidence base in older adults rather than a lower true burden. The lower survey numbers in older age groups likely understate the true burden. ### How to read this - Age patterns differ between symptom surveys and diagnostic surveys. Under-recognition of anxiety in older adults likely makes their survey numbers an underestimate rather than a true low. - Cross-sectional surveys can't fully distinguish age effects (how anxiety changes with the individual's age) from cohort effects (how anxiety differs between generations who lived through different eras). ### Limits - Cross-sectional surveys can't fully separate age effects from generational effects. - Adolescent and older-adult measurement each have known gaps. - Symptom screeners and diagnostic interviews produce different age curves. - Older-adult anxiety often co-occurs with cognitive decline and medical illness, complicating measurement. ### Sources - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - CDC Youth Risk Behavior Surveillance System (https://www.cdc.gov/healthyyouth/data/yrbs/index.htm) - CDC Household Pulse Survey, Mental Health (https://www.cdc.gov/nchs/covid19/pulse/mental-health.htm) - U.S. Surgeon General, Advisory on Protecting Youth Mental Health, 2021 (https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/index.html) - Kessler RC, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the NCS-R. Archives of General Psychiatry, 2005. PMID: 15939839 (https://pubmed.ncbi.nlm.nih.gov/15939839/) - Wolitzky-Taylor KB, et al. Anxiety disorders in older adults: a comprehensive review. Depression and Anxiety, 2010. PMID: 20077432 (https://pubmed.ncbi.nlm.nih.gov/20077432/) - CDC Youth Risk Behavior Surveillance System, decade trend reports (https://www.cdc.gov/healthyyouth/data/yrbs/index.htm) ## Anxiety prevalence by sex URL: https://anxietyresearch.org/reports/anxiety-prevalence-by-sex Topic cluster: Epidemiology Last updated: 2026-05-18 The roughly two-to-one female-to-male ratio for anxiety disorders is one of the most consistent findings in psychiatric epidemiology, replicated across surveys, decades, and countries. The explanation is not settled: biological, social, and measurement factors all appear to contribute, and none fully account for the gap on its own. McLean and colleagues' 2011 review is the most-cited synthesis. ### Key takeaways - In US national surveys, women's past-year prevalence of any anxiety disorder is roughly twice that of men (NIMH; NCS-R). - The ratio is largest for specific phobia and generalized anxiety disorder, and smaller for OCD and social anxiety. - The gap appears in early adolescence and persists across adulthood (Kessler et al., 2005). - The pattern is consistent internationally in WHO World Mental Health Surveys. - Contributing factors include biological (hormonal cycling, HPA-axis differences), social (stress exposure, trauma rates), and measurement (reporting differences). ### The size of the gap NCS-R data put women's past-year prevalence of any anxiety disorder at roughly 23 percent versus roughly 14 percent for men, a ratio of about 1.7 to 1. The ratio is largest for specific phobia and generalized anxiety disorder, where it approaches 2 to 3 to 1 in some samples. It is smaller for OCD and social anxiety, and roughly equal for OCD. The pattern is consistent across US and international samples, including WHO World Mental Health Surveys covering 17 countries (Seedat et al., 2009). ### When the gap emerges Sex differences in anxiety are minimal in early childhood, emerge in the pre-pubertal-to-pubertal transition (roughly age 6-11 depending on the anxiety subtype), and persist across adulthood. The developmental timing points to a mix of biological and social contributors: hormonal shifts occur in the same window as changes in social expectations, peer dynamics, and, unfortunately, differential exposure to trauma. ### Candidate explanations Biological candidates include estrogen and progesterone cycling effects on stress response, HPA-axis differences, and sex-specific brain-development trajectories. The evidence for each is moderate; none accounts for the gap alone. Social candidates include differential rates of interpersonal trauma (women have higher rates of sexual assault, which is a risk factor for anxiety and PTSD), differential caregiving and financial stressors, and gender-socialization patterns around expressing distress. Measurement candidates include reporting differences (men may be less likely to endorse anxiety-related items) and stigma differences (help-seeking is more socially costly for men in some contexts). Most researchers argue that all three families of explanations contribute. ### What this doesn't prove Higher prevalence does not mean anxiety is a women's illness. The disorder-level gap is real, but a large share of men also meet criteria for anxiety disorders and receive less care than the base rates would predict. Cross-cultural work shows the female-higher pattern is robust but its size varies, which suggests social context contributes even when biological factors are present. ### What the data show - Across national surveys, women have roughly twice the past-year prevalence of most anxiety disorders compared with men. The pattern is consistent across generalized anxiety disorder, panic disorder, specific phobia, and social anxiety disorder, and it appears in both US and international data. NIMH figures drawing on the National Comorbidity Survey Replication show this roughly two-to-one ratio for any anxiety disorder. Kessler and colleagues' NCS-R analyses documented the sex gap emerging in early adolescence, before puberty in the classical sense but around the time of pubertal transition. - Researchers have proposed several contributing explanations, and the honest summary is that no single one fully accounts for the gap. McLean and colleagues' 2011 review synthesized the candidates. Biological factors, including hormonal influences (particularly estrogen and progesterone cycling, and post-partum shifts) and HPA-axis differences, have been studied and have moderate support. Social factors, including differences in exposure to certain stressors and trauma rates, add explanatory weight. Measurement itself may play a part, because men may be less likely to report anxiety symptoms or to seek care for them, which would widen the apparent gap but does not by itself explain its magnitude. - The sex difference also interacts with treatment. Because women are diagnosed more often, they're also more often represented in treatment settings and in the clinical trials that shape the evidence base. Clinical trials of anxiety treatments have historically been more female-heavy, which means treatment recommendations are anchored to samples in which women are the majority. ### How to read this - The two-to-one ratio is a robust, repeatedly replicated finding. The explanation for it isn't settled, and differences in how men and women report distress may inflate the measured gap. - The ratio applies to disorder-level prevalence; subclinical anxiety symptoms show smaller sex differences. ### Limits - Most surveys record sex in binary categories and don't capture gender identity. - Under-reporting by men is plausible but hard to quantify. - The ratio varies somewhat by specific disorder. - Research on anxiety in transgender and non-binary populations is small and mostly clinical rather than epidemiological. ### Sources - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - Kessler et al., NCS-R, 2005 (https://pubmed.ncbi.nlm.nih.gov/15939837/) - McLean CP, et al. Gender differences in anxiety disorders. Journal of Psychiatric Research, 2011. PMID: 21439576 (https://pubmed.ncbi.nlm.nih.gov/21439576/) - Seedat S, et al. Cross-national associations between gender and mental disorders in the World Health Organization World Mental Health Surveys. Archives of General Psychiatry, 2009. PMID: 19652121 (https://pubmed.ncbi.nlm.nih.gov/19652121/) - Kessler RC, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the NCS-R. Archives of General Psychiatry, 2005. PMID: 15939839 (https://pubmed.ncbi.nlm.nih.gov/15939839/) ## Anxiety and substance use URL: https://anxietyresearch.org/reports/anxiety-and-substance-use Topic cluster: Epidemiology Last updated: 2026-05-18 Anxiety disorders and substance use disorders co-occur at rates several times what chance alone would predict. Alcohol use disorder is the most common co-occurring condition. The self-medication pattern is real, but so is the reverse: heavy substance use, and especially withdrawal, drives anxiety. Integrated treatment outperforms sequential care in the trials that have compared them. ### Key takeaways - In the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-III), adults with a past-year anxiety disorder had roughly two to three times the odds of also meeting criteria for an alcohol use disorder compared with adults without anxiety. - Alcohol use disorder is the most common co-occurring SUD across national surveys. - Cannabis, stimulant, and sedative use disorders also co-occur with anxiety at elevated rates. - Alcohol and benzodiazepine withdrawal reliably produce anxiety and panic-like symptoms. - Integrated care for both conditions produces better outcomes than treating them sequentially in randomized trials. ### How large is the overlap Grant and colleagues' analyses of the National Epidemiologic Survey on Alcohol and Related Conditions consistently find that anxiety disorders co-occur with alcohol use disorder at roughly two to three times the rate expected from base rates alone. The overlap is present in every specific anxiety disorder studied, with somewhat higher odds ratios for generalized anxiety disorder and social anxiety disorder. NSDUH tracks the same pattern for a range of substances. ### Which way does causation run It runs both ways. The self-medication hypothesis, articulated by Khantzian and refined in decades of subsequent work, holds that some people use alcohol or other substances to reduce anxiety in the short term. Longitudinal cohorts, including work by Kushner and colleagues on alcohol and anxiety, support this pattern for some patients. At the same time, alcohol and benzodiazepine withdrawal states produce anxiety and panic-like symptoms directly, so heavy or repeated substance use can drive new-onset anxiety symptoms. In an individual case the history usually points to one direction or the other, but survey data alone can't distinguish them at the population level. ### Substances beyond alcohol Cannabis, stimulants (including prescribed and illicit), and sedatives all co-occur with anxiety at elevated rates. The cannabis picture is mixed: some users report anxiolytic effects, some develop cannabis-induced anxiety, and heavy chronic use is associated with panic. Stimulants can produce or worsen anxiety through their direct pharmacology and through the stress of withdrawal. Sedatives, especially benzodiazepines, produce anxiety on withdrawal that is often misattributed to the underlying disorder returning. ### Why integrated care matters The older model treated substance use first and anxiety later, or the reverse. Randomized comparisons of integrated versus sequential care generally favor integrated approaches, and both SAMHSA (Treatment Improvement Protocol 42) and the American Psychiatric Association support treating co-occurring conditions together. Practically, this can mean combining medication for anxiety with a substance-use-focused therapy, or delivering CBT skills that address both. Benzodiazepine prescribing for anxiety in patients with alcohol or opioid use disorder is generally avoided because of overdose and dependence risks (see the [evidence summary on benzodiazepines](/evidence-summaries/benzodiazepines) for the safety context). ### What the data show - Anxiety disorders and substance use disorders co-occur at high rates. National survey data, including the SAMHSA National Survey on Drug Use and Health and the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), consistently show that people with an anxiety disorder are more likely than the general population to also meet criteria for a substance use disorder, and the reverse is also true. Grant and colleagues' analyses of NESARC-III put the odds of comorbid alcohol use disorder in adults with a past-year anxiety disorder at roughly two to three times that of adults without anxiety, depending on which anxiety disorder is being examined. - The relationship runs in more than one direction. For some people, alcohol or other substances are used to manage anxiety symptoms in the short term, a pattern sometimes called self-medication, which can develop into a substance use disorder over time. For others, heavy substance use, and especially withdrawal, can produce or worsen anxiety. Alcohol withdrawal in particular can produce intense anxiety and panic-like symptoms. Untangling which came first is often not possible from survey data. - Comorbid anxiety and substance use is associated with a harder course than either condition alone. People with both conditions have higher rates of relapse, more functional impairment, and higher healthcare utilization. Clinical guidance increasingly favors treating both conditions together rather than insisting one be resolved before the other is addressed. ### How to read this - Survey data shows the association clearly but can't establish the direction of cause. The two conditions influence each other, and a single person's history may run either way. - The reported overlap depends on how the survey defines substance use disorder and which time window is used. ### Limits - Substance use is under-reported in surveys, which likely understates the overlap. - The measured overlap rate depends on which substances and which diagnostic thresholds are counted. - Most trials of integrated treatment are of modest size and focus on specific substances. ### Sources - SAMHSA National Survey on Drug Use and Health (https://www.samhsa.gov/data/) - National Institute on Drug Abuse (https://nida.nih.gov/) - National Institute on Alcohol Abuse and Alcoholism (https://www.niaaa.nih.gov/) - Grant BF, et al. Epidemiology of DSM-5 Alcohol Use Disorder: Results from NESARC-III. JAMA Psychiatry, 2015. PMID: 26039070 (https://pubmed.ncbi.nlm.nih.gov/26039070/) - Kushner MG, et al. The relationship between anxiety disorders and alcohol use disorders: a review of major perspectives and findings. Clinical Psychology Review, 2000. PMID: 10721495 (https://pubmed.ncbi.nlm.nih.gov/10721495/) - SAMHSA, Substance Use Disorder Treatment for People With Co-Occurring Disorders (Treatment Improvement Protocol 42) (https://store.samhsa.gov/product/tip-42-substance-use-treatment-persons-co-occurring-disorders/pep20-02-01-004) ## Anxiety and chronic illness URL: https://anxietyresearch.org/reports/anxiety-and-chronic-illness Topic cluster: Epidemiology Last updated: 2026-05-18 Anxiety runs high in patients with major chronic illness, at rates roughly two to three times the general population depending on the condition. The relationship is bidirectional and biologically plausible in every direction. It matters clinically because comorbid anxiety worsens adherence, complicates rehabilitation, and, in cardiac and diabetic populations, has been associated with worse medical outcomes. ### Key takeaways - Point-prevalence of clinically significant anxiety symptoms is roughly 20 to 40 percent across cardiovascular disease, diabetes, COPD, and cancer populations, several times the general-population rate. - Anxiety after acute myocardial infarction is associated with worse cardiac outcomes in meta-analyses (Roest et al., 2010). - Diabetes distress and generalized anxiety often coexist and independently affect glycemic control. - Anxiety in COPD is associated with worse exacerbation-related outcomes and lower pulmonary rehabilitation completion. - Screening (typically GAD-7) is now recommended in several chronic-disease care pathways. ### Cardiovascular disease Anxiety and cardiovascular disease have one of the best-documented links. Roest and colleagues' 2010 meta-analysis of prospective post-myocardial-infarction cohorts found that patients with elevated anxiety had roughly 1.4 times the risk of subsequent cardiac events and mortality compared with non-anxious patients, adjusting for known predictors. Panic disorder in particular has been studied as a driver of emergency-department utilization (see the report on [panic and emergency care](/reports/panic-and-emergency-care)). Mechanistic candidates include altered autonomic function, inflammation, and behavioral pathways such as adherence to prevention regimens. ### Diabetes Diabetes distress, a construct related to but distinct from generalized anxiety, is common in adults living with diabetes. Meta-analyses find elevated rates of anxiety disorders in adults with type 1 and type 2 diabetes compared with the general population, and anxiety is independently associated with poorer glycemic control in longitudinal studies. Screening for anxiety and depression is now embedded in several US diabetes care standards. ### COPD Anxiety is unusually common in COPD, in part because the physical experience of breathlessness overlaps directly with the physical experience of panic. Meta-analyses estimate clinically significant anxiety in roughly a third of COPD patients. Anxiety in COPD is associated with higher rates of exacerbation-related hospitalization and lower completion of pulmonary rehabilitation, both of which affect long-term outcomes. ### Cancer Anxiety around cancer diagnosis, treatment, and follow-up is common and, at moderate levels, appropriate to the situation. Distinguishing clinically significant anxiety disorders from adjustment reactions is one of the practical challenges in oncology psychology. NCCN's distress-thermometer approach and disease-specific psycho-oncology guidelines have moved toward routine screening, and integrated psycho-oncology services are increasingly standard at cancer centers. ### Chronic pain Anxiety and chronic pain co-occur at rates well above chance and reinforce each other prospectively in cohort studies. Bair and colleagues' systematic reviews consistently find bidirectional prospective associations. The overlap complicates both diagnosis (pain and anxiety share physical features) and treatment (some medications used for one also affect the other, with a mix of helpful and complicating results). ### What this doesn't prove The associations are consistent across large observational datasets. What the observational data cannot prove is that treating anxiety, on its own, changes the medical outcome. Some randomized trials of collaborative care for anxiety plus depression in chronic disease clinics have found benefits on symptom measures and functional outcomes; medical-outcome effects are more modest and less consistent. ### What the data show - Anxiety occurs more often in people with chronic medical conditions than in the general population. The pattern has been documented for cardiovascular disease, diabetes, chronic obstructive pulmonary disease, cancer, and chronic pain, among others. The elevated prevalence is consistent enough across conditions that the link isn't specific to any one illness. Point-prevalence of clinically significant anxiety in these populations commonly falls in the 20 to 40 percent range, several times the general-population rate. - Several mechanisms likely contribute, and they aren't mutually exclusive. A serious diagnosis and the daily burden of managing a chronic illness are themselves significant stressors. Some conditions and some medications produce physical symptoms that overlap with anxiety, such as a racing heart or breathlessness. And shared biological pathways, including inflammation and stress-system activity, have been studied as common ground between anxiety and several chronic diseases. - Comorbid anxiety matters for the medical condition, not only for mental health. Roest and colleagues' 2010 meta-analysis of post-myocardial-infarction cohorts found that anxiety was associated with increased risk of subsequent cardiac events. Anxiety is associated with lower adherence to medical treatment, higher health-care use, and, in some conditions, worse medical outcomes. This is why screening for anxiety, typically with the GAD-7, is increasingly recommended within the care of several chronic diseases. ### How to read this - The association is well established. The direction of cause and the mechanism vary by condition and are still being studied. Symptom overlap between anxiety and physical illness makes precise measurement difficult. - The medical significance of comorbid anxiety differs by condition; anxiety after myocardial infarction, for example, has stronger prognostic data than anxiety in some other chronic conditions. ### Limits - Most of the evidence is observational. - The overlap of physical and anxiety symptoms complicates diagnosis (breathlessness, tachycardia, GI distress). - Estimates vary by condition and by how anxiety is measured (screening tool vs diagnostic interview). - Randomized trials of anxiety treatment in chronic-illness populations are smaller and less common than descriptive studies. ### Sources - CDC, data on chronic disease and mental health (https://www.cdc.gov/chronicdisease/) - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - American Psychiatric Association (https://www.apa.org/) - Roest AM, et al. Anxiety and risk of incident coronary heart disease: a meta-analysis. Journal of the American College of Cardiology, 2010. PMID: 20620710 (https://pubmed.ncbi.nlm.nih.gov/20620710/) - Roy-Byrne PP, et al. Anxiety disorders and comorbid medical illness. General Hospital Psychiatry, 2008. PMID: 18291291 (https://pubmed.ncbi.nlm.nih.gov/18291291/) - Bair MJ, et al. Impact of pain on depression treatment response in primary care. Psychosomatic Medicine, 2004. PMID: 14747644 (https://pubmed.ncbi.nlm.nih.gov/14747644/) - Spitzer RL, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of Internal Medicine, 2006. PMID: 16717171 (https://pubmed.ncbi.nlm.nih.gov/16717171/) ## Anxiety and depression overlap URL: https://anxietyresearch.org/reports/anxiety-and-depression-overlap Topic cluster: Epidemiology Last updated: 2026-05-18 Roughly two out of three adults with current major depression also meet criteria for at least one anxiety disorder. The overlap is one of the most consistent findings in psychiatric epidemiology, and comorbid presentations are consistently harder to treat than either condition alone. Fortunately, the first-line treatments for anxiety and depression overlap almost completely, which simplifies planning even when the presentation is complex. ### Key takeaways - About 60 to 70 percent of adults with a current major depressive episode also meet criteria for at least one anxiety disorder (Lamers et al., NESDA, 2011; NCS-R analyses). - Lifetime co-occurrence is even higher; most adults who ever have major depression will also meet criteria for an anxiety disorder at some point. - Comorbid presentations are associated with greater impairment, longer time to remission, and higher relapse risk in longitudinal studies. - First-line treatments overlap: SSRIs, SNRIs, and CBT all carry evidence for both conditions. - Sequenced Treatment Alternatives to Relieve Depression (STAR*D) found that patients with anxious depression had lower remission rates and slower response. ### How the overlap is measured The overlap rate depends on what you count and when. Point-in-time comorbidity, meaning both conditions present at the same time, is roughly 60 to 70 percent among adults with current major depression. Lifetime comorbidity, meaning both conditions at some point in a person's history, is higher: NCS-R analyses put it at roughly 75 percent for major depression with any anxiety disorder. The single most-cited current-comorbidity paper is Lamers and colleagues' 2011 analysis of the Netherlands Study of Depression and Anxiety (NESDA). ### What comorbidity means for course and prognosis The STAR*D study, the largest US effectiveness trial in depression treatment, followed more than 4,000 adults through sequential antidepressant treatment. Fava and colleagues found that anxious depression, defined by a high anxiety-symptom score at baseline, was associated with lower remission rates, slower response, and higher side-effect burden than non-anxious depression, at every level of treatment. Longitudinal cohorts consistently show longer episode duration, more functional impairment, and higher relapse rates in comorbid presentations. ### Why the treatments overlap SSRIs and SNRIs are first-line for both major depression and every common anxiety disorder in US and international guidelines. CBT is first-line for both, though the specific protocols differ (exposure-focused for anxiety, behavioral-activation and cognitive-restructuring emphasis for depression). This makes treatment planning simpler for comorbid presentations than for other combinations. When one first-line option works for both conditions, sequencing becomes a matter of clinical fit rather than choosing between two different treatment traditions. ### What this doesn't prove The overlap describes group-level averages. It doesn't establish that anxiety and depression share a single underlying cause. The evidence supports a dimensional view in which shared risk factors, including neuroticism and shared environmental stressors, contribute to both, alongside condition-specific factors. Whether anxiety and depression represent one disorder with two presentations, or two disorders with overlapping mechanisms, remains debated in the research literature. ### What the data show - Anxiety and depression overlap heavily. National survey data show that a large share of adults who meet criteria for a major depressive episode in a given year also meet criteria for at least one anxiety disorder, and a large share of people with an anxiety disorder experience depression. The National Comorbidity Survey Replication, the survey that gives the word 'comorbidity' its place in this field, documented the overlap in detail, and Lamers and colleagues' analysis of the Netherlands Study of Depression and Anxiety put the current comorbidity at roughly two-thirds of adults with major depression. - The overlap isn't only common, it's consequential. Comorbid anxiety and depression is associated with greater symptom severity, more functional impairment, a longer time to remission, and a higher risk of relapse than either condition alone. The STAR*D study, the largest US effectiveness trial of sequential antidepressant treatment, found that anxious depression, meaning depression with a high anxiety-symptom score, was associated with lower remission rates and slower response across levels of treatment. The two conditions also share symptoms, including sleep disturbance, difficulty concentrating, and irritability, which can make them hard to separate in an individual presentation. - Because the overlap is so common, treatment guidance increasingly favors approaches that address both conditions. The first-line treatments overlap as well, since SSRIs, SNRIs, and CBT all have evidence in both anxiety and depression, which simplifies treatment when the two co-occur. See the [evidence summaries on CBT for anxiety](/evidence-summaries/cbt-for-anxiety) and [SSRIs and SNRIs for anxiety](/evidence-summaries/ssris-snris-for-anxiety) for the treatment-side detail. ### How to read this - Estimates of the overlap vary with definitions, time windows, and survey methods. Read the overlap as a strong and consistent pattern rather than a single fixed rate. - Anxious depression is a specification of major depression with prominent anxiety features; the DSM-5-TR includes it as a modifier. ### Limits - Shared symptoms complicate measurement. - Survey estimates depend on the diagnostic instrument used. - The overlap rate differs between point-in-time and lifetime measures. - Most treatment trials focus on one disorder as primary; direct trials of integrated treatment approaches are less common. ### Sources - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - Kessler et al., NCS-R, 2005 (https://pubmed.ncbi.nlm.nih.gov/15939837/) - Lamers F, et al. Comorbidity patterns of anxiety and depressive disorders in a large cohort study: the Netherlands Study of Depression and Anxiety (NESDA). Journal of Clinical Psychiatry, 2011. PMID: 21495080 (https://pubmed.ncbi.nlm.nih.gov/21495080/) - Fava M, et al. Difference in treatment outcome in outpatients with anxious versus nonanxious depression: a STAR*D report. American Journal of Psychiatry, 2008. PMID: 18172020 (https://pubmed.ncbi.nlm.nih.gov/18172020/) - Kessler RC, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the NCS-R. Archives of General Psychiatry, 2005. PMID: 15939839 (https://pubmed.ncbi.nlm.nih.gov/15939839/) ## Anxiety symptoms in college students URL: https://anxietyresearch.org/reports/anxiety-and-college-students Topic cluster: Epidemiology Last updated: 2026-05-18 Anxiety is now the most common presenting concern in college mental health, ahead of depression and relationship problems. Symptom rates on the Healthy Minds Study and ACHA-NCHA have roughly doubled since 2013, and campus counseling utilization has grown faster than staffing. The combined picture is real demand plus rising help-seeking, and the two together strain a system built for a smaller share of students. ### Key takeaways - Anxiety has been the most common presenting concern at college counseling centers since roughly the mid-2010s (CCMH annual reports). - The Healthy Minds Study finds moderate-to-severe anxiety symptoms in roughly a third of college students in recent waves, up from about 17 percent in 2013. - Campus counseling caseloads have grown 30 to 40 percent faster than enrollment over the past decade (CCMH). - Help-seeking has increased more than symptom prevalence has, meaning the measured rise reflects both more anxiety and more willingness to name and treat it. - The Surgeon General has called youth mental health a national priority (2021 Advisory). ### What the surveys measure Two big surveys anchor the picture. The Healthy Minds Study (University of Michigan) samples students from participating colleges each year and uses standardized screeners including the GAD-7 for anxiety. The ACHA National College Health Assessment uses a broader health survey with an anxiety module. Both find rising rates over the past decade and reasonably similar current levels, roughly a third of students reporting significant anxiety symptoms. See the report on [anxiety symptoms by age](/reports/anxiety-symptoms-by-age) for how this fits into the broader lifespan pattern. ### What the campus counseling data show The Center for Collegiate Mental Health's annual report, drawn from more than 150 college counseling centers, has tracked utilization since 2010. Two findings are consistent: anxiety has been the most common presenting concern for years, and caseloads have grown faster than enrollment. Many centers have moved to session limits, brief-treatment models, or triage systems to manage demand. This has knock-on effects on wait times and referral to off-campus care. ### How much of the rise is real anxiety vs more help-seeking Both. Symptom-level surveys of unselected student samples show real increases in the fraction of students meeting screener cutoffs for anxiety, which suggests a genuine underlying rise. At the same time, help-seeking rates have grown faster than symptom prevalence has, meaning a larger share of students with anxiety now reach services. Reduced stigma, wider awareness, and expanded campus resources probably explain most of that. Both trends push measured caseloads up, and both are worth taking seriously. ### What this doesn't prove Rising symptoms on a screener are not the same as rising disorder prevalence in a diagnostic sense; symptom scales are sensitive to recent context, mood, and reporting norms. Findings from four-year college populations may not generalize to community college students, students at minority-serving institutions, or non-student young adults of the same age. Comparing years requires comparing the same instrument administered the same way, which is not always possible. ### What the data show - Anxiety is now the most common concern that brings college students to campus mental health services. Surveys of US college students, including the Healthy Minds Study and the American College Health Association's national assessment (ACHA-NCHA), have tracked rising rates of anxiety symptoms among students over the past decade and a rising share of students seeking help. Healthy Minds' data show moderate-to-severe anxiety symptoms in roughly a third of students in recent waves, up from about 17 percent in 2013. - Demand has grown faster than the capacity to meet it. Campus counseling centers report rising utilization that outpaces staffing, which produces waitlists and session limits at many institutions. The Center for Collegiate Mental Health's annual reporting has documented this capacity gap, with caseloads growing 30 to 40 percent faster than enrollment over the past decade. - Several explanations are discussed, and they likely act together. The college-age years are a peak period for the onset of anxiety disorders, independent of any recent trend. Reduced stigma means more students are willing to name anxiety and seek care, which raises measured rates even where the underlying condition is unchanged. Broader pressures on young adults, including academic and financial stress, have also been proposed as contributors, and the [Surgeon General's 2021 Advisory on Protecting Youth Mental Health](https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/index.html) drew this evidence together at the national level. The data show the rise clearly. The full mix of causes is still being worked out. ### How to read this - Rising measured rates reflect both more anxiety and more help-seeking. The two are difficult to separate, so a rising number isn't, on its own, proof of a rising disorder rate. - Campus utilization data measure the students who reach the counseling center, not the total need on campus. ### Limits - College survey samples vary by institution and response rate. - Findings from student samples may not generalize to non-students of the same age. - Utilization data reflects services used, not total need. - Symptom screeners (like GAD-7) are not diagnostic; they measure recent symptom burden. ### Sources - Healthy Minds Study (https://healthymindsnetwork.org/) - American College Health Association, NCHA (https://www.acha.org/NCHA) - Center for Collegiate Mental Health, Penn State (https://ccmh.psu.edu/) - U.S. Surgeon General, Advisory on Protecting Youth Mental Health, 2021 (https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/index.html) - Lipson SK, et al. Trends in college student mental health and help-seeking. Journal of Affective Disorders, 2022. PMID: 34600966 (https://pubmed.ncbi.nlm.nih.gov/34600966/) ## Panic attacks and emergency-care use URL: https://anxietyresearch.org/reports/panic-and-emergency-care Topic cluster: Healthcare utilization Last updated: 2026-05-18 Panic disorder is one of the most common reasons anxiety touches an emergency department. Panic-attack symptoms mirror a heart attack, so a first episode reasonably leads to an ED visit, and a substantial share of chest-pain visits with no cardiac cause turn out to be panic. Recognition of panic in the ED is uneven, and repeat visits are common when the first visit doesn't connect the patient to outpatient care. ### Key takeaways - Chest pain is one of the top reasons for adult ED visits in the United States (roughly 6-7 million ED visits per year, CDC NHAMCS). - Studies of chest-pain patients with no acute cardiac cause find that a substantial share, roughly a quarter to a third in specific samples, meet criteria for panic disorder (Fleet et al., American Journal of Medicine, 1996; Katerndahl, Journal of Nervous and Mental Disease, 2004). - Panic disorder in the ED is often under-recognized; patients frequently leave without a psychiatric diagnosis or referral. - Untreated panic is associated with high healthcare utilization, including repeat ED visits, until effective outpatient treatment starts. - First-line treatment for panic disorder is CBT with interoceptive exposure and/or an SSRI (see [evidence summary](/evidence-summaries/panic-disorder-treatment)). ### Why panic ends up in the ED A panic attack produces intense autonomic symptoms: chest tightness, palpitations, shortness of breath, a feeling of impending doom. The physical experience is indistinguishable from a cardiac event without a workup. A person having a first attack, especially at rest or overnight, will reasonably go to the ED. Chest pain is the second-most-common reason for adult ED visits after abdominal pain in NHAMCS. The evaluation is medically necessary; the ED is doing the right thing to work it up. ### How much of non-cardiac chest pain is panic Fleet and colleagues' 1996 study of an ambulatory cardiology clinic sample found panic disorder in roughly a quarter of patients with non-cardiac chest pain. Katerndahl's subsequent work in family medicine and ED settings found similar rates. Not every patient with non-cardiac chest pain has panic, and not every patient with panic will meet criteria for panic disorder, but the fraction is large enough that panic should be on the differential once cardiac and other medical causes are ruled out. ### What happens after the ED visit Cost and utilization studies consistently find that untreated panic disorder is associated with high healthcare utilization: repeat ED visits, extensive workups, and multiple medical specialty consultations. Effective outpatient treatment, specifically CBT with interoceptive exposure or an SSRI, reduces this utilization pattern in cohort studies. The ED-to-outpatient linkage is the point where the pathway commonly breaks: many patients leave the ED without a psychiatric diagnosis, referral, or follow-up plan. ### What this doesn't prove Association between panic and repeat ED use is not proof that missed diagnosis is the sole driver. Some patients have both cardiac disease and panic. Some panic-like symptoms are due to thyroid disease, stimulants, alcohol withdrawal, or arrhythmias, and reasonable workup is required. The ED is not the setting where panic disorder is best treated; the point of good ED care is stabilizing the patient, ruling out medical causes, and connecting them to outpatient care. ### What the data show - Panic and acute anxiety are a recognized driver of emergency department visits. A panic attack produces intense physical symptoms, including chest pain, a pounding heart, shortness of breath, and a feeling of dread, and those symptoms closely resemble a cardiac emergency. Many people experiencing a first panic attack reasonably go to an emergency department, and chest pain is one of the most common reasons for emergency visits overall, with roughly 6 to 7 million adult ED visits per year for chest pain per CDC's National Hospital Ambulatory Medical Care Survey. - Research on emergency care for panic, including Fleet and colleagues' influential 1996 American Journal of Medicine study and Katerndahl's follow-up work, shows two consistent patterns. First, a meaningful share of patients who present with chest pain and no cardiac cause are experiencing panic or anxiety, and panic disorder is under-recognized in that setting. Fleet's cardiology-clinic sample found panic disorder in roughly a quarter of patients with non-cardiac chest pain. Second, an emergency visit often doesn't connect the patient to ongoing outpatient care, so repeat visits are common when the underlying anxiety isn't addressed. - The clinical point cuts both ways. New cardiac-type symptoms should be medically evaluated and not assumed to be anxiety, especially a first episode. Once a cardiac and other medical workup is clear, recognizing panic and linking the patient to treatment can prevent a cycle of repeat emergency visits. ### How to read this - National emergency-visit data changes with each release of the underlying survey. Any specific volume figure should be cited from the most recent year available rather than carried forward. - Panic disorder is a specific DSM-5-TR diagnosis; not every panic attack means a panic disorder. Isolated panic attacks are common; the disorder requires recurrent attacks plus persistent worry or behavior change. ### Limits - Emergency datasets record visit reasons and discharge diagnoses imperfectly. - Panic is both under-coded, when it's missed, and sometimes assumed without a full workup. The two errors push the count in opposite directions. - Fleet-era percentages are from cardiology-clinic samples and may not generalize to all EDs. - Randomized trials of ED-based recognition and referral pathways for panic are limited in size. ### Sources - CDC National Hospital Ambulatory Medical Care Survey (https://www.cdc.gov/nchs/ahcd/index.htm) - American College of Emergency Physicians (https://www.acep.org/) - Fleet RP, et al. Panic disorder in emergency department chest pain patients: prevalence, comorbidity, suicidal ideation, and physician recognition. American Journal of Medicine, 1996. PMID: 8629651 (https://pubmed.ncbi.nlm.nih.gov/8629651/) - Katerndahl DA. Panic and prevalence: an overview of the physical and psychiatric findings in panic disorder. Journal of Nervous and Mental Disease, 2004. PMID: 15043128 (https://pubmed.ncbi.nlm.nih.gov/15043128/) - Katerndahl DA, Realini JP. Where do panic attack sufferers seek care? Journal of Family Practice, 1995. PMID: 7562719 (https://pubmed.ncbi.nlm.nih.gov/7562719/) ## Anxiety in primary care URL: https://anxietyresearch.org/reports/anxiety-and-primary-care Topic cluster: Healthcare utilization Last updated: 2026-05-18 Primary care is where most US anxiety care actually happens. The primary care clinician is the most common first point of contact for adults with anxiety symptoms, and non-psychiatric prescribers write the majority of anxiety prescriptions. Recognition is uneven, treatment intensity varies widely, and collaborative care is the best-studied response, though it isn't yet the default. ### Key takeaways - The US Preventive Services Task Force recommends anxiety screening in adults age 19-64 (Grade B recommendation, 2023). - The GAD-7 is a validated 7-item screener with sensitivity around 89 percent and specificity around 82 percent at a cutoff of 10 for GAD (Spitzer et al., 2006). - Non-psychiatric clinicians write the majority of US antidepressant prescriptions, per MEPS analyses. - Collaborative care, tested in the CALM randomized trial (Roy-Byrne et al., 2010), improves anxiety outcomes in primary care. - The primary care visit averages about 15-20 minutes, which limits therapy delivery in that setting. ### Why primary care is the anxiety front door Two structural facts drive this. First, most Americans have a primary care clinician and can reach them faster than they can reach a psychiatrist (see the report on [psychiatric wait times](/reports/psychiatric-wait-times)). Second, HRSA shortage-area data show that most rural US counties have no practicing psychiatrist, which leaves the primary care clinician as the only option for many patients (see [rural and urban access](/reports/rural-and-urban-access)). By design or by necessity, primary care carries the bulk of the load. ### The GAD-7 and how it's used The GAD-7 is a seven-item self-report screener developed by Spitzer and colleagues in 2006. A score of 10 or higher is the standard cutoff for likely generalized anxiety disorder, with sensitivity around 89 percent and specificity around 82 percent in the original validation. It also has reasonable performance for panic disorder, social anxiety, and PTSD, though it was designed for GAD. Use in primary care is embedded in many electronic health records and is one of the practical tools primary care clinicians have to catch anxiety that otherwise presents as fatigue, insomnia, or somatic complaints. ### USPSTF's 2023 anxiety-screening recommendation In 2023 the US Preventive Services Task Force issued a Grade B recommendation for screening adults age 19-64 for anxiety, based on evidence that screening in primary care can identify unrecognized anxiety and that treatment initiated after screening improves symptoms. The recommendation is for adults not already known to have an anxiety disorder. USPSTF issued an I (insufficient evidence) statement for adults 65 and older, reflecting the thinner evidence base in older adults. See the [USPSTF recommendation](https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/anxiety-adults-screening) for the full framework. ### Collaborative care: the evidence Collaborative care embeds a behavioral-health care manager in the primary care team, with weekly caseload review by a consulting psychiatrist. The care manager conducts brief therapy and tracks outcomes with a validated measure. The most-cited anxiety-specific trial is CALM (Roy-Byrne et al., JAMA, 2010), which randomized primary care patients with panic disorder, GAD, social anxiety, or PTSD to collaborative care or usual care and found substantially better anxiety outcomes at 6 and 12 months. Collaborative care for depression has an even larger evidence base and is standard in some health systems; the anxiety-specific rollout has been slower. ### What this doesn't prove Screening finds anxiety; it does not, on its own, treat it. The value of screening depends on what happens after a positive screen: whether the clinician follows up, whether the patient can access treatment, and whether treatment is delivered at adequate intensity. The primary-care share of anxiety prescriptions is a structural feature of the system rather than a claim about optimal care, and any specific figure depends on the data source and the year. ### What the data show - Primary care, not the psychiatrist's office, is where most anxiety care in the United States happens. Primary care clinicians are the most common first point of contact for adults seeking help with anxiety, and they prescribe a large share of the medication used to treat it. This is partly by design and partly by necessity. In much of the country, especially rural areas, there's no psychiatrist within reach, and the primary care clinician is the only option. - Two issues shape anxiety care in this setting. The first is recognition. Anxiety can be missed in a brief primary care visit, especially when a patient presents with physical complaints such as fatigue, sleep problems, or pain. Brief validated screening tools, principally the GAD-7 (Spitzer et al., 2006, with sensitivity around 89 percent and specificity around 82 percent at a cutoff of 10 for GAD), improve detection. The US Preventive Services Task Force has recommended screening adults age 19-64 for anxiety with a Grade B recommendation. The second is depth. A primary care visit is short, which limits the time available for the structured talk-therapy components that the evidence supports. - Models that support primary care, including collaborative care, in which a care manager and a consulting psychiatrist back up the primary care clinician, have evidence for improving outcomes. Roy-Byrne and colleagues' 2010 CALM trial (JAMA) randomized primary care patients with anxiety disorders to collaborative care versus usual care and found meaningfully better anxiety outcomes at 6 and 12 months. Collaborative care is one response to the structural reality that primary care carries most of the load. ### How to read this - The statement that most anxiety care happens in primary care is a well-supported structural fact. Exact figures for the share of prescriptions written by non-psychiatrists depend on the dataset and the year. - GAD-7 is a screening tool, not a diagnostic instrument; a positive screen indicates further evaluation is warranted, not a diagnosis. ### Limits - Primary care data varies by source. - Recognition rates are hard to measure directly (chart review versus screening at scale gives different answers). - Collaborative care is well studied but isn't widely available across the country. - USPSTF's 2023 anxiety recommendation is for adults 19-64; the anxiety screening recommendation for older adults remains "insufficient evidence" (I statement). ### Sources - U.S. Preventive Services Task Force, recommendation on screening for anxiety in adults (https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/anxiety-adults-screening) - AHRQ Medical Expenditure Panel Survey (https://www.ahrq.gov/data/meps.html) - Spitzer RL, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of Internal Medicine, 2006. PMID: 16717171 (https://pubmed.ncbi.nlm.nih.gov/16717171/) - Roy-Byrne P, et al. Delivery of evidence-based treatment for multiple anxiety disorders in primary care: a randomized controlled trial (CALM). JAMA, 2010. PMID: 20483968 (https://pubmed.ncbi.nlm.nih.gov/20483968/) - Kroenke K, et al. Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 2007. PMID: 17339617 (https://pubmed.ncbi.nlm.nih.gov/17339617/) ## Psychiatric hospitalization trends URL: https://anxietyresearch.org/reports/psychiatric-hospitalization-trends Topic cluster: Healthcare utilization Last updated: 2026-05-18 US public psychiatric inpatient capacity has fallen by roughly 95 percent from its 1955 peak. Community capacity did not expand to match, and the consequences show up as ED boarding, high readmission rates, and a system that struggles to place patients in crisis. Anxiety disorders are rarely a primary reason for admission, but severe anxiety frequently accompanies the conditions that are. ### Key takeaways - Public psychiatric inpatient beds fell from about 340 per 100,000 in 1955 to roughly 11-12 per 100,000 in recent years, a decline of over 95 percent per Treatment Advocacy Center analyses. - The 30-day readmission rate after psychiatric hospitalization is roughly 15-20 percent in AHRQ HCUP analyses. - Psychiatric boarding, patients waiting hours to days in EDs for an inpatient bed, is documented as a persistent system problem by ACEP. - Anxiety is rarely a primary discharge diagnosis for psychiatric admission but is often a comorbid diagnosis alongside major depression, psychosis, or substance use. ### The long decline of inpatient capacity In 1955, US public psychiatric hospitals held roughly 340 beds per 100,000 population. By the 2010s and 2020s, that figure had fallen below 12 per 100,000 in most Treatment Advocacy Center analyses. The decline reflects deinstitutionalization, which began in the mid-twentieth century for reasons that were partly humanitarian (large state hospitals were often warehousing rather than treating patients) and partly practical (the introduction of chlorpromazine and other antipsychotics made outpatient management possible for many patients). The intent was to shift care to community mental health centers. The construction of that community system was never funded at the scale originally proposed. ### Boarding: where the capacity gap shows up When someone in psychiatric crisis presents to an ED and needs inpatient admission, they may wait in the ED for hours or days while staff search for an available bed. This is boarding. ACEP's polling and single-system studies consistently document longer boarding times for psychiatric patients than for medical patients, with disproportionate impact on pediatric patients, patients with public insurance, and patients from underserved communities. Boarding is bad for patients, bad for ED throughput, and one of the clearest measurable consequences of the inpatient capacity gap. ### Readmission and what it says about outpatient care AHRQ HCUP analyses put the 30-day readmission rate after psychiatric hospitalization in the 15 to 20 percent range, higher than the medical average. High psychiatric readmission generally points at problems in the transition from inpatient to outpatient care: gaps in medication continuation, delayed follow-up appointments, and difficulty accessing outpatient prescribers (see the report on [psychiatric wait times](/reports/psychiatric-wait-times) for the outpatient supply side). ### Where anxiety fits Anxiety disorders are rarely the primary reason for psychiatric admission on their own. Admission is typically driven by suicidal or homicidal risk, severe depression, psychosis, mania, or acute substance-use crises. Anxiety enters the picture through comorbidity: severe anxiety frequently accompanies the conditions that do lead to admission (see [anxiety and depression overlap](/reports/anxiety-and-depression-overlap) and [anxiety and substance use](/reports/anxiety-and-substance-use)). Panic disorder is more relevant to ED utilization than to inpatient admission (see [panic and emergency care](/reports/panic-and-emergency-care)). ### What the data show - The United States has far fewer psychiatric inpatient beds than it did in the mid-twentieth century. Analyses by the Treatment Advocacy Center and others put the public psychiatric bed count at roughly 340 per 100,000 population in 1955 and roughly 11 to 12 per 100,000 in recent years, a decline of over 95 percent. The long decline in inpatient capacity, sometimes called deinstitutionalization, was driven by policy changes, the introduction of new medications, and a shift toward community-based care. The intent was sound, but community capacity didn't expand to match, which left a gap. - That gap shows up in emergency departments. When a person in psychiatric crisis needs admission and no bed is available, they may wait in an emergency department for hours or days, a situation known as psychiatric boarding. The American College of Emergency Physicians has documented boarding as a persistent system problem across most US EDs, with longer boarding times for pediatric patients and for patients with public insurance. - Anxiety disorders aren't, in most cases, a leading cause of psychiatric hospitalization on their own. Admission is more often driven by risk of harm, severe depression, psychosis, or crises involving comorbid conditions. But anxiety is relevant to the inpatient picture through comorbidity, since severe anxiety frequently accompanies the conditions that do lead to admission. AHRQ HCUP data show 30-day psychiatric readmission rates in the 15 to 20 percent range in recent analyses, which points to gaps in the outpatient care that follows discharge. ### How to read this - Inpatient capacity and boarding data come from several sources with differing methods. Treat the figures as directional rather than exact. Anxiety's role here is mostly indirect, through comorbidity. - "Psychiatric bed" is defined differently across sources: state hospital beds, private psychiatric hospital beds, and general-hospital psychiatric units are counted differently. ### Limits - National inpatient datasets lag by 1-2 years. - The term 'psychiatric bed' is defined differently across sources. - Anxiety's contribution is hard to isolate because it usually acts through comorbid conditions. - Boarding data are collected by specialty societies and single-system studies, not from a national registry. ### Sources - American Hospital Association inpatient capacity reports (https://www.aha.org/) - AHRQ HCUP (https://www.hcup-us.ahrq.gov/) - American College of Emergency Physicians (https://www.acep.org/) - Treatment Advocacy Center, reports on psychiatric inpatient capacity trends (https://www.tac.org/reports_publications/) - American College of Emergency Physicians, ED boarding polling and position statements (https://www.acep.org/) - AHRQ HCUP Statistical Briefs on psychiatric hospitalizations and readmissions (https://www.hcup-us.ahrq.gov/reports/statbriefs/statbriefs.jsp) ## Anxiety and work impairment URL: https://anxietyresearch.org/reports/anxiety-and-work-impairment Topic cluster: Functional impact Last updated: 2026-05-18 Anxiety disorders sit in the top tier of causes of workplace disability in the United States. Greenberg and colleagues' original 1999 US burden-of-illness estimate put the annual anxiety-disorder cost at about $42 billion, with workplace productivity the largest share. Updated analyses through the 2010s and 2020s haven't reversed that framing. WHO burden-of-disease work places anxiety and depression together as leading causes of years lived with disability globally. ### Key takeaways - Greenberg and colleagues' 1999 US burden-of-illness estimate put the total cost of anxiety disorders at about $42 billion per year (1990 dollars), with workplace productivity the largest share. - Presenteeism (reduced productivity while at work) is often the larger component of workplace cost than absenteeism. - In WHO Global Burden of Disease analyses, anxiety disorders are among the top causes of years lived with disability worldwide. - Mental health conditions are a growing share of Social Security Disability Insurance awards over the past two decades. - Effective anxiety treatment reduces work impairment in randomized trials. ### The two components of workplace cost Anxiety's workplace cost is usually broken into two parts. Absenteeism is easier to see: missed shifts, sick days, and unplanned leave. Presenteeism is harder to see: hours at work with reduced concentration, slower task completion, more errors, and lower creative output. Analyses using instruments like the Work Productivity and Activity Impairment questionnaire consistently find that presenteeism accounts for more of the total workplace cost of anxiety and depression than absenteeism does. ### The Greenberg 1999 estimate and its updates Greenberg and colleagues' 1999 paper in the Journal of Clinical Psychiatry compiled the first comprehensive US estimate of the economic burden of anxiety disorders, at about $42 billion per year (1990 dollars), roughly a third of it from workplace productivity. The estimate has been updated for depression (Greenberg's 2015 and 2021 depression-specific updates), but anxiety-specific US updates are less common. International reviews, including Konnopka and Kรถnig's 2020 review in PharmacoEconomics, reach comparable conclusions about the size and distribution of the anxiety-disorder cost. ### Disability data Social Security Disability Insurance and Supplemental Security Income data show that mental health conditions have grown as a share of new disability awards over the past two decades, with mood and anxiety disorders together representing a substantial share. The disability figures are downstream of clinical severity and of workplace accommodations, and different countries handle disability differently, so international comparisons need care. ### Whether treatment reduces work impairment Randomized trials of anxiety treatment consistently find improvements in work-related functioning alongside symptom improvement, with effect sizes tracking those of symptom reduction. The CALM trial (Roy-Byrne et al., 2010), which tested collaborative care for anxiety in primary care, reported improvements in work functioning as well as symptom scores. Whether treatment scales up to reduce population-level disability claims is harder to establish and depends on access as much as on effectiveness. ### What the data show - Anxiety disorders carry a measurable cost at work, and the cost shows up in two forms. The first is absenteeism, missed days of work. The second is presenteeism, reduced productivity while present, which is harder to see but, in several analyses, it's the larger share of the total cost. Kessler and colleagues' work with the National Comorbidity Survey Replication data quantified anxiety-related lost work days and days of impaired function. - Burden-of-illness research has placed the economic cost of anxiety disorders in the United States in the tens of billions of dollars a year, with lost workplace productivity a major component. The foundational US burden estimate, by Greenberg and colleagues in the Journal of Clinical Psychiatry (1999), placed the annual cost at about $42 billion in 1990 dollars, and subsequent updates have not reversed the direction. Konnopka and colleagues' international review in 2020 reached comparable conclusions on the workplace-productivity share. - Anxiety also intersects with disability. Anxiety disorders are among the conditions cited in disability claims, often alongside depression, and the share of Social Security Disability Insurance awards involving a mental health condition has grown over time. The functional reach of anxiety doesn't stop at the clinic; it extends into earnings, employment stability, and the broader economy. ### How to read this - Burden-of-illness estimates depend heavily on the costing method and the year. Treat the 'tens of billions' framing as an order-of-magnitude statement, not a precise figure. - Presenteeism is estimated indirectly, usually through self-report scales like the Work Productivity and Activity Impairment questionnaire, and so carries measurement uncertainty. ### Limits - Presenteeism is difficult to measure. - Cost estimates vary widely by method and by what is counted (direct medical costs, indirect productivity, mortality-related costs). - Disability data reflects claims filed and awarded, not the total functional impact of anxiety. - The Greenberg 1999 estimate is dated; updated US-specific analyses are less numerous than for depression. ### Sources - Greenberg et al., economic burden of anxiety disorders (https://pubmed.ncbi.nlm.nih.gov/?term=greenberg+economic+burden+anxiety) - U.S. Bureau of Labor Statistics (https://www.bls.gov/) - Social Security Administration disability data (https://www.ssa.gov/disability/) - Greenberg PE, et al. The economic burden of anxiety disorders in the 1990s. Journal of Clinical Psychiatry, 1999. PMID: 10453795 (https://pubmed.ncbi.nlm.nih.gov/10453795/) - Konnopka A and Kรถnig H. Economic burden of anxiety disorders: a systematic review and meta-analysis. PharmacoEconomics, 2020. PMID: 31646432 (https://pubmed.ncbi.nlm.nih.gov/31646432/) - Kessler RC, et al. The prevalence and effects of adult attention deficit/hyperactivity disorder on work performance in a nationally representative sample of workers. Journal of Occupational and Environmental Medicine, 2005. PMID: 15951814 (https://pubmed.ncbi.nlm.nih.gov/15951814/) - WHO Global Burden of Disease Study, mental disorder burden estimates (https://www.who.int/data/gho/data/themes/global-burden-of-disease) ## Anxiety and school absenteeism URL: https://anxietyresearch.org/reports/anxiety-and-school-absenteeism Topic cluster: Functional impact Last updated: 2026-05-18 Chronic absenteeism, defined as missing 10 percent or more of school days, roughly doubled after the pandemic and remains elevated. Anxiety is one meaningful strand of the picture. School refusal, in which a young person cannot attend school because of emotional distress, is most often driven by anxiety, and early intervention improves both attendance and mental health outcomes. ### Key takeaways - Chronic absenteeism reached about 28 percent of US students in the 2021-22 school year, up from 15 percent pre-pandemic, and remains elevated (US Department of Education). - School refusal, a specific pattern of school avoidance driven by emotional distress, affects roughly 1-5 percent of school-age children. - Anxiety disorders (separation anxiety, social anxiety, generalized anxiety) account for the largest share of school-refusal cases. - CBT with a graded return-to-school component has moderate evidence for school refusal (Heyne et al.; Kearney). - School-based mental health programs improve attendance in evaluation studies (Sanchez et al., 2018). ### What school refusal is (and isn't) School refusal is a specific pattern in which a young person cannot attend school because of emotional distress, most often anxiety-driven. It is not truancy; truancy involves not wanting to attend, often with the parent unaware. School refusal involves the parent knowing the child is home, the child often distressed about not being able to attend, and physical symptoms (stomachaches, headaches, nausea) frequently front-and-center. Kearney's work has clarified the functional analysis of school-refusal behavior, distinguishing subtypes based on what the child is avoiding or seeking. ### How anxiety drives school avoidance Several anxiety subtypes contribute. Separation anxiety is most common in younger children. Social anxiety often drives absence in adolescents, particularly around class presentations, cafeteria social situations, and physical education. Generalized anxiety can drive avoidance through fear of failure or perfectionism. In each case, the short-term relief of staying home reinforces avoidance, which allows the anxiety to grow. Missed school also raises academic stress, adding a new source of anxiety. ### The chronic-absenteeism picture and the post-pandemic rise Chronic absenteeism, defined by most states as missing 10 percent or more of school days, roughly doubled during the pandemic and remains elevated. The US Department of Education's [chronic absenteeism data story](https://www2.ed.gov/datastory/chronicabsenteeism.html) documents the pattern. Attribution is complex: the shift to remote learning changed norms about attendance, mental-health symptoms rose (see [college-student anxiety data](/reports/anxiety-and-college-students) for the older-age analog), and family and economic disruption affected attendance directly. Anxiety is one strand among several. ### What works For clinical school refusal, CBT with a graded return-to-school protocol has the strongest evidence, sometimes combined with SSRI treatment for more severe presentations. School-based mental health programs, in which counselors or clinicians are embedded in schools, have small to moderate benefits on attendance and mental health outcomes in Sanchez and colleagues' 2018 systematic review. Early intervention is important: the longer the pattern persists, the harder it is to reverse. ### What the data show - Anxiety is a recognized contributor to school absence. The clearest connection is school refusal, sometimes called school avoidance, a pattern in which a child or adolescent has serious difficulty attending school because of emotional distress. Anxiety, including separation anxiety, social anxiety, and generalized anxiety, is one of the common drivers of school refusal. Estimates of school-refusal prevalence in US and international samples range from about 1 to 5 percent of school-age children, with anxiety disorders identified as the primary contributor in the majority of clinical samples. - The link also appears at the broader level of chronic absenteeism, defined in education data as missing 10 percent or more of school days. The US Department of Education reports that chronic absenteeism reached about 28 percent of students in the 2021-22 school year, up from about 15 percent before the pandemic, and remains elevated. Chronic absenteeism has many causes, and anxiety is one strand among them, but it's a strand that's often missed, because an anxious child may present with physical complaints, such as stomachaches or headaches, rather than naming fear. - The relationship matters because school absence and anxiety can reinforce each other. Missing school reduces a young person's exposure to the situations they fear, which brings short-term relief but allows the anxiety to grow, and falling behind academically adds a new source of stress. School-based mental health programs and early intervention are the responses with the most support; a 2018 systematic review by Sanchez and colleagues found small to moderate positive effects on attendance and mental health outcomes. ### How to read this - Anxiety is one cause of school absenteeism among several. Absenteeism data records that a student was absent, not why, so the anxiety contribution is inferred rather than directly counted. - School refusal is a specific clinical pattern, distinct from truancy; the two often need different responses. ### Limits - Education datasets record attendance, not the reason for an absence. - The anxiety contribution is estimated from clinical research, not from attendance records. - The clinical distinction between school refusal and truancy is not always clean in real cases. - School-based-mental-health-program studies vary widely in intervention type and quality. ### Sources - CDC Youth Risk Behavior Surveillance System (https://www.cdc.gov/healthyyouth/data/yrbs/index.htm) - American Academy of Pediatrics (https://www.aap.org/) - U.S. Department of Education, data on chronic absenteeism (https://www2.ed.gov/datastory/chronicabsenteeism.html) - U.S. Surgeon General, Advisory on Protecting Youth Mental Health, 2021 (https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/index.html) - Heyne D, et al. Assessment and treatment of school refusal in children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 2001 and subsequent reviews. PMID: 11455109 (https://pubmed.ncbi.nlm.nih.gov/11455109/) - Kearney CA. School absenteeism and school refusal behavior in youth: a contemporary review. Clinical Psychology Review, 2008. PMID: 17573151 (https://pubmed.ncbi.nlm.nih.gov/17573151/) - Sanchez AL, et al. The effectiveness of school-based mental health services for elementary-aged children: a meta-analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 2018. PMID: 29525359 (https://pubmed.ncbi.nlm.nih.gov/29525359/) ## Anxiety and sleep disruption URL: https://anxietyresearch.org/reports/anxiety-and-sleep Topic cluster: Functional impact Last updated: 2026-05-18 Insomnia and anxiety travel together. More than half of adults with an anxiety disorder report chronic sleep problems, and longitudinal studies show the relationship runs both directions: sleep loss raises anxiety, and anxiety corrodes sleep. Cognitive behavioral therapy for insomnia is the first-line treatment for chronic insomnia and consistently improves anxiety alongside sleep. ### Key takeaways - Roughly 40 to 60 percent of adults with anxiety disorders report clinically significant insomnia in cross-sectional studies. - Longitudinal cohort studies show sleep problems and anxiety predict each other prospectively (Cox and Olatunji, 2016; Neckelmann et al., 2007). - Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for chronic insomnia per the American Academy of Sleep Medicine guideline. - CBT-I improves comorbid anxiety symptoms as well as sleep in meta-analyses. - Sleep loss reliably increases next-day emotional reactivity in experimental studies. ### How common is insomnia in anxiety Estimates depend on how insomnia is defined and which anxiety disorder is being studied, but the pattern is consistent: adults with anxiety disorders report clinically significant insomnia at roughly 40 to 60 percent, several times the general-population rate. Generalized anxiety disorder and panic disorder show some of the highest rates, since worry and physiological arousal both interfere with sleep onset. See the review by [Cox and Olatunji (Journal of Anxiety Disorders, 2016)](https://pubmed.ncbi.nlm.nih.gov/26630013/) for a synthesis. ### The direction-of-cause question Cross-sectional studies can't tell you which came first. Longitudinal cohort studies can. Neckelmann and colleagues followed more than 25,000 adults in Norway for 11 years and found that persistent insomnia at baseline predicted the onset of anxiety disorders during follow-up, independent of baseline anxiety symptoms. Experimental sleep-restriction studies in healthy volunteers show that even a few nights of shortened sleep raise self-reported anxiety, negative affect, and threat sensitivity. The picture that emerges is bidirectional: each condition raises risk for the other, and once both are present they reinforce each other. ### What CBT-I is, and what the evidence says it does CBT-I is a structured, time-limited therapy of usually six to eight sessions. It combines sleep-restriction therapy (limiting time in bed to the amount of sleep actually obtained, then gradually extending), stimulus-control instructions (using the bed only for sleep), cognitive restructuring around sleep beliefs, and relaxation. Meta-analyses find it produces moderate to large improvements in sleep and, when anxiety is present, moderate improvements in anxiety symptoms alongside. See Belleville and colleagues' 2011 meta-analysis for the anxiety-outcome estimate. General sleep-hygiene advice, such as "avoid screens before bed" or "keep a consistent schedule," is not the same as CBT-I. The AASM guideline explicitly notes that sleep hygiene alone is not first-line for chronic insomnia. ### Medications for anxiety-related insomnia Sedative-hypnotics such as zolpidem and eszopiclone are FDA-approved for insomnia but carry dependence, next-day sedation, and safety concerns in older adults per the Beers Criteria. Benzodiazepines are effective for short-term use but not recommended for chronic insomnia in current guidelines. Sedating antidepressants such as trazodone and mirtazapine are used off-label for insomnia. Melatonin has small effects on sleep-onset latency but is not a first-line treatment. For anxiety with insomnia, structured CBT-I combined with an anxiety-focused psychotherapy or medication is the approach current guidelines support. Medication decisions belong with a clinician. ### What this doesn't prove The evidence shows that treating sleep helps anxiety, on average. It doesn't establish which patients benefit most, or that CBT-I is sufficient by itself for a diagnosed anxiety disorder. The average patient in a CBT-I trial does not have severe co-occurring conditions, so effect sizes in real-world populations may be smaller. Access to CBT-I remains a limit; there are far fewer trained CBT-I therapists than the need would suggest. ### What the data show - Sleep problems and anxiety are tightly linked. Insomnia, meaning difficulty falling asleep, staying asleep, or both, is among the most commonly reported symptoms in anxiety disorders. Surveys and clinical studies consistently find that people with anxiety report poor sleep at much higher rates than the general population; estimates in the 40 to 60 percent range recur across studies of generalized anxiety disorder and panic disorder. - The relationship runs in both directions, which cohort studies that follow people over time have shown. Anxiety can drive sleep problems, since a worried, activated mind resists sleep. And poor sleep can drive anxiety, since sleep loss raises emotional reactivity and makes the next day's worries harder to manage. Over time the two can settle into a self-reinforcing cycle. Neckelmann and colleagues' 2007 Norwegian cohort study of more than 25,000 adults followed for 11 years found that insomnia predicted the onset of anxiety disorders independent of baseline symptoms. - This two-way link has a practical payoff. Cognitive behavioral therapy for insomnia, known as CBT-I, is the first-line treatment for chronic insomnia per the [American Academy of Sleep Medicine clinical practice guideline](https://aasm.org/clinical-resources/practice-standards/), and a growing body of evidence shows it improves anxiety symptoms as well as sleep when the two co-occur. Treating sleep directly, rather than waiting for it to resolve once anxiety is treated, is a reasonable and evidence-supported approach. See the [evidence summary on anxiety and sleep](/evidence-summaries/anxiety-and-sleep) for the deeper trial-by-trial breakdown. ### How to read this - The anxiety and sleep link is robust and bidirectional. A single study can't establish the direction of cause for any individual, since the two feed each other. - CBT-I is a specific structured therapy, not general sleep hygiene advice; the two are often confused. ### Limits - Most of the evidence relies on self-reported sleep. - Objective sleep measurement (polysomnography, actigraphy) is less common in the epidemiological literature. - Estimates vary by how insomnia is defined (short-term vs chronic; symptom-level vs disorder-level). - Access to trained CBT-I clinicians is limited; digital CBT-I programs are expanding but their long-term data are still accumulating. ### Sources - NIH NHLBI on sleep (https://www.nhlbi.nih.gov/health/sleep) - American Academy of Sleep Medicine (https://aasm.org/) - Cox RC and Olatunji BO. A systematic review of sleep disturbance in anxiety and related disorders. Journal of Anxiety Disorders, 2016. PMID: 26630013 (https://pubmed.ncbi.nlm.nih.gov/26630013/) - Neckelmann D, et al. Chronic insomnia as a risk factor for developing anxiety and depression. Sleep, 2007. PMID: 17682656 (https://pubmed.ncbi.nlm.nih.gov/17682656/) - Belleville G, et al. Meta-analytic review of the impact of cognitive-behavior therapy for insomnia on concomitant anxiety. Clinical Psychology Review, 2011. PMID: 21482338 (https://pubmed.ncbi.nlm.nih.gov/21482338/) - American Academy of Sleep Medicine, clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults (https://aasm.org/clinical-resources/practice-standards/) ## Anxiety and relationship impairment URL: https://anxietyresearch.org/reports/anxiety-and-relationships Topic cluster: Functional impact Last updated: 2026-05-18 Anxiety disorders correlate with lower relationship satisfaction in meta-analyses of thousands of couples. Whisman's work shows the effect is roughly the same size as depression's. The picture in parenting is similar: anxious parents shape children's environments in ways that raise child-anxiety risk, though the pathway is not deterministic. Couple- and family-based interventions have modest but real evidence as complements to individual treatment. ### Key takeaways - Whisman's meta-analyses put the correlation between generalized anxiety disorder and marital dissatisfaction in the small-to-moderate range, comparable to major depression's. - Partner accommodation of anxiety symptoms (reassurance-giving, avoidance-supporting) is common and can maintain the disorder. - Parental anxiety raises risk of child anxiety through both heritable and environmental pathways (Micco et al., 2009). - Couple-based CBT and family-based interventions have small-to-moderate additional benefit over individual therapy in some anxiety presentations. - Caregiver burden in family members of adults with severe anxiety is a documented but under-studied area. ### Partner relationships Whisman's meta-analytic work, drawing on decades of couples research, established that anxiety disorders correlate with lower marital satisfaction at effect sizes comparable to depression's. The effect is present for generalized anxiety disorder, panic disorder, and social anxiety, with somewhat different mechanisms in each: excessive worry and reassurance-seeking in GAD, avoidance of shared activities in agoraphobia, and social withdrawal in social anxiety. Partner accommodation of anxiety symptoms is common and, while well-intentioned, can maintain the disorder by allowing avoidance to persist. ### Parenting and child anxiety Children of anxious parents have roughly two to three times the odds of developing an anxiety disorder themselves compared with children of non-anxious parents, based on Micco and colleagues' 2009 meta-analysis. The pathway is mixed. Shared genetics account for a share. Environmental transmission adds to it: parental modeling of anxious responses, family accommodation of child avoidance (letting a child skip feared situations, which reduces short-term distress but reinforces avoidance), and family communication patterns that emphasize threat all contribute. This is a probabilistic pattern, not a deterministic one. ### Couple- and family-based interventions Couple-based CBT for anxiety involves the partner as a co-therapist in exposure planning and skills practice, and can address partner-accommodation patterns directly. Small-to-moderate additional benefits over individual CBT have been reported in specific presentations, particularly panic disorder with agoraphobia. Family-based CBT is the standard delivery format for child anxiety disorders, with strong evidence for effectiveness. The evidence for adding couple or family involvement to individual treatment in adults is smaller and more variable but generally favorable. ### Caregiver burden Family members and partners of adults with severe or chronic anxiety disorders often carry a real practical burden: adjusting schedules, providing reassurance, managing avoidance, taking on tasks the person feels unable to do. This is documented in caregiver-burden literature but studied less in anxiety than in psychosis or dementia. Support for the caregivers themselves is a recognized part of good practice, though it is often unavailable in routine outpatient settings. ### What the data show - Anxiety disorders affect relationships, and the effect isn't limited to one kind of relationship. Research has documented it in partner, family, and parenting relationships, where anxiety disorders are associated with lower relationship satisfaction and with strain on partners and family members, who often take on extra practical and emotional load. Whisman's meta-analyses have quantified the marital-dissatisfaction association for generalized anxiety disorder, panic disorder, and social anxiety, finding small-to-moderate correlations that are roughly comparable to the correlation between marital dissatisfaction and major depression. Some anxiety presentations directly shape relational life, for example when avoidance limits shared activities, when reassurance-seeking becomes a recurring pattern, or when partner accommodation (adjusting shared plans to avoid triggers) becomes the shape of daily life. - The effect runs both ways. Relationship conflict and stress are themselves risk factors for anxiety, and a supportive relationship can be a protective factor. So a strained relationship and an anxiety disorder can each worsen the other. In parenting, the picture is similar: parental anxiety raises child-anxiety risk through a mix of shared genetics and environmental effects (modeling of anxious responses, accommodation of child avoidance, family communication patterns). Micco and colleagues' 2009 meta-analysis in Depression and Anxiety documented the parent-to-child transmission at the population level. - The treatment relevance is twofold. Involving a partner or family member in treatment, where appropriate, can help, and couple-based CBT and family-based interventions have modest evidence as a complement to individual treatment for some presentations. The burden on the people around a person with anxiety is real too, which is why support for family members is a recognized, if often overlooked, part of the picture. ### How to read this - Relationship effects are documented but vary widely by person and presentation. This is population-level research, not a prediction for any particular couple or family. - Correlational effect sizes describe averages; a specific couple may lie well above or below. ### Limits - Relationship outcomes are hard to measure objectively. - Most of the evidence is observational. - Findings vary by anxiety subtype. - Couples and family therapy trials in anxiety are small and heterogeneous compared to individual-therapy trials. ### Sources - American Psychiatric Association (https://www.apa.org/) - NIMH, Any Anxiety Disorder (https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder) - Whisman MA. Marital distress and DSM-IV psychiatric disorders in a population-based national survey. Journal of Abnormal Psychology, 2007. PMID: 17696710 (https://pubmed.ncbi.nlm.nih.gov/17696710/) - Micco JA, et al. Anxiety and depressive disorders in offspring at high risk for anxiety: a meta-analysis. Journal of Anxiety Disorders, 2009. PMID: 19464830 (https://pubmed.ncbi.nlm.nih.gov/19464830/) - Zaider TI and Heimberg RG. Relationships between anxiety and depression: contemporary views. Depression and Anxiety, treatment chapters (https://onlinelibrary.wiley.com/journal/15206394) ## Nebraska's psychiatrist desert: 82 counties, no psychiatrist URL: https://anxietyresearch.org/reports/nebraska-mental-health-access Topic cluster: Access and systems Last updated: 2026-07-03 Nebraska has about 175 licensed psychiatrists for nearly two million people, and they cluster almost entirely in its cities. The state's own workforce center reports that only 11 of Nebraska's 93 counties have a practicing psychiatrist. More than a million Nebraskans, roughly half the state, live in a federally designated mental health shortage area. If you live outside Omaha or Lincoln, psychiatric care is mostly something that happens somewhere else. ### Key takeaways - Only 11 of Nebraska's 93 counties have a practicing psychiatrist, per the state's own workforce center. - Nebraska had 175 licensed psychiatrists in 2024, up 8 percent from 2020, but rural Nebraska has just under 4 psychiatrists per 100,000 residents against a national average around 10.5. - HRSA designated 91 mental health professional shortage areas in Nebraska covering 1,068,645 people, with 44.5 percent of need met. - No Nebraska-specific psychiatric wait-time study has been published; for many Nebraskans the real wait is the distance to the nearest psychiatrist. ### What this report covers Public data on mental health prevalence, psychiatrist supply, and geographic access in Nebraska, with every figure cited. Where sources disagree, we show the disagreement. For how Nebraska compares with the rest of the country, see the [Nebraska state statistics page](/statistics/states/nebraska) and the national report on [rural and urban access](/reports/rural-and-urban-access). ### How common are mental health conditions in Nebraska About 24.4 percent of Nebraska adults, roughly 359,000 people, experienced any mental illness in the most recent federal survey period; 10.1 percent, about 149,000 adults, had a major depressive episode; and 5.7 percent reported serious thoughts of suicide (SAMHSA NSDUH, 2022-2023 state estimates). Mental Health America's 2025 rankings place Nebraska 30th overall and 35th for access to care. For what [prevalence](/glossary/prevalence) does and doesn't measure, see the glossary. ### The supply picture Nebraska had 175 licensed psychiatrists in 2024, up 8 percent from 162 in 2020, according to the [Behavioral Health Education Center of Nebraska (BHECN)](https://www.unmc.edu/bhecn/) at the University of Nebraska Medical Center. Growth is real but thin: rural Nebraska has just under 4 psychiatrists per 100,000 residents against a national average around 10.5, and researchers estimate roughly 26 per 100,000 are needed to meet demand. The state trains psychiatrists in only two residency programs, at Creighton and UNMC, and retains about 54 percent of those it trains. ### The geography problem This is Nebraska's defining access issue, and the numbers are stark however you cut them: - Only 11 of 93 counties have a practicing psychiatrist (BHECN, 2020-2024 workforce snapshot). - A peer-reviewed rural health study counted 78 counties with no practicing psychiatrist. - 88 of 93 counties are considered behavioral health shortage areas, and about 29 counties have no behavioral health professional of any kind (reporting compiled from BHECN and Rural and Remote Health; counts vary by year and definition, which is why we show them separately). As of December 31, 2025, HRSA designated 91 mental health professional shortage areas in Nebraska covering 1,068,645 people, with 44.5 percent of need met ([KFF analysis of HRSA data](https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-areas-hpsas/)). Nebraska's need-met figure beats the national average of 27.3 percent, but the population living inside a shortage area is roughly four times [Maine's](/reports/maine-mental-health-access). About 27 percent of Nebraskans live in rural areas (2020 Census), and the state's uninsured rate is 7.1 percent (2024 ACS). Roughly 23,000 Nebraskans contacted the 988 Suicide and Crisis Lifeline in 2024, rising past 35,000 contacts in fiscal year 2025 (Nebraska DHHS, Nebraska Public Media). ### Wait times No Nebraska-specific psychiatric wait-time study has been published. The 2025 AMN Healthcare/Merritt Hawkins survey found new-patient physician appointment waits averaging 31 days in large metros across surveyed specialties (psychiatry was not among them). In practice, Nebraskans outside the Omaha and Lincoln metros often face the longer wait embedded in distance: the nearest psychiatrist may be several counties away. For the national picture, see the report on [psychiatric wait times in the United States](/reports/psychiatric-wait-times). ### What this doesn't prove County counts vary by source year and by definition (licensed vs practicing, psychiatrist vs any behavioral health professional). Statewide per-capita figures blend urban abundance with rural absence. No public dataset measures how long Nebraskans actually wait for a first psychiatric appointment, and telehealth utilization for mental health is not reported at the state level. ### If you need care now In crisis, call or text 988. Nebraska's rural reality makes [telehealth](/reports/telepsychiatry-and-anxiety-care) one of the few distance-independent options; federally qualified health centers and the state's network of behavioral health regions also provide lower-cost paths to care. This report is educational and is not medical advice. ### Sources - BHECN/UNMC, The Psychiatry Workforce in Nebraska: 2020 to 2024, and 2026 Behavioral Health Workforce Capacity Report (https://www.unmc.edu/bhecn/) - SAMHSA, 2022-2023 NSDUH State Estimates, Nebraska tables (https://www.samhsa.gov/data/) - KFF State Health Facts, Mental Health Care HPSAs, data as of Dec 31, 2025 (https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-areas-hpsas/) - Mental Health America, The State of Mental Health in America 2025 (https://mhanational.org/issues/state-mental-health-america) - Rural and Remote Health, article 3392, Nebraska psychiatrist distribution (https://www.rrh.org.au/journal/article/3392) - America's Health Rankings, Mental Health Providers and Uninsured, Nebraska (https://www.americashealthrankings.org/) - U.S. Census Bureau, 2020 Census and QuickFacts Nebraska (https://www.census.gov/quickfacts/NE) - Nebraska Public Media / Nebraska DHHS, 988 contact volumes, 2024-2025 (https://dhhs.ne.gov) - AMN Healthcare / Merritt Hawkins, 2025 Survey of Physician Appointment Wait Times (https://www.amnhealthcare.com/) ## Mental health care access in Maine: what the data shows in 2026 URL: https://anxietyresearch.org/reports/maine-mental-health-access Topic cluster: Access and systems Last updated: 2026-07-03 Maine is a paradox. It ranks second in the nation for access to mental health care, mostly because so many residents are insured. At the same time, federal shortage data shows Maine meets only 14.4 percent of its need for mental health practitioners, one of the lowest rates in the country, and a 2024 point-in-time study found roughly 13,000 Mainers waiting an average of 8 months for care. Coverage is not the problem. Clinicians are. ### Key takeaways - Maine ranks 2nd of 51 for access to mental health care in Mental Health America's 2025 rankings, behind only Vermont, yet 49th of 51 for prevalence of mental illness. - HRSA-designated shortage areas in Maine meet only 14.41 percent of estimated mental health need, well below the national average of 27.29 percent. - A fall 2024 point-in-time study found about 13,000 Mainers waiting an average of 8 months for a provider, counseling, or prescribing services. - Psychiatry was the only mental health profession in Maine that shrank between 2019 and 2024, while the state's total mental health workforce grew about 60 percent. ### What this report covers This report summarizes the most recent public data on mental health prevalence, provider supply, and waiting times in Maine. Every figure is cited to its source. Where sources conflict, we say so. For how Maine compares with the rest of the country, see the [Maine state statistics page](/statistics/states/maine) and the national report on [rural and urban access](/reports/rural-and-urban-access). ### How common are mental health conditions in Maine About 25.8 percent of Maine adults, roughly 290,000 people, experienced any mental illness in the most recent federal survey period, and 9.4 percent, about 106,000 adults, had a major depressive episode (SAMHSA NSDUH, 2022-2023 state estimates). Mental Health America's 2025 rankings place Maine 49th of 51 for prevalence, meaning Maine has among the highest rates of mental illness in the country. For what [prevalence](/glossary/prevalence) does and doesn't measure, see the glossary. ### The access paradox The same MHA rankings place Maine 2nd of 51 for access to care, behind only Vermont. Only about 13 percent of Maine adults with a mental illness reported not receiving needed care, an improvement widely attributed to the state's 2019 Medicaid expansion. Maine's uninsured rate is 5.5 percent (2024 ACS), better than the national 8.2 percent. So by insurance measures, Maine looks strong. The supply-side data tells a different story. ### The clinician bottleneck As of December 31, 2025, HRSA designated 66 mental health professional shortage areas in Maine, covering 269,696 residents. Within those areas, only 14.41 percent of estimated need is met, well below the national average of 27.29 percent ([KFF analysis of HRSA data](https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-areas-hpsas/)). Counting psychiatrists in Maine is surprisingly hard, and honest reporting requires two numbers. The Maine Board of Licensure in Medicine counted 343 active psychiatrist licenses in January 2025, but the board does not track whether those license-holders actually practice in Maine. Federal labor statistics, which exclude the self-employed, counted just 60 practicing psychiatrists in 2022, down from 110 in 2019 (both figures via [The Maine Monitor](https://www.themainemonitor.org/mental-health-providers/)). The true number of practicing psychiatrists sits somewhere between, and the trend line is the alarming part: psychiatry was the only mental health profession in Maine that shrank between 2019 and 2024, while the state's total mental health workforce grew about 60 percent, driven mostly by social workers. ### How long people wait A fall 2024 point-in-time study led by the Maine chapter of the National Association of Social Workers found about 13,000 patients seeking a provider, counseling, or prescribing services, waiting an average of 8 months. More than 8,000 had waited between 6 and 24 months for mental health counseling specifically (The Maine Monitor, February 2025). For context, the 2025 AMN Healthcare/Merritt Hawkins survey put the average wait for a new-patient physician appointment across specialties at 31 days in large metro areas; psychiatry was not among the surveyed specialties, but Maine's measured waits run many multiples of that benchmark. For the national picture, see the report on [psychiatric wait times in the United States](/reports/psychiatric-wait-times). ### The rural dimension Maine is the second most rural state in the nation, with about 61 percent of residents living in rural areas (2020 Census). Recruiting clinicians to rural counties, and increasingly to Portland because of housing costs, is cited by workforce researchers as the central barrier. More than a third of rural Maine physicians are over 60. For how the same geography problem plays out in the plains, see the companion report on [Nebraska's psychiatrist desert](/reports/nebraska-mental-health-access). ### What this doesn't prove These datasets measure different things in different years, and none of them captures every clinician type. The 8-month wait figure comes from a point-in-time count of people actively seeking care, not a random sample. Telepsychiatry-specific utilization data for Maine is not publicly available. None of this data tells an individual person how long they personally will wait. ### If you need care now If you are in crisis, call or text 988. For non-crisis care, options that shorten the wait include [telehealth](/reports/telepsychiatry-and-anxiety-care) practices licensed in Maine, federally qualified health centers with sliding-scale fees, and Maine's [211 directory](https://211maine.org). This report is educational and is not medical advice. ### Sources - SAMHSA, 2022-2023 NSDUH State Estimates, Maine tables (https://www.samhsa.gov/data/) - Mental Health America, The State of Mental Health in America 2025, state rankings (https://mhanational.org/issues/state-mental-health-america) - KFF State Health Facts, Mental Health Care HPSAs, data as of Dec 31, 2025 (https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-areas-hpsas/) - The Maine Monitor, Maine's mental health workforce grew 60%. Psychiatry shrank. Feb 16, 2025 (https://www.themainemonitor.org/mental-health-providers/) - America's Health Rankings (UnitedHealth Foundation), Mental Health Providers and Uninsured measures, Maine, 2024-2025 (https://www.americashealthrankings.org/) - AMN Healthcare / Merritt Hawkins, 2025 Survey of Physician Appointment Wait Times (https://www.amnhealthcare.com/) - U.S. Census Bureau, 2020 Census urban-rural classification (https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural.html) - CNN Health, The best and worst states for mental health care, Oct 1, 2025 (https://www.cnn.com/health) # Evidence summaries ## CBT for anxiety disorders URL: https://anxietyresearch.org/evidence-summaries/cbt-for-anxiety Cognitive behavioral therapy shows moderate to large effects across generalized anxiety, social anxiety, panic, and specific phobias in meta-analyses. ### Research question How well does cognitive behavioral therapy work for adults with anxiety disorders, and how does it compare with other treatments? ### What the evidence suggests CBT is the most thoroughly studied psychological treatment for anxiety, and the evidence base is large. A 2018 meta-analysis by Carpenter and colleagues pooled 41 randomized placebo-controlled trials and found that CBT produced moderate effect sizes for anxiety disorders compared with placebo conditions. The effects were strongest for generalized anxiety disorder and obsessive-compulsive disorder and somewhat smaller for panic disorder and social anxiety disorder. Earlier meta-analyses by Hofmann and Smits reached compatible conclusions across a broader sweep of trials. The size of the effect depends heavily on the comparison group. Against a waitlist, where people simply wait for treatment, CBT shows large effects. Against an active comparison, such as another credible therapy or a pill placebo, the effect is smaller but still present. This pattern is normal, and it's one reason a single trial's headline number can mislead. The comparison sets the meaning of the result. For panic disorder, social anxiety disorder, and specific phobias, the exposure component of CBT carries much of the benefit. Trials that include structured, adequately dosed exposure tend to show larger effects than those relying mainly on cognitive techniques alone. Follow-up data suggest gains from CBT hold up for many patients after treatment ends, though relapse isn't rare and some patients use booster sessions. ### What the evidence does not prove The evidence doesn't show that CBT works equally well for every person or every anxiety disorder. It doesn't establish that any single CBT protocol is clearly superior to the others. It doesn't prove that benefits last indefinitely without further contact, since long-term follow-up beyond one to two years is limited. Many trials exclude people with significant comorbid conditions, so results may not transfer fully to the many patients who carry more than one diagnosis. ### Practical interpretation Clinical guidelines from the American Psychiatric Association and NICE place CBT among first-line treatments for adult anxiety disorders. In practice, the choice between CBT, medication, and combined care depends on the specific disorder, symptom severity, patient preference, prior treatment response, and whether a trained CBT therapist is available. Access is a real limit, since CBT requires a trained provider and active participation, and trained therapists are in short supply in many areas. ### Sources - Carpenter et al., Depression and Anxiety, 2018 (https://pubmed.ncbi.nlm.nih.gov/29451967/) - Hofmann and Smits, 2008 meta-analysis (https://pubmed.ncbi.nlm.nih.gov/18363421/) - APA practice guidelines (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) - NICE GAD guideline (CG113) (https://www.nice.org.uk/guidance/cg113) ## SSRIs and SNRIs for anxiety URL: https://anxietyresearch.org/evidence-summaries/ssris-snris-for-anxiety SSRIs and SNRIs produce small to moderate effect sizes versus placebo across most adult anxiety disorders and are first-line pharmacotherapy in major guidelines. ### Research question How well do SSRIs and SNRIs work for adult anxiety disorders, and what are the main safety considerations? ### What the evidence suggests SSRIs and SNRIs are the most studied medications for anxiety, and meta-analyses consistently find they outperform placebo across the major anxiety disorders. A 2015 meta-analysis by Bandelow and colleagues, pooling treatment trials across the anxiety disorders, found small to moderate effect sizes for both classes against placebo. A 2019 network meta-analysis in the Lancet by Slee and colleagues compared drug treatments for generalized anxiety disorder and found that several SSRIs and SNRIs outperformed placebo, with the differences between agents within the class being small. Response builds over weeks, not days. A meaningful effect typically appears over four to six weeks, and a full response can take eight to twelve weeks. This slow onset matters clinically, because a patient who expects a fast result may stop early and conclude the medication failed before it had a fair trial. There's no strong evidence that any one SSRI is meaningfully better than another for anxiety overall. Because efficacy is broadly similar within the class, the choice between agents is usually driven by the side-effect profile, possible drug interactions, prior response, and cost. ### What the evidence does not prove The evidence doesn't predict which individual patient will respond to which medication. It doesn't show that benefits persist after the medication stops; relapse after discontinuation is common, which is why guidelines often suggest continuing treatment for a period after improvement. Average effect sizes also conceal wide variation, so a group result doesn't forecast any one person's experience. ### Clinical considerations SSRIs and SNRIs carry an FDA boxed warning about an increased risk of suicidal thoughts in children, adolescents, and young adults, and close monitoring is standard in the early weeks of treatment. Sexual side effects are common and are a frequent reason patients stop. Stopping abruptly, especially with shorter-acting agents such as paroxetine and venlafaxine, can cause discontinuation syndrome, so these medications are lowered gradually with a planned taper. SNRIs can raise blood pressure at higher doses. ### Practical interpretation SSRIs and SNRIs are first-line medication options for generalized anxiety disorder, social anxiety disorder, and panic disorder in US and international guidelines. Medication and CBT are both reasonable first choices, and the decision usually rests on patient preference, access, severity, and prior response. None of this is individualized advice. Medication decisions belong with a prescribing clinician. ### Sources - Bandelow et al., International Clinical Psychopharmacology, 2015. PMID: 25932596 (https://pubmed.ncbi.nlm.nih.gov/25932596/) - Slee et al., Lancet, 2019. PMID: 30712879 (https://pubmed.ncbi.nlm.nih.gov/30712879/) - US Food and Drug Administration, revisions to product labeling for antidepressants (suicidality warning) (https://www.fda.gov/media/74149/download) - APA practice guidelines (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) ## Exposure therapy URL: https://anxietyresearch.org/evidence-summaries/exposure-therapy Exposure-based therapy produces large effect sizes for specific phobias and panic disorder. Inhibitory learning frameworks now guide protocol design. ### Research question How effective is exposure-based therapy for anxiety disorders, panic, and specific phobias? ### What the evidence suggests Exposure therapy has one of the strongest evidence bases in mental health treatment. For specific phobias, controlled trials and meta-analyses report large effects, and in vivo exposure, where a person faces the feared object or situation directly, consistently outperforms less direct approaches. For panic disorder, exposure to feared bodily sensations, called interoceptive exposure, is a core and well-supported component of treatment. For social anxiety disorder, exposure to feared social situations is central to effective CBT. The way researchers understand how exposure works has shifted. The older model held that exposure worked through habituation, meaning anxiety simply faded with repeated contact. The current dominant framework is the inhibitory learning model, developed by Craske and colleagues. It holds that exposure works by building new learning that competes with the original fear, rather than erasing it. The shift has practical consequences for how exposure is designed, including a focus on violation of expectancy, variability across contexts, and removal of within-session safety behaviors. Single-session and intensive formats have evidence for specific phobias. Virtual reality exposure is an active research area with promising but more limited evidence than in vivo exposure. ### What the evidence does not prove The evidence doesn't show that exposure is equally tolerable for every patient. It's demanding by design, and dropout can be higher than for less intensive treatments. It doesn't establish that virtual reality exposure matches in vivo exposure across all conditions. It also doesn't prove that exposure works well without therapist guidance for most patients. ### Clinical considerations Pre-session use of benzodiazepines or alcohol may interfere with the new learning that exposure depends on, and is generally avoided during active exposure work. Exposure is most effective when it's structured, planned, and adequately dosed rather than brief or avoidant. ### Practical interpretation Exposure-based therapy is a first-line approach for specific phobias and a core component of effective CBT for panic disorder and social anxiety disorder. It's best delivered by a trained therapist, especially at the start. The discomfort it involves is part of how it works, which makes a clear rationale and a good therapeutic relationship important. ### Sources - Craske et al., Behaviour Research and Therapy, 2014. PMID: 24864005 (https://pubmed.ncbi.nlm.nih.gov/24864005/) - Wolitzky-Taylor et al., Clinical Psychology Review, 2008. PMID: 18410984 (https://pubmed.ncbi.nlm.nih.gov/18410984/) - APA practice guidelines (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) ## Panic disorder treatment URL: https://anxietyresearch.org/evidence-summaries/panic-disorder-treatment CBT with interoceptive exposure and SSRIs both have strong individual evidence. Combination treatment may offer short-term advantages over monotherapy. ### Research question What treatments have the strongest evidence for panic disorder, with or without agoraphobia? ### What the evidence suggests Two treatments have the strongest individual evidence for panic disorder: cognitive behavioral therapy that includes interoceptive exposure, and SSRIs. Meta-analyses find both produce meaningful reductions in panic frequency and severity compared with control conditions. CBT for panic specifically targets the panic cycle, in which a person misreads normal or anxious body sensations as dangerous, which raises fear, which intensifies the sensations. Combined treatment, meaning CBT plus medication, has been studied directly. A meta-analysis by Mitte and other reviews suggest that combining the two can offer an advantage over either alone in the short term, particularly for more severe presentations, though the long-term difference between combined treatment and CBT alone is smaller. Some evidence suggests CBT may protect against relapse better than medication alone once treatment ends, possibly because the skills remain available after the course is over. Benzodiazepines reduce acute panic symptoms quickly and are sometimes used short-term, but they're limited for long-term management by tolerance and dependence, and guidelines don't favor them as a standalone long-term treatment. ### What the evidence does not prove The evidence doesn't identify which specific patient will respond best to therapy, medication, or the combination. It doesn't establish a single optimal sequence for combined treatment. Long-term follow-up beyond one to two years is limited. ### Clinical considerations New panic symptoms, especially a first episode, deserve medical evaluation rather than an assumption of anxiety. Cardiac conditions, thyroid problems, and the effects of caffeine, stimulants, and alcohol withdrawal can mimic panic. Ruling out medical causes is part of responsible assessment. ### Practical interpretation For panic disorder, CBT with interoceptive exposure and SSRIs are both reasonable first-line choices, and combined treatment is an option for more severe cases. The decision depends on severity, patient preference, access to a trained therapist, and prior response. Treatment decisions belong with a clinician. ### Sources - Mitte, Journal of Affective Disorders, 2005. PMID: 16005982 (https://pubmed.ncbi.nlm.nih.gov/16005982/) - Bandelow et al., International Clinical Psychopharmacology, 2015. PMID: 25932596 (https://pubmed.ncbi.nlm.nih.gov/25932596/) - APA practice guideline for the treatment of panic disorder (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) ## Generalized anxiety disorder URL: https://anxietyresearch.org/evidence-summaries/generalized-anxiety-disorder SSRIs, SNRIs, and CBT are all supported as first-line for GAD. Buspirone, pregabalin, hydroxyzine, and adjunctive quetiapine each have a defined evidence niche. ### Research question What does the evidence show about treating generalized anxiety disorder in adults? ### What the evidence suggests Generalized anxiety disorder, or GAD, has a solid evidence base for both psychological and medication treatment. CBT shows moderate to large effects for GAD in meta-analyses, and GAD is among the anxiety disorders where CBT performs most strongly. On the medication side, the 2019 Lancet network meta-analysis by Slee and colleagues compared drug treatments for GAD directly and found several agents more effective than placebo, including the SSRIs escitalopram and sertraline, the SNRIs duloxetine and venlafaxine, and pregabalin. Pregabalin has consistent evidence in GAD and is approved for the condition in Europe, though it isn't FDA-approved for GAD in the United States. Buspirone has older supporting evidence in GAD and is sometimes used, particularly when SSRIs aren't tolerated or as an add-on. Hydroxyzine, an antihistamine, has some evidence for short-term symptom relief. Quetiapine has trial evidence in GAD but is rarely a first-line choice because of its metabolic and other side-effect burden. A consistent finding across GAD studies is that relapse is common after treatment stops. Continued treatment, whether therapy skills or medication, is often needed to maintain gains. ### What the evidence does not prove The evidence doesn't predict the best treatment for an individual. It doesn't show that any one first-line option is clearly superior to the others. It doesn't establish how long treatment should continue, which remains an individualized clinical decision. ### Practical interpretation For GAD, CBT, SSRIs, and SNRIs are all first-line options in US and international guidelines. The choice depends on patient preference, severity, access, side-effect tolerance, and prior response. Because relapse is common, plans for how long to continue treatment are part of the conversation. This is general information, not individualized advice. ### Sources - Slee et al., Lancet, 2019. PMID: 30712879 (https://pubmed.ncbi.nlm.nih.gov/30712879/) - Carpenter et al., Depression and Anxiety, 2018. PMID: 29451967 (https://pubmed.ncbi.nlm.nih.gov/29451967/) - NICE GAD guideline (CG113) (https://www.nice.org.uk/guidance/cg113) ## Social anxiety disorder URL: https://anxietyresearch.org/evidence-summaries/social-anxiety-disorder CBT with exposure and SSRIs are first-line. Group CBT shows comparable effects to individual CBT in several trials. ### Research question What treatments have evidence for social anxiety disorder in adults? ### What the evidence suggests Social anxiety disorder responds to both psychological and medication treatment, and a large network meta-analysis has compared the options directly. A 2014 network meta-analysis in Lancet Psychiatry by Mayo-Wilson and colleagues found that individual CBT was the most effective intervention for social anxiety disorder, with effects that were larger and appeared more durable than medication. SSRIs and SNRIs were also more effective than placebo and remain a recommended option, particularly for patients who prefer medication or cannot access CBT. Within CBT, exposure to feared social situations is central. Treatment helps a person test their fearful predictions about social judgment and gather evidence that contradicts them. Group CBT has evidence as well, and several trials find group formats produce effects broadly comparable to individual CBT, which matters for access because group delivery can reach more people. ### What the evidence does not prove The evidence doesn't show that CBT works for every patient, and dropout occurs. It doesn't establish the best treatment for a specific individual. The finding that CBT may be more durable than medication comes from comparing groups; it doesn't predict any one person's long-term course. ### Practical interpretation For social anxiety disorder, CBT with exposure is a strong first-line choice, and SSRIs and SNRIs are also first-line. The network meta-analysis evidence pointing to CBT's durability is one consideration among several, alongside patient preference, severity, and access to a trained therapist. Treatment decisions belong with a clinician. ### Sources - Mayo-Wilson et al., Lancet Psychiatry, 2014. PMID: 26361000 (https://pubmed.ncbi.nlm.nih.gov/26361000/) - Bandelow et al., International Clinical Psychopharmacology, 2015. PMID: 25932596 (https://pubmed.ncbi.nlm.nih.gov/25932596/) - NICE social anxiety guideline (CG159) (https://www.nice.org.uk/guidance/cg159) ## Anxiety and sleep URL: https://anxietyresearch.org/evidence-summaries/anxiety-and-sleep Insomnia and anxiety frequently co-occur. CBT for insomnia improves both sleep and anxiety symptoms in many randomized trials. ### Research question What is the relationship between anxiety and sleep problems, and what treatments help when the two occur together? ### What the evidence suggests Anxiety and sleep problems are closely linked, and the relationship runs in both directions. Insomnia is one of the most common symptoms reported in anxiety disorders. At the same time, poor sleep predicts the later development and worsening of anxiety. This two-way pattern appears in cohort studies that follow people over time, where sleep problems at one point predict anxiety later, and anxiety predicts sleep problems later. The treatment with the strongest evidence for chronic insomnia is cognitive behavioral therapy for insomnia, known as CBT-I. Major sleep guidelines, including those from the American Academy of Sleep Medicine, recommend CBT-I as the first-line treatment for chronic insomnia. Trials show CBT-I improves sleep, and a growing body of evidence finds it also reduces anxiety symptoms when the two conditions occur together. Treating sleep directly, rather than assuming it will resolve once anxiety is treated, appears to help. Medication for sleep alone, such as sedative-hypnotics, can produce short-term sleep improvement but doesn't reliably address the underlying anxiety, and some sleep medications carry their own dependence and next-day-impairment concerns. ### What the evidence does not prove The evidence doesn't establish the direction of cause in any individual case, since anxiety and insomnia feed each other. It doesn't show that treating sleep alone resolves an anxiety disorder. Long-term data on combined treatment approaches remain limited. ### Practical interpretation When anxiety and insomnia occur together, the evidence supports treating sleep directly rather than waiting for it to resolve. CBT-I is the first-line treatment for chronic insomnia and may help anxiety symptoms as well. A clinician can help decide whether to address sleep, anxiety, or both at once. This is general information, not individualized advice. ### Sources - American Academy of Sleep Medicine, clinical practice guideline for the treatment of chronic insomnia disorder in adults (https://aasm.org/clinical-resources/practice-standards/) - Cox and Olatunji, Journal of Anxiety Disorders, 2016. PMID: 26630013 (https://pubmed.ncbi.nlm.nih.gov/26630013/) - National Heart, Lung, and Blood Institute, information on insomnia (https://www.nhlbi.nih.gov/health/insomnia) ## Exercise and anxiety symptoms URL: https://anxietyresearch.org/evidence-summaries/exercise-and-anxiety Moderate-intensity aerobic exercise produces small to moderate reductions in anxiety symptoms in randomized trials. ### Research question Does exercise reduce anxiety symptoms, and how should that evidence be understood? ### What the evidence suggests Exercise has a real but modest effect on anxiety symptoms. Randomized trials and meta-analyses find that regular aerobic exercise produces small to moderate reductions in anxiety symptoms compared with no-exercise control conditions. A 2017 meta-analysis by Stubbs and colleagues examined exercise for people with anxiety and stress-related disorders and found a beneficial effect on symptoms. Other reviews of exercise in both clinical and general populations report results in the same direction. The effect is most consistent for subclinical anxiety and for mild to moderate symptoms. Moderate-intensity aerobic activity is the most studied form. The exact dose that works best isn't settled, and trials vary widely in length, intensity, and how they measure outcomes, which produces real heterogeneity in the pooled results. ### What the evidence does not prove The evidence doesn't show that exercise replaces first-line treatments for moderate to severe anxiety disorders. It doesn't establish a single optimal type, intensity, or duration of exercise. Many exercise trials are short and have methodological limits, including the difficulty of blinding, since people know whether they're exercising. ### Practical interpretation Exercise is reasonably supported as one helpful component of managing anxiety, particularly for milder symptoms, and it carries broad general health benefits. The evidence doesn't position it as a stand-alone substitute for CBT or medication in a diagnosed anxiety disorder. How exercise fits into a treatment plan is best worked out with a clinician. ### Sources - Stubbs et al., Psychiatry Research, 2017. PMID: 28088704 (https://pubmed.ncbi.nlm.nih.gov/28088704/) - Aylett et al., BMC Health Services Research, 2018. PMID: 30509251 (https://pubmed.ncbi.nlm.nih.gov/30509251/) ## Mindfulness-based interventions URL: https://anxietyresearch.org/evidence-summaries/mindfulness-based-interventions MBSR and MBCT are structured eight-week clinician-led courses with the strongest evidence base. Commercial mindfulness apps are a different intervention with weaker, mixed evidence. ### Research question What does the evidence show about mindfulness-based interventions for anxiety? ### What the evidence suggests Mindfulness-based interventions, mainly mindfulness-based stress reduction and mindfulness-based cognitive therapy, have been studied for anxiety in a number of trials. Meta-analyses find these structured programs produce small to moderate reductions in anxiety symptoms. A widely cited 2014 review in JAMA Internal Medicine by Goyal and colleagues found that mindfulness meditation programs produced modest improvements in anxiety, and that the effects were roughly comparable to those of other active interventions rather than clearly superior to them. An important distinction in this evidence is between structured eight-week programs and informal or app-based mindfulness. The stronger evidence is for the structured eight-week formats delivered by a trained instructor. Evidence for app-based mindfulness is more mixed, and the results of those trials are less consistent. ### What the evidence does not prove The evidence doesn't show that mindfulness is superior to established first-line treatments such as CBT or medication for diagnosed anxiety disorders. It doesn't establish that brief or app-based mindfulness reproduces the effects of a structured eight-week course. Major clinical guidelines don't position mindfulness as a stand-alone treatment for moderate to severe anxiety disorders. ### Practical interpretation Mindfulness-based interventions, particularly the structured eight-week programs, are a reasonable option as part of a broader plan, especially for milder symptoms or alongside other treatment. They aren't a replacement for first-line care in a diagnosed anxiety disorder. A clinician can help weigh where mindfulness fits for a given person. ### Sources - Goyal et al., JAMA Internal Medicine, 2014. PMID: 24395196 (https://pubmed.ncbi.nlm.nih.gov/24395196/) - Hofmann et al., Journal of Consulting and Clinical Psychology, 2010. PMID: 20350028 (https://pubmed.ncbi.nlm.nih.gov/20350028/) ## Benzodiazepines, benefits and risks URL: https://anxietyresearch.org/evidence-summaries/benzodiazepines Benzodiazepines provide rapid acute relief. Long-term use carries tolerance, dependence, withdrawal, and combination-use risk that has shaped a tightening regulatory posture. ### Research question What does the evidence show about the benefits and risks of benzodiazepines for anxiety? ### What the evidence suggests Benzodiazepines reduce acute anxiety quickly. Unlike SSRIs and SNRIs, which take weeks, benzodiazepines act within minutes to an hour. For short-term relief of severe anxiety or for occasional use, that speed is a genuine clinical benefit, and trials confirm short-term symptom reduction. The concern is with longer-term use. With regular use the body develops tolerance, so the same dose produces less effect over time. Physical dependence can develop, and stopping after sustained use can cause a withdrawal syndrome. For this reason, clinical guidelines generally recommend benzodiazepines for short-term or situational use rather than as a long-term stand-alone treatment for an anxiety disorder. The safety profile has been the subject of formal regulatory action. In 2020 the FDA required a class-wide update to benzodiazepine labeling to strengthen warnings about abuse, misuse, dependence, withdrawal, and the risks of stopping. Benzodiazepines also carry an FDA boxed warning about the danger of combining them with opioids, a combination that can cause severe sedation and slowed breathing. ### What the evidence does not prove The evidence doesn't show that every patient who uses a benzodiazepine develops tolerance or dependence. It doesn't establish that benzodiazepines have no role; they remain useful for specific, time-limited situations. The point is that the risk profile rises substantially with long-term use. ### Clinical considerations Benzodiazepines are listed as potentially inappropriate medications for older adults in the American Geriatrics Society Beers Criteria, because of risks including falls, confusion, and cognitive effects. Stopping abruptly after regular use can be medically serious, including a risk of seizures, so any taper from sustained use should be planned and supervised by a clinician. ### Practical interpretation Benzodiazepines have a real but limited role: short-term or occasional use, often while a longer-term treatment such as an SSRI takes effect. They're generally not a first-line long-term treatment for anxiety disorders. Decisions about starting, continuing, or stopping a benzodiazepine belong with a prescribing clinician. ### Sources - US Food and Drug Administration, Drug Safety Communication, benzodiazepine class labeling update, 2020 (https://www.fda.gov/drugs/drug-safety-and-availability/fda-requiring-boxed-warning-updated-improve-safe-use-benzodiazepine-drug-class) - US Food and Drug Administration, Drug Safety and Availability index (boxed warning on concurrent use of benzodiazepines and opioids) (https://www.fda.gov/drugs/drug-safety-and-availability) - American Geriatrics Society, Beers Criteria for potentially inappropriate medication use in older adults (https://pubmed.ncbi.nlm.nih.gov/37139824/) ## Telepsychiatry for anxiety care URL: https://anxietyresearch.org/evidence-summaries/telepsychiatry-for-anxiety Outcomes of telepsychiatry for assessment, medication management, and several psychotherapies are comparable to in-person care. Regulatory rules on prescribing and state licensure shape what telepsychiatry can deliver. ### Research question How do the outcomes of telepsychiatry compare with in-person care for anxiety? ### What the evidence suggests Telepsychiatry, meaning psychiatric care delivered through secure video visits, has been studied for two decades, and the evidence base grew quickly after 2020. Systematic reviews and comparative studies report that telepsychiatry produces outcomes broadly comparable to in-person care for psychiatric assessment, medication management, and several psychotherapies, including CBT delivered by video. Patient satisfaction in these studies is generally high. The clearest practical benefit is access. Telepsychiatry extends care into rural areas and other places where the in-person specialty workforce is thin. Studies of telepsychiatry in shortage areas find it can connect patients to care that would otherwise involve long travel or long waits. ### What the evidence does not prove The evidence doesn't show that telepsychiatry is equivalent to in-person care for every situation. It has clear limits for acute risk assessment, for physical examination, and for emergency stabilization. Evidence for fully virtual delivery of exposure-based protocols is still developing. ### Important practical considerations Telepsychiatry operates within real regulatory limits. A clinician must be licensed in the state where the patient is physically located at the time of the visit. The prescribing of controlled substances by telemedicine is governed by the federal Ryan Haight Act and by DEA rules that have been subject to ongoing change. These rules shape what telepsychiatry can and cannot do. ### Practical interpretation For many people with anxiety, telepsychiatry is a reasonable and evidence-supported way to receive assessment, medication management, and therapy, and it can substantially improve access. It isn't suited to psychiatric emergencies. Whether telepsychiatry fits a given person's needs is a clinical decision. ### Sources - Hilty et al., Telemedicine and e-Health, 2013. PMID: 23697504 (https://pubmed.ncbi.nlm.nih.gov/23697504/) - Bashshur et al., Telemedicine and e-Health, 2016. PMID: 26624248 (https://pubmed.ncbi.nlm.nih.gov/26624248/) - US Drug Enforcement Administration, information on telemedicine and the Ryan Haight Act (https://www.deadiversion.usdoj.gov/) # Treatment research ## Buspirone evidence URL: https://anxietyresearch.org/treatment-research/buspirone-evidence Last updated: 2026-05-18 What buspirone is, what the controlled trials show for generalized anxiety disorder, and where it fits in current practice. ### Research question What does the evidence show about buspirone for anxiety disorders, and where does it fit in treatment? ### What the evidence suggests - Buspirone is an anti-anxiety medication that works differently from both antidepressants and benzodiazepines. It acts mainly on serotonin receptors and is used primarily for generalized anxiety disorder. - Buspirone has been studied for generalized anxiety disorder since the 1980s, and controlled trials support a modest benefit over placebo for GAD symptoms. Its evidence base is strongest in GAD. Evidence in panic disorder and social anxiety disorder is weak, and buspirone isn't generally used as a primary treatment for those conditions. - Buspirone has two features that set it apart from benzodiazepines. It doesn't cause physical dependence, and stopping it doesn't produce a withdrawal syndrome. It also doesn't cause the sedation that benzodiazepines do. The trade-off is speed. Buspirone takes about two to four weeks to reach a clear effect, so it doesn't provide the rapid relief a benzodiazepine does. - Because of these features, buspirone is often used in specific situations rather than as a universal first choice. It's considered when an SSRI or SNRI isn't tolerated, as an add-on when an antidepressant has produced only a partial response, or for a patient for whom the dependence risk of a benzodiazepine is a particular concern. Clinical guidelines for GAD include buspirone as a recognized option. ### Important limitations The buspirone trial literature is smaller than the SSRI or SNRI literature, and much of it predates modern reporting standards. Effect-size comparisons across treatment classes should be read with that asymmetry in mind. ### What the evidence does not prove The evidence doesn't show that buspirone matches the effect size of SSRIs or SNRIs across anxiety disorders. It doesn't support buspirone as a strong treatment for panic disorder or social anxiety disorder. It doesn't predict which individual patient will respond. ### Practical interpretation Buspirone is a reasonable, lower-risk option for generalized anxiety disorder, particularly when antidepressants aren't tolerated or when an add-on is needed. It isn't a fast-acting medication and isn't a first-line choice for panic or social anxiety disorder. Medication decisions belong with a prescribing clinician. ### Sources - Slee A, et al. Pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. Lancet, 2019. PMID: 30797576 (https://pubmed.ncbi.nlm.nih.gov/30797576/) - NICE GAD guideline (CG113) (https://www.nice.org.uk/guidance/cg113) - US Food and Drug Administration, buspirone product labeling (https://www.accessdata.fda.gov/drugsatfda_docs/label/2010/018731s051lbl.pdf) ## Combined treatment evidence URL: https://anxietyresearch.org/treatment-research/combined-treatment-evidence Last updated: 2026-05-18 Does combining psychotherapy and medication work better than either one alone for anxiety disorders, and when does it matter most. ### Research question Does combining psychotherapy and medication work better than either one alone for anxiety disorders? ### What the evidence suggests - Combined treatment, usually meaning CBT plus an SSRI or SNRI, has been compared directly with each treatment alone in trials and meta-analyses. The general pattern is that combined treatment can offer an advantage over monotherapy in the short term, and the advantage tends to be clearer for more severe presentations. For milder anxiety, the added benefit of combining treatments is smaller, and either treatment alone is often sufficient. - The long-term picture is more nuanced. Once treatment ends, the gap between combined treatment and CBT alone narrows. Some evidence suggests CBT carries a relapse-protection advantage that medication alone doesn't, because the skills remain available to the patient after treatment stops. So combined treatment's short-term edge doesn't always translate into a lasting edge over CBT. - Sequencing also matters. Some patients respond well to medication first and add therapy once symptoms have eased enough to engage with it. Others can tolerate exposure-based CBT only after a partial medication response. The evidence doesn't identify one correct sequence for everyone. ### Important limitations Combined-treatment trials are smaller in number than monotherapy trials and tend to have limited long-term follow-up. Comparator design varies across studies, which complicates cross-trial comparison. ### What the evidence does not prove The evidence doesn't show that combined treatment is the best choice for every patient. For milder anxiety it often adds cost and burden without a proportional gain. It doesn't establish a single optimal order for starting therapy and medication. ### Practical interpretation Combined CBT and medication is a reasonable choice for moderate to severe anxiety, where the short-term advantage is clearest. For milder anxiety, a single first-line treatment is often enough. Whether to combine treatments, and in what order, is an individualized clinical decision that weighs severity, preference, access, and prior response. ### Sources - Bandelow B, et al. Efficacy of treatments for anxiety disorders: a meta-analysis. International Clinical Psychopharmacology, 2015. PMID: 25932596 (https://pubmed.ncbi.nlm.nih.gov/25932596/) - NICE GAD guideline (CG113) (https://www.nice.org.uk/guidance/cg113) - American Psychiatric Association, clinical practice guidelines (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) ## Measurement-based care URL: https://anxietyresearch.org/treatment-research/measurement-based-care Last updated: 2026-05-18 What measurement-based care is, what the evidence shows, and why it's increasingly treated as a marker of good-quality mental health care. ### Research question Does using a validated symptom scale at each visit improve mental health treatment outcomes? ### What the evidence suggests - Measurement-based care is the practice of using a validated rating scale at each visit to track a patient's symptoms over time, and using those scores to guide treatment decisions. For anxiety, the most common scale is the GAD-7. - Measurement-based care has been studied mostly in depression, where trials and reviews find it improves outcomes compared with usual care that relies on clinical impression alone. The reasoning is straightforward. A number measured the same way at each visit catches a lack of progress earlier than memory or general impression does. That earlier signal prompts earlier adjustment, such as changing a dose, switching a medication, or adding therapy. - The evidence base specific to anxiety is smaller than the depression evidence, but it points in the same direction. Measurement-based care is increasingly recommended as a general standard of good practice across mental health conditions, and several professional groups and quality frameworks have endorsed it. - Measurement-based care is also relevant to how care is delivered at scale. It gives a clinic a consistent way to see whether patients are improving, which supports quality monitoring and stepped-care models. It fits naturally with telepsychiatry, where a brief scale can be completed before a video visit. ### Important limitations The anxiety-specific evidence is less developed than the depression evidence. Measurement-based care also depends on consistent use, and in practice scales aren't always administered at every visit. ### What the evidence does not prove The evidence doesn't show that scores alone should drive decisions. A number supports clinical judgment, it doesn't replace it. ### Practical interpretation Measurement-based care is a low-cost practice that helps catch stalled progress earlier and supports more responsive treatment. It's increasingly treated as a marker of good-quality mental health care. The scores are a tool for the clinician and patient to use together, not a substitute for clinical assessment. ### Sources - Fortney JC, et al. A tipping point for measurement-based care. Psychiatric Services, 2017. PMID: 27582237 (https://pubmed.ncbi.nlm.nih.gov/27582237/) - Spitzer RL, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of Internal Medicine, 2006. PMID: 16717171 (https://pubmed.ncbi.nlm.nih.gov/16717171/) ## Treatment adherence research URL: https://anxietyresearch.org/treatment-research/treatment-adherence-research Last updated: 2026-05-18 How well people stay on anxiety treatment, why so many stop early, and why adherence is a central reason real-world results lag trial results. ### Research question How well do people stay on anxiety treatment, and what does the research say about why adherence matters? ### What the evidence suggests - Adherence to anxiety treatment, both medication and therapy, is incomplete in real-world settings. A meaningful share of patients stop antidepressants within the first few months, often before the medication has had a full chance to work. Therapy has dropout as well, since CBT requires attending sessions and doing practice between them. - The research identifies several recurring reasons. Side effects are a major driver of stopping medication, particularly sexual side effects with SSRIs. The slow onset of antidepressants matters too. A person who expects fast relief and feels none in the first two weeks may stop before the four-to-six-week window in which benefit usually appears. Cost, insurance gaps, the burden of appointments, stigma, and simply feeling better and deciding the treatment is no longer needed all contribute. - Adherence is also why real-world results often fall short of trial results. A treatment can have strong efficacy in a trial and a weaker effect in practice, not because the treatment changed, but because fewer people took it as intended. ### Important limitations Adherence research depends on imperfect measurement. Self-report and pharmacy records each have limits, and the two often disagree. Most adherence studies are observational rather than randomized. ### What the evidence does not prove The research doesn't show a single intervention that solves adherence. It doesn't establish that non-adherence is mainly a patient failing. The causes are often structural, including cost and access. ### Practical interpretation Adherence is a central reason treatments underperform outside of trials. Understanding the common reasons people stop, especially the slow onset of antidepressants and the burden of side effects, helps set realistic expectations. Measurement-based care and clear communication about what to expect can support adherence. Decisions about continuing or changing treatment belong with a clinician. ### Sources - World Health Organization. Adherence to Long-Term Therapies: Evidence for Action. 2003 (https://iris.who.int/handle/10665/42682) - Sansone RA, Sansone LA. Antidepressant adherence: are patients taking their medications? Innovations in Clinical Neuroscience, 2012. PMID: 22808448 (https://pubmed.ncbi.nlm.nih.gov/22808448/) ## Why outcomes vary between individuals URL: https://anxietyresearch.org/treatment-research/why-outcomes-vary Last updated: 2026-05-18 Why two people with the same anxiety diagnosis given the same treatment can have very different results, and what that means for reading a study. ### Research question Why do two people with the same anxiety diagnosis, given the same treatment, often have different outcomes? ### What the evidence suggests - Anxiety research reports averages. A meta-analysis that finds a moderate effect size for a treatment is describing what happened across a whole group. Inside that average, outcomes vary widely. Some people improve a great deal, some a little, and some not at all. The average doesn't predict where any one person will land. - Several sources of variation are well documented. People differ biologically in how they respond to a given medication. They differ in the severity and duration of their anxiety at the start. They differ in comorbidity, since a person with anxiety plus depression or a substance use disorder often follows a harder course than a person with anxiety alone. They differ in social circumstances, including stress, support, and the ability to attend appointments. And they differ in adherence to the treatment that was prescribed. - This isn't a weakness in the research. It's a real feature of the conditions. It's also why clinical trials report a range and not only an average, and why measures like the number needed to treat exist. Those measures make explicit that a treatment helps some people and not others. ### Important limitations Most trials are designed for group-level inference. Individual predictive models remain less accurate than average effect estimates, and follow-up windows are often shorter than the course of illness. ### What the evidence does not prove Population-level research can't tell an individual which treatment will work for them. There is no validated test that reliably predicts an individual's response to a specific anxiety treatment in advance. Genetic and biomarker prediction is an active research area, but it isn't yet a routine clinical tool. ### Practical interpretation The honest reading of anxiety research is that it narrows the options and ranks them by average benefit, but it can't make the choice for a specific person. This is the central reason treatment is a collaborative process between patient and clinician, often involving trying a first-line option, measuring the response, and adjusting. A study can inform that process. It can't replace it. ### Sources - Kessler RC, et al. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 2005. PMID: 15939837 (https://pubmed.ncbi.nlm.nih.gov/15939837/) - Cochrane Handbook for Systematic Reviews of Interventions (interpreting effect sizes and heterogeneity) (https://training.cochrane.org/handbook) ## Panic disorder treatment research URL: https://anxietyresearch.org/treatment-research/panic-disorder-treatment-evidence Last updated: 2026-05-18 How the effective treatments for panic disorder compare, how they're combined, and how they're sequenced in practice. ### Research question Panic disorder has more than one effective treatment. How do they compare, how are they combined, and how are they sequenced? ### What the evidence suggests - The two treatments with the strongest evidence for panic disorder are cognitive behavioral therapy that includes interoceptive exposure, and SSRIs. Head-to-head, meta-analyses don't show one to be reliably superior to the other for short-term symptom reduction. Both work for many patients. - The more interesting research question is about combining and sequencing them. Studies of combined treatment, CBT plus medication, suggest a short-term advantage over either treatment alone, and the advantage is clearest for more severe panic. That advantage narrows over the longer term. Some evidence points to CBT carrying a relapse-protection benefit that medication alone doesn't, because the interoceptive exposure skills remain available to the patient after treatment ends. - Sequencing has been studied directly in panic. Some patients tolerate exposure-based CBT more easily after a partial medication response has taken the edge off the worst symptoms. Others prefer to start with CBT and add medication only if it's needed. The research doesn't name one correct order. - One research direction worth knowing about is the use of CBT to support coming off medication. Stepping medication down while CBT skills consolidate has been studied as a way to capture both the speed of medication and the durability of therapy. The picture is promising but isn't yet settled. - Guidelines from the APA and NICE list CBT with interoceptive exposure and SSRIs as first-line for panic disorder, and they treat combined treatment as a reasonable option for more severe presentations. Benzodiazepines are positioned for short-term use, not long-term monotherapy. ### Important limitations Long-term comparative follow-up beyond one to two years is limited. Trials often exclude people with significant comorbidity, so the sequencing findings may not transfer cleanly to complex cases. ### What the evidence does not prove The research can't identify which treatment, or which sequence, is best for a specific patient. It doesn't establish a single correct order for starting CBT and medication, and it doesn't show that combined treatment is necessary in every case. ### Practical interpretation For panic disorder, the practical questions are usually not whether anything works but which treatment to start with and whether to combine. Both CBT with interoceptive exposure and SSRIs are reasonable starting points, and combined treatment is an option when panic is severe. These are decisions for a patient and clinician to make together. ### Sources - American Psychiatric Association, practice guideline for the treatment of panic disorder (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) - Bandelow B, et al. Efficacy of treatments for anxiety disorders: a meta-analysis. International Clinical Psychopharmacology. 2015. PMID: 25932596 (https://pubmed.ncbi.nlm.nih.gov/25932596/) - Mitte K. A meta-analysis of the efficacy of psycho- and pharmacotherapy in panic disorder with and without agoraphobia. Journal of Affective Disorders. 2005. PMID: 16005982 (https://pubmed.ncbi.nlm.nih.gov/16005982/) ## GAD pharmacotherapy beyond SSRIs and SNRIs URL: https://anxietyresearch.org/treatment-research/gad-pharmacotherapy-beyond-ssri-snri Last updated: 2026-05-18 What the research shows about the medications used in GAD when first-line SSRIs and SNRIs don't work or aren't tolerated. ### Research question SSRIs and SNRIs are first-line medication for generalized anxiety disorder, but they don't work for everyone. What does the evidence show about the other medications used in GAD? ### What the evidence suggests - Pregabalin has the most substantial evidence among the non-SSRI, non-SNRI options. Multiple randomized trials support it, and the 2019 Lancet network meta-analysis of GAD drug treatments included it among the effective agents. Pregabalin is approved for GAD in Europe. It isn't FDA-approved for GAD in the United States, so its use here for anxiety is off-label. It has a faster onset than SSRIs and doesn't cause the sexual side effects that drive much SSRI discontinuation, but it has its own profile: dose-related sedation, dizziness, and weight gain, plus a controlled-substance status (Schedule V) tied to misuse and dependence concerns, particularly in patients with prior substance use disorder. - Buspirone has older trial evidence in GAD. It doesn't cause dependence or sedation, and it doesn't produce a withdrawal syndrome. Its effect size is modest, and its evidence is specific to GAD rather than to panic or social anxiety. It's often used as an add-on to a partially effective SSRI, or as a primary option when a patient can't tolerate SSRIs. - Hydroxyzine, an antihistamine, has randomized evidence for short-term symptom relief in GAD. It isn't habit-forming. Its main drawback is sedation, and it's generally used for shorter-term or as-needed relief rather than as a long-term primary treatment. - Quetiapine, an antipsychotic, has randomized trial evidence in GAD. Despite that evidence, it's rarely an early choice, because its side-effect burden, including metabolic effects, weight gain, and sedation, is substantial relative to the alternatives. Guidelines generally hold it in reserve. - These medications enter the picture in a fairly consistent pattern: after a first-line SSRI or SNRI has been given an adequate trial, at an adequate dose, for an adequate length of time, and has either not worked or not been tolerated. The choice among them depends on the specific reason the first-line option failed, the patient's other conditions, and the side-effect trade-offs each one carries. ### Important limitations Head-to-head trials comparing these agents directly are limited, and long-term safety data are stronger for SSRIs and SNRIs than for these second-line options. Several of the U.S. uses described here are off-label, which means prescribing rests on the trial evidence and clinical judgment rather than on an FDA indication for GAD. ### What the evidence does not prove The research doesn't establish a clear ranking among these options for an individual patient. It doesn't show that any of them outperforms an SSRI or SNRI that has been given an adequate trial. ### Practical interpretation These medications come into play after first-line treatment hasn't worked or hasn't been tolerated. None of this is individualized advice. Medication decisions belong with a prescribing clinician. ### Sources - Slee A, et al. Pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. Lancet. 2019. PMID: 30712879 (https://pubmed.ncbi.nlm.nih.gov/30712879/) - NICE guideline CG113, generalised anxiety disorder and panic disorder in adults (https://www.nice.org.uk/guidance/cg113) - US Food and Drug Administration, product labeling for buspirone, hydroxyzine, and quetiapine (https://www.accessdata.fda.gov/scripts/cder/daf/) ## Mindfulness-based interventions in anxiety treatment URL: https://anxietyresearch.org/treatment-research/mindfulness-based-interventions-evidence Last updated: 2026-05-18 Where mindfulness fits in anxiety treatment, how guidelines frame it, and why the format of the mindfulness program matters. ### Research question Mindfulness-based interventions are widely promoted for anxiety. Where do they fit alongside the established treatments, how do guidelines treat them, and what separates the formats that have evidence from the ones that don't? ### What the evidence suggests - The single most useful distinction in the mindfulness evidence is the format. The trials that support mindfulness for anxiety mostly studied structured programs: mindfulness-based stress reduction and mindfulness-based cognitive therapy, each delivered as a roughly eight-week course with a trained instructor and a defined curriculum. Those structured programs show small to moderate reductions in anxiety symptoms. - App-based and informal mindfulness is a different and weaker evidence base. The trials are more mixed, the effects are smaller and less consistent, and the programs vary enormously in content and dose. The word "mindfulness" covers both, which is why a general claim that mindfulness helps anxiety is too loose to be useful. A structured eight-week course and a ten-minute phone app aren't the same intervention, and the evidence doesn't treat them as interchangeable. - Major clinical guidelines don't position mindfulness as a first-line, stand-alone treatment for a moderate to severe anxiety disorder. CBT and first-line medication hold that position. Where mindfulness appears in guidance, it's generally as one option among broader approaches, or as a complement to first-line treatment, rather than a replacement for it. Meta-analytic comparisons tend to find mindfulness roughly comparable to other active interventions rather than superior to them. - Mindfulness-based interventions are most reasonably used as one component of a plan, especially for milder symptoms, for general stress, or alongside first-line treatment. They carry low risk and broader well-being benefits. ### Important limitations Many mindfulness trials lack active controls and have small sample sizes. Heterogeneity in protocol, instructor training, and outcome measures complicates synthesis. Because blinding is difficult in mindfulness trials, expectation effects are hard to rule out. ### What the evidence does not prove The research doesn't show mindfulness outperforming first-line treatments. It doesn't establish that brief or app-based mindfulness reproduces the results of the structured programs. ### Practical interpretation The practical caution is to match expectations to the format. A structured eight-week MBSR or MBCT course is the version with evidence behind it. A reader choosing mindfulness as their main response to a diagnosed anxiety disorder should know that guidelines don't put it in the first-line slot, and that the casual app version isn't carrying the same evidence as the structured course. ### Sources - Goyal M, et al. Meditation programs for psychological stress and well-being: a systematic review and meta-analysis. JAMA Internal Medicine. 2014. PMID: 24395196 (https://pubmed.ncbi.nlm.nih.gov/24395196/) - Hofmann SG, et al. The effect of mindfulness-based therapy on anxiety and depression: a meta-analytic review. Journal of Consulting and Clinical Psychology. 2010. PMID: 20350028 (https://pubmed.ncbi.nlm.nih.gov/20350028/) - American Psychiatric Association, clinical practice guidelines (https://www.psychiatry.org/psychiatrists/practice/clinical-practice-guidelines) # Research literacy ## What is a meta-analysis? URL: https://anxietyresearch.org/research-literacy/what-is-a-meta-analysis Last updated: 2026-05-18 A meta-analysis pools results from multiple studies that asked the same question. The output is two numbers. The first is a single overall estimate of the effect. The second, called heterogeneity, measures how much the included studies disagreed with each other. The second number often matters as much as the first. ### What it is A meta-analysis isn't just more data is better. It is a formal pooling that weights studies by sample size and precision, and that quantifies disagreement between studies (heterogeneity). A meta-analysis differs from a single study in that it summarizes a literature. It differs from a systematic review in that it adds quantitative synthesis on top of the structured search and appraisal process. ### Why it sits at the top of the evidence hierarchy A well-conducted meta-analysis of randomized trials gives a more precise and stable answer than any single trial. It also makes biases such as publication bias and small-study effects visible rather than hidden. ### How it can mislead Garbage in, garbage out. A meta-analysis of weak studies produces a precise-looking estimate of a weak signal. Publication bias can inflate effect sizes. Different inclusion criteria can give different answers. ### Why this matters Most claims about what works for anxiety draw on meta-analyses. Reading them carefully helps separate strong from weak conclusions even when the headline number sounds similar. ### Sources - Cochrane Handbook for Systematic Reviews of Interventions (https://training.cochrane.org/handbook) - Higgins and Thompson on heterogeneity (https://pubmed.ncbi.nlm.nih.gov/12111919/) - PRISMA reporting standard (https://www.prisma-statement.org/) ## Correlation vs causation URL: https://anxietyresearch.org/research-literacy/correlation-vs-causation Last updated: 2026-05-18 A correlation is a statistical association between two variables. A causal claim says that changing one variable changes the other. Most cross-sectional and observational studies show correlations. ### What separates a correlation from a causal claim Demonstrating causation requires a study design that can rule out reverse causation, confounding, and selection bias. ### Anxiety-specific example Studies showing a correlation between adolescent social media use and anxiety symptoms can't, on their own, establish that social media use causes anxiety. Reverse causation, anxious teens use more social media, is a live alternative. Confounding factors such as sleep disruption and family environment are also alive. Selection effects matter too. ### Designs that support causal inference Randomized controlled trials, natural experiments, and well-instrumented causal-inference studies are the designs that most strongly support causal claims. ### Why this matters Many anxiety headlines turn a correlation into a causal claim in a single sentence. Reading the underlying study design quickly tells you whether that move was earned. ### Sources - Hill criteria for causation (https://pubmed.ncbi.nlm.nih.gov/14283879/) ## Statistical vs clinical significance URL: https://anxietyresearch.org/research-literacy/statistical-vs-clinical-significance Last updated: 2026-05-18 A statistically significant finding is one that is unlikely to be the result of chance alone, given the sample size and the statistical test used. A clinically significant finding is one that meaningfully changes patient outcomes or decisions. ### Why the two can diverge Large studies can find statistically significant effects that are clinically trivial. Small studies can find effects that look promising but lack precision. ### Key concepts Effect size, confidence interval, number needed to treat, and minimal clinically important difference are the core concepts that bridge statistical and clinical significance. ### Anxiety trial example A 2-point drop on a 21-point anxiety scale can be statistically significant in a large trial but may not represent a meaningful clinical change for an individual patient. ### Why this matters A trial can be technically positive and clinically modest. Looking at effect sizes rather than only p-values keeps this distinction visible. ### Sources - Cochrane Handbook for Systematic Reviews of Interventions (https://training.cochrane.org/handbook) - GRADE working group (https://www.gradeworkinggroup.org/) ## Why mental health studies disagree URL: https://anxietyresearch.org/research-literacy/why-mental-health-studies-disagree Last updated: 2026-05-18 Studies on the same intervention can reach different conclusions for several non-mysterious reasons: different patient populations, different outcome measures, different follow-up windows, different definitions of response or remission, different control conditions, and chance. ### Where the disagreement actually comes from When two systematic reviews of the same intervention reach different conclusions, the difference is usually traceable to inclusion criteria and outcome definitions, not to one being right and the other wrong. ### Practical advice When conflicting headlines appear, look at the underlying review's inclusion criteria, outcome measure, and follow-up window before reading the result. ### Why this matters Apparent contradictions in anxiety research often dissolve once you read the methods section. Knowing where to look saves time and prevents whiplash. ### Sources - Ioannidis on conflicting reviews (https://pubmed.ncbi.nlm.nih.gov/16060722/) - Cochrane Handbook (https://training.cochrane.org/handbook) ## Why one study shouldn't change treatment URL: https://anxietyresearch.org/research-literacy/why-one-study-should-not-change-treatment Last updated: 2026-05-18 Single studies, even high-profile ones, shouldn't change clinical practice in isolation. Reasons include chance findings, p-hacking, publication bias, replication failure, and the difference between population-level effects and individual treatment decisions. ### Why guidelines move slower than headlines Clinical guidelines move slower than the literature for this reason. Guideline bodies wait for replication and synthesis before changing recommendations. ### What a reasonable response to a striking single study looks like Note it, watch for replication, and don't change a medication or therapy plan on the basis of one paper. ### Why this matters Studies that look definitive often look smaller in the next replication. A pace that waits for confirmation is part of how clinical decision-making manages that pattern. ### Sources - Ioannidis, Why Most Published Research Findings Are False (https://pubmed.ncbi.nlm.nih.gov/16060722/) - Open Science Collaboration replication studies (https://pubmed.ncbi.nlm.nih.gov/26315443/) - GRADE working group (https://www.gradeworkinggroup.org/) ## How to read psychiatric study headlines URL: https://anxietyresearch.org/research-literacy/how-to-read-psychiatric-headlines Last updated: 2026-05-18 Headlines about mental health studies often outrun the underlying evidence. A short triage checklist helps a non-clinician read coverage more carefully. ### The checklist Was it animal or human? Mouse data doesn't transfer directly to clinical practice. Was it randomized or observational? Observational studies show association, not causation. How big was the sample? Single small studies are easy to overinterpret. Was it preregistered? Preregistered protocols reduce p-hacking risk. Was the outcome measure clinically meaningful, such as symptom-scale change or remission rate, or a surrogate, such as a biomarker or brain imaging signal? Was the comparator a placebo, an active control, or treatment as usual? Each gives a different read. What does the headline absolute risk reduction look like versus the relative risk reduction? Relative numbers always look bigger. Has the finding been replicated? ### Why this matters Most readers of anxiety news won't read the source paper. A reliable triage saves time and helps the source paper get a fair reading when it does matter. ### Sources - STROBE reporting standard (https://www.strobe-statement.org/) - CONSORT reporting standard (http://www.consort-statement.org/) ## Levels of evidence URL: https://anxietyresearch.org/research-literacy/levels-of-evidence Last updated: 2026-05-18 Different study designs support different strengths of claim. The evidence hierarchy organizes that variation. ### The hierarchy, from weakest to strongest Expert opinion and case reports are useful for hypothesis generation and weak for causal claims. Cross-sectional and case-control studies show association at a point in time. Cohort studies follow groups over time and are stronger for inference. Randomized controlled trials eliminate many sources of bias through random assignment. Systematic reviews and meta-analyses pool randomized trials with formal methods. Clinical practice guidelines synthesize evidence with explicit grading and recommendations. ### Why guidelines lag the literature Guideline bodies wait for replication, weigh harms, and require formal voting and review. This is a feature, not a bug. ### How AnxietyResearch.org weights sources The site prioritizes systematic reviews and meta-analyses, then clinical guidelines, then large epidemiologic datasets, then individual peer-reviewed studies. See the Methods page for the full hierarchy. ### Why this matters When two sources contradict, the higher level of evidence usually carries more weight in clinical reasoning. Knowing the order makes that judgment easier. ### Sources - GRADE working group (https://www.gradeworkinggroup.org/) - AHRQ Effective Health Care Program (https://effectivehealthcare.ahrq.gov/) # Glossary ## Randomized controlled trial URL: https://anxietyresearch.org/glossary/randomized-controlled-trial Category: Study design A randomized controlled trial, often shortened to RCT, is a study that tests whether a treatment works by splitting people into groups at random. One group gets the treatment. Another group gets a comparison, such as a placebo or usual care. Why it matters: Most of what we know about anxiety treatments comes from RCTs. When a guideline calls a therapy or a medication "first-line," that ranking usually rests on a stack of RCTs. A study that isn't randomized can still be useful, but it can't rule out the chance that the groups were different before anything started. How it appears: Researchers use RCTs to establish the efficacy of new medications or psychotherapies for anxiety. In more detail: The word "random" is the important part. A computer, not a researcher, decides which group each person joins. This matters because it spreads out the things that could tilt the result, such as age, how severe someone's anxiety is, or whether they have other health problems. When the groups start out alike, a difference at the end is more likely to come from the treatment itself. Example: A trial of cognitive behavioral therapy for panic disorder might enroll 200 people. A computer assigns half to start CBT now and half to a waitlist. After twelve weeks, the researchers compare panic attack counts between the two groups. What to watch for: A single RCT is one data point. Small size, short follow-up, and a weak comparison group, such as a waitlist instead of a real alternative treatment, all limit how much weight a trial deserves. A treatment looks strongest when several solid RCTs point the same way. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Systematic review URL: https://anxietyresearch.org/glossary/systematic-review Category: Study design A systematic review is a careful, organized summary of all the studies that have looked at the same question. It follows a written plan, so the process can be checked and repeated by someone else. Why it matters: When you want to know what the evidence says about an anxiety treatment, a systematic review is usually a better starting point than any single study. Guideline groups such as the American Psychiatric Association and NICE lean heavily on them. The Cochrane Library is one well-known home for systematic reviews. How it appears: Often used to summarize the overall effectiveness of a particular class of anxiety treatment over decades of research. In more detail: A regular review is just an expert sharing an opinion about the research. A systematic review works differently. The authors decide in advance what question they're asking, where they'll search, and which studies count. Then they search, screen, and grade every study against those rules. The point is to remove cherry-picking, where a writer mentions only the studies that fit their view. Example: A systematic review of exposure therapy for specific phobias would search the medical databases, find every trial that tested it, and report what the whole body of evidence shows. That includes the weak studies, not only the strong ones. What to watch for: A systematic review is only as good as the studies inside it. If the underlying trials are small or poorly run, the review should say so. Two reviews of the same topic can also reach different answers if they used different rules for which studies to include. Read the methods, not only the conclusion. Source: PRISMA 2020 statement, reporting guideline for systematic reviews. (https://www.prisma-statement.org/) ## Meta-analysis URL: https://anxietyresearch.org/glossary/meta-analysis Category: Study design A meta-analysis is a math method that combines the results of several studies into one overall number. It often sits inside a systematic review. Why it matters: Most headline claims about anxiety treatment, such as "CBT has a moderate to large effect," come from meta-analyses. Pooling studies gives a steadier answer than any one trial. It can also expose problems, such as publication bias, where studies with disappointing results never get published. How it appears: Used to quantify the exact degree of symptom reduction expected from therapies across diverse populations. In more detail: Each study on its own gives a rough answer. A meta-analysis pools those answers and gives more weight to the larger and more precise studies. The result is a single estimate of how well a treatment works, plus a measure of how much the studies agreed or disagreed with each other. Example: Ten trials each test an SSRI for generalized anxiety disorder. Some show a large benefit, some a small one. A meta-analysis combines all ten into one effect size, so you can see the average across the whole set rather than picking one trial. What to watch for: A meta-analysis can look precise even when it isn't. If it pools weak studies, the result is a tidy number built on shaky data. Strong disagreement between the studies, called heterogeneity, is a warning sign. And different inclusion rules produce different answers, so two meta-analyses of the same treatment won't always match. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Network meta-analysis URL: https://anxietyresearch.org/glossary/network-meta-analysis Category: Study design A network meta-analysis compares several treatments at once, including some that were never tested directly against each other in the same study. Why it matters: There are many anxiety treatments and few trials that test them head-to-head. Network meta-analysis helps fill that gap. A well-known example is the 2019 Lancet network meta-analysis of drug treatments for generalized anxiety disorder, which compared many medications in one analysis. How it appears: Used in anxiety research to rank SSRIs, SNRIs, and psychotherapies on a common scale. In more detail: A regular meta-analysis usually compares one treatment against one comparison. A network meta-analysis links many studies into a web. If treatment A was tested against B, and B against C, the method can estimate how A compares to C, even though no trial put A and C side by side. The output is often a ranking of treatments. Example: Trials exist for an SSRI against placebo, and for an SNRI against placebo, but few that test the SSRI against the SNRI directly. A network meta-analysis uses the shared placebo comparison to estimate how the two drugs stack up. What to watch for: The indirect comparisons rest on an assumption that the linked trials are similar enough to connect. If the trials studied very different patients, the network can mislead. Treatment rankings also tend to look more certain than they are. A small gap between two ranked treatments is often not meaningful. Source: Slee A, et al. Pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. Lancet. 2019. PMID: 30797576. (https://pubmed.ncbi.nlm.nih.gov/30797576/) ## Cohort study URL: https://anxietyresearch.org/glossary/cohort-study Category: Study design A cohort study follows a group of people over time to see what happens to them. It watches; it doesn't assign treatment. Why it matters: Some questions can't be tested with a randomized trial, for practical or ethical reasons. You can't randomly assign people to a stressful job or a chronic illness. Cohort studies are how researchers study questions like whether anxiety in early adulthood links to other health problems years later. How it appears: Useful for tracking how early life stress relates to the development of anxiety disorders in adulthood. In more detail: Researchers pick a group, the cohort, and track it forward in time. They compare people who had some exposure against people who didn't. Because no one is randomly assigned, a cohort study shows links, not proof of cause. It is an observational study, which means the researchers observe rather than control what happens. Example: A study follows 5,000 adults for ten years. It records who had an anxiety disorder at the start, then checks who later developed heart disease. It can report whether the two are linked. What to watch for: A cohort study can't rule out other explanations. If the anxious people in the study also slept less or had less access to care, any of those could drive the result instead. Researchers adjust for such factors with statistics, but adjustment is never perfect. Treat a cohort finding as a strong hint, not a verdict. Source: STROBE Statement, reporting guideline for observational studies. (https://www.strobe-statement.org/) ## Cross-sectional study URL: https://anxietyresearch.org/glossary/cross-sectional-study Category: Study design A cross-sectional study takes a snapshot of a population at one moment in time. It measures things once, not over a stretch of time. Why it matters: Most prevalence numbers, such as "about 19 percent of US adults had an anxiety disorder in the past year," come from large cross-sectional surveys. These studies are good at counting how common something is. They are weak at explaining cause. How it appears: Frequently used in public health surveys to estimate how many adults currently experience anxiety symptoms. In more detail: This design captures a single point. It can tell you how many people have anxiety right now, or how anxiety and some other trait appear together today. It can't tell you which one came first, because it never watches anything change. Example: A national survey asks 30,000 adults in one year whether they have anxiety symptoms and whether they sleep poorly. It can report that the two go together. It can't say whether poor sleep led to anxiety, anxiety led to poor sleep, or something else caused both. What to watch for: The main trap is reading a cross-sectional link as cause and effect. A snapshot can't show direction. It also can't separate brand-new cases from long-standing ones. Use cross-sectional studies for "how common" questions, not "what causes" questions. Source: STROBE Statement, reporting guideline for observational studies. (https://www.strobe-statement.org/) ## Case-control study URL: https://anxietyresearch.org/glossary/case-control-study Category: Study design A case-control study starts with the outcome. It compares people who already have a condition, the cases, with similar people who don't, the controls, then looks back to find differences. Why it matters: Case-control studies help researchers explore possible risk factors, such as childhood experiences or family history, especially for less common presentations. They're quicker and cheaper than long studies that follow people forward for years. How it appears: Used in early-stage research linking specific exposures to anxiety disorder onset. In more detail: This design works backward. Instead of following people forward in time, it begins with the result and searches the past for clues. It is efficient for rare conditions, because the researchers don't have to wait for new cases to appear. Example: Researchers gather 300 adults with panic disorder and 300 similar adults without it. They ask both groups about earlier life events, then compare what they find. What to watch for: Looking backward brings two problems. Memory is unreliable, and people with a condition may recall their past differently than people without it, which is called recall bias. Choosing a fair control group is also hard. A case-control study points to leads worth testing, not to firm answers. Source: STROBE Statement, reporting guideline for observational studies. (https://www.strobe-statement.org/) ## Intention-to-treat analysis URL: https://anxietyresearch.org/glossary/intention-to-treat-analysis Category: Study design Intention-to-treat means a trial counts every person in the group they were first assigned to, even if they stopped the treatment or dropped out partway through. Why it matters: Anxiety treatments, both therapy and medication, have real dropout rates. Medication can cause side effects. Therapy takes time and effort. An intention-to-treat analysis reflects what happens in ordinary conditions, not only among the people who completed everything. How it appears: Reported alongside per-protocol analyses in SSRI and CBT trials. In more detail: It sounds odd to keep counting someone who quit. There is a good reason. People often drop out because the treatment didn't help or caused side effects. If a trial counted only the people who finished, it would quietly discard the bad results and make the treatment look better than it is. Intention-to-treat keeps the comparison honest. Example: A trial assigns 100 people to an SSRI. Twenty stop early because of side effects. An intention-to-treat analysis still counts all 100 in the SSRI group when measuring the result. What to watch for: Its opposite is a per-protocol analysis, which counts only the people who completed the treatment as planned. Per-protocol results usually look rosier. When a trial reports both, the intention-to-treat number is the more trustworthy one for everyday decisions. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Pragmatic trial URL: https://anxietyresearch.org/glossary/pragmatic-trial Category: Study design A pragmatic trial tests a treatment in everyday conditions, with the kind of patients and clinics you'd see in normal care. Why it matters: A treatment can shine in a strict research setting and then disappoint in a busy primary care clinic. Pragmatic trials help close that gap. They're especially useful for questions about how care is delivered, such as measurement-based care or telepsychiatry. How it appears: Used to evaluate measurement-based care programs and telepsychiatry workflows. In more detail: Many trials are tightly controlled. They enroll a narrow group of patients, exclude anyone with other health problems, and deliver the treatment in ideal conditions. That answers the question "can this work." A pragmatic trial is built to answer a different question, "does this work in real life." It uses broad eligibility, regular clinics, and outcomes that matter to patients. Example: A pragmatic trial of telepsychiatry for anxiety would enroll typical patients across ordinary clinics, including people who also have other conditions, and measure outcomes patients care about, such as symptom relief and staying in treatment. What to watch for: The trade-off is control. Because pragmatic trials let real life in, they carry more variation and can be harder to interpret. Their opposite, the tightly controlled explanatory trial, gives a cleaner answer to a narrower question. Both are useful. They answer different things. Source: Ford I, Norrie J. Pragmatic Trials. New England Journal of Medicine. 2016. PMID: 27532833. (https://pubmed.ncbi.nlm.nih.gov/27532833/) ## Blinding URL: https://anxietyresearch.org/glossary/blinding Category: Study design Blinding means hiding who got which treatment, so that knowledge doesn't skew the results. Why it matters: Anxiety is measured mostly through what people report about their own symptoms, often on scales like the GAD-7. Self-reported outcomes are easy to nudge with expectation, which makes blinding especially important. It's also a known challenge for therapy trials. You can't fully hide from someone whether they're receiving talk therapy, so therapy trials often blind the outcome assessor instead. How it appears: Double-blind studies are standard for evaluating new pharmacological treatments for anxiety. In more detail: If patients know they got the real drug, their hopes can shift how they feel and what they report. If the researchers measuring the outcome know who got what, their judgment can drift too. Blinding hides the assignment. A trial can blind the patients, the clinicians, the people scoring the outcomes, or several of these at once. Example: In a drug trial, the placebo pill looks identical to the real one. Neither the patient nor the prescriber knows which is which until the trial ends. What to watch for: When a trial can't blind patients, expectation can inflate the result. Check who was blinded. A trial where the outcome assessor was blinded is stronger than one where everyone knew the assignment, even if the patients themselves couldn't be blinded. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Control group URL: https://anxietyresearch.org/glossary/control-group Category: Study design The control group is the comparison group in a study. It doesn't get the treatment being tested, so researchers have something to measure the treatment against. Why it matters: The type of control group changes what a trial can claim. A waitlist control, where people simply wait, sets a low bar, and treatments look strong against it. An active control, where people get a real alternative treatment or a placebo, sets a higher and fairer bar. How it appears: Patients receiving standard care or a placebo while another group receives a novel therapy. In more detail: A result means little on its own. If anxiety scores drop in a treatment group, you need to know what would have happened without the treatment. Anxiety often eases over time on its own. People tend to join trials when symptoms are at their worst, and symptoms usually improve from a peak. The control group captures that background change, so the treatment's real effect can be separated out. Example: A CBT trial with a waitlist control may show a large effect. The same CBT compared against another active therapy may show a smaller one. Both results can be true. They answer different questions. What to watch for: When you read a trial result, find out what the control group received. A big effect against a waitlist isn't the same as a big effect against an active treatment. The comparison sets the meaning of the number. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Placebo URL: https://anxietyresearch.org/glossary/placebo Category: Study design A placebo is an inactive treatment, such as a sugar pill, used for comparison. It contains no active drug. Why it matters: The placebo response in anxiety trials is large. Many people improve on placebo alone. This is why a medication has to clearly beat placebo, not merely help people, before it counts as effective. When you read that a drug has a "small to moderate effect," that effect is usually measured on top of an already sizable placebo response. How it appears: New anxiety medications must demonstrate efficacy significantly greater than that of a placebo. In more detail: Placebos exist because belief and expectation have real effects on how people feel. When someone takes a pill they expect to help, their symptoms often improve somewhat, even with no active medicine in it. This is the placebo effect. To know whether a drug works beyond expectation, a trial compares it against a placebo that looks and feels the same. Example: In an SSRI trial for panic disorder, both groups improve. The placebo group improves from expectation and natural recovery. The SSRI group improves more. The gap between the two is the drug's true effect. What to watch for: A trial with no placebo or comparison can't separate the drug from expectation. A strong placebo response is also not fake. The people genuinely feel better. It just isn't proof that the active treatment is what helped them. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Prevalence URL: https://anxietyresearch.org/glossary/prevalence Category: Statistics Prevalence is the share of a population that has a condition at a given time. It answers the question "how many people have this." Why it matters: Almost every headline anxiety statistic is a prevalence figure. When you read that about 19 percent of US adults had an anxiety disorder in the past year, that's past-year prevalence. When you read 31 percent at some point in life, that's lifetime prevalence. Confusing the two is one of the most common errors in reading anxiety data. How it appears: Reported as the percentage of adults diagnosed with an anxiety disorder within a given year. In more detail: Prevalence is usually written as a percentage. It can cover different windows of time. Point prevalence is how many people have the condition on a single day. Past-year prevalence is how many had it at any time in the last twelve months. Lifetime prevalence is how many have ever had it. The window changes the number, so the window matters. Example: A survey of 30,000 adults finds that 5,700 of them met criteria for an anxiety disorder in the past year. The past-year prevalence is 19 percent. What to watch for: Always check the window and the method. A prevalence number from a survey that uses a full diagnostic interview will differ from one that uses a short symptom screener. Higher isn't the same as more accurate. Symptom screeners tend to report higher numbers than diagnostic interviews. Source: CDC, Principles of Epidemiology in Public Health Practice. (https://www.cdc.gov/csels/dsepd/ss1978/) ## Incidence URL: https://anxietyresearch.org/glossary/incidence Category: Statistics Incidence is the rate at which new cases of a condition appear in a population over a period of time. It counts only the new cases. Why it matters: Incidence is harder to measure for anxiety than prevalence, because it means following people over time and catching the moment a disorder begins. Most anxiety statistics you see are prevalence. When a report says anxiety is rising, check whether it means more new cases each year, which is incidence, or simply more people carrying the condition, which is prevalence. How it appears: Used to measure whether anxiety disorders are being diagnosed more frequently year over year. In more detail: Incidence and prevalence are easy to mix up. Prevalence counts everyone who has the condition. Incidence counts only the people who developed it during a set window, usually a year. Incidence describes how fast a condition is appearing. Prevalence describes how much of it is present overall. Example: In a town of 100,000 adults, 800 develop an anxiety disorder for the first time during one year. The incidence is 0.8 percent per year. What to watch for: A condition can have steady incidence but rising prevalence if people stay ill longer or recover more slowly. So a rising prevalence doesn't always mean a condition is striking more people. Read carefully which measure a report uses. Source: CDC, Principles of Epidemiology in Public Health Practice. (https://www.cdc.gov/csels/dsepd/ss1978/) ## Odds ratio URL: https://anxietyresearch.org/glossary/odds-ratio Category: Statistics An odds ratio compares the odds of an outcome in one group with the odds in another. It's a way to measure how strongly two things are linked. Why it matters: Odds ratios show up when researchers study what is linked to anxiety, such as family history, sleep problems, or chronic illness. An odds ratio gives the direction and rough strength of a link. It doesn't prove that one thing caused the other. How it appears: Used to describe the likelihood of experiencing anxiety given a specific risk factor, like lack of sleep. In more detail: Odds aren't the same as risk. Risk is the chance something happens. Odds are the chance it happens divided by the chance it doesn't. An odds ratio of 1 means no difference between the groups. Above 1 means the outcome is more likely in the first group. Below 1 means it is less likely. Odds ratios appear most often in case-control studies. Example: A study reports that people with chronic pain have an odds ratio of 2.0 for also having an anxiety disorder. The odds of anxiety are about twice as high in the chronic pain group. What to watch for: When an outcome is common, an odds ratio can look larger than the real difference in risk. A reader can easily mistake an odds ratio of 2.0 for "twice the risk," which isn't quite what it means. An odds ratio always needs a confidence interval to show how precise it is. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Relative risk URL: https://anxietyresearch.org/glossary/relative-risk Category: Statistics Relative risk compares the chance of an outcome in one group with the chance in another. A relative risk of 1 means no difference. Why it matters: Relative risk appears in trial results and in cohort studies of anxiety. It's easier to read correctly than an odds ratio, which is why many researchers and readers prefer it. How it appears: Shows the relative risk of relapse when medication is discontinued versus maintained. In more detail: Relative risk, sometimes called risk ratio, is more intuitive than the odds ratio. It is the risk in one group divided by the risk in the other. A relative risk of 1.5 means the outcome is 50 percent more likely in the first group. A relative risk of 0.7 means it is 30 percent less likely. It's used in cohort studies and randomized trials, where actual risk can be measured directly. Example: In a cohort study, 6 percent of one group develops an anxiety disorder over five years, compared with 3 percent of another. The relative risk is 2.0. The first group's risk is twice as high. What to watch for: Relative risk hides the starting size of the risk. A relative risk of 2.0 sounds alarming, but if the baseline risk is tiny, doubling it is still a small absolute change. Always ask for the absolute numbers alongside the relative ones. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Hazard ratio URL: https://anxietyresearch.org/glossary/hazard-ratio Category: Statistics A hazard ratio compares how quickly an outcome happens in one group versus another over time. It's about timing, not only whether something happens. Why it matters: Hazard ratios appear in studies of relapse and of staying in treatment. A study might report how long people stay well after finishing therapy, and whether one treatment delays relapse longer than another. How it appears: Reported in long-term maintenance trials for anxiety and depression. In more detail: Relative risk asks whether an outcome happened by the end of a study. A hazard ratio asks how fast it happened along the way. It comes from studies that track the time until an event, such as a relapse. A hazard ratio of 1 means the same speed in both groups. Below 1 means slower. Above 1 means faster. Example: A study of relapse after anxiety treatment reports a hazard ratio of 0.6 for people who continued medication. Continuing treatment was linked to a slower rate of relapse over the follow-up period. What to watch for: A hazard ratio describes a rate over time, not a simple final count. It assumes the gap between groups stays roughly steady across the whole follow-up, which isn't always true. As with other ratios, check the confidence interval. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Confidence interval URL: https://anxietyresearch.org/glossary/confidence-interval Category: Statistics A confidence interval is a range of values that probably contains the true answer. It shows how precise a study's result is. Why it matters: The confidence interval is one of the most useful things to look at in any anxiety study, and one of the most ignored. Two trials can report the same effect size, one with a tight interval and one with a wide interval. The first is far more trustworthy. How it appears: Accompanies effect size estimates to show the range within which the true treatment effect likely falls. In more detail: A study reports one number, but that number is an estimate from a sample, not the exact truth. The confidence interval, usually a 95 percent interval, is the range the true value most likely falls within. A narrow interval means a precise estimate. A wide interval means an uncertain one. Example: A trial reports that a treatment lowered anxiety scores by 4 points, with a 95 percent confidence interval of 2 to 6. The best estimate is 4, but the true effect is probably somewhere between 2 and 6. What to watch for: If a confidence interval for a difference includes zero, the study can't rule out no effect. If an interval for a ratio includes 1, the same is true. A wide interval is a sign of a small or noisy study. Read the interval, not only the headline number. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## p-value URL: https://anxietyresearch.org/glossary/p-value Category: Statistics A p-value is a number that tells you how surprising a result would be if the treatment actually did nothing. A small p-value means the result would be unlikely by chance alone. Why it matters: P-values appear in nearly every anxiety trial, and they're widely misread. A p-value doesn't tell you the chance the treatment works. It doesn't tell you how big the effect is. It speaks only to chance. How it appears: Reported in clinical trial results alongside effect size and confidence interval. In more detail: Researchers often use a cutoff of 0.05. If the p-value is below 0.05, they call the result statistically significant. This doesn't mean the result is important or large. It means only that it is unlikely to be a fluke of chance, given the data. Example: A trial finds a treatment lowered anxiety scores more than placebo, with a p-value of 0.01. A difference this large would be uncommon if the treatment truly did nothing. What to watch for: A small p-value with a tiny effect is common in large studies and may not matter to a patient. A large p-value doesn't prove a treatment fails; the study may simply have been too small to detect a real effect. Pair the p-value with the effect size and the confidence interval. Source: Wasserstein RL, Lazar NA. The ASA Statement on p-Values. The American Statistician. 2016. (https://doi.org/10.1080/00031305.2016.1154108) ## Statistical significance URL: https://anxietyresearch.org/glossary/statistical-significance Category: Statistics A result is statistically significant when it's unlikely to be explained by chance alone. It is usually decided by a p-value below a set cutoff, often 0.05. Why it matters: News coverage often treats "statistically significant" as proof that a treatment works well. That's a misreading. A significant result in a very large trial can describe an effect too small for any patient to notice. How it appears: Often denoted by a p-value, indicating that a treatment's effect on anxiety wasn't a random occurrence. In more detail: Statistical significance answers one narrow question: could this result be a coincidence of sampling? If the answer is probably not, the result is called significant. That is all the term means. It says nothing about whether the effect is large, useful, or meaningful to a patient. Example: A trial of 5,000 people finds a treatment lowers anxiety scores by half a point on a 21-point scale, with a significant p-value. The result is real, but half a point is unlikely to change how anyone feels. What to watch for: Significance depends heavily on sample size. Big studies can make trivial effects significant. Small studies can miss real effects and call them non-significant. Always pair statistical significance with clinical significance, which asks whether the effect is big enough to matter. Source: Wasserstein RL, Lazar NA. The ASA Statement on p-Values. The American Statistician. 2016. (https://doi.org/10.1080/00031305.2016.1154108) ## Clinical significance URL: https://anxietyresearch.org/glossary/clinical-significance Category: Statistics Clinical significance asks whether a result is big enough to make a real difference to a patient's life. It is different from statistical significance. Why it matters: This is the gap that headlines miss most often. A drug or therapy can beat placebo by a statistically significant margin and still produce a change too small to feel. Clinical significance is the better question for a reader deciding whether a treatment is worth it. How it appears: Used to argue that a specific intervention reduces anxiety enough to improve a patient's functional abilities. In more detail: A study can show a result that is statistically significant, meaning unlikely to be chance, but still too small to matter. Clinical significance is about size and meaning. Does the change move someone from struggling to functioning? Would a patient or clinician actually notice it? Tools like the minimal clinically important difference help answer this. Example: A treatment lowers GAD-7 anxiety scores by 1 point, and the result is statistically significant in a large trial. But a noticeable change on the GAD-7 is closer to 4 points, so a 1-point change has little clinical significance. What to watch for: Ask two questions of a result, not one. Is it unlikely to be chance? And is it large enough to matter? A result can pass the first test and fail the second. Source: APA Dictionary of Psychology, "clinical significance." (https://dictionary.apa.org/) ## Effect size URL: https://anxietyresearch.org/glossary/effect-size Category: Statistics An effect size is a number that describes how large a treatment's effect is. It lets you compare results across different studies and different measures. Why it matters: When this site says CBT shows moderate to large effects for anxiety, that statement is about effect sizes pooled across trials. Effect size is how researchers compare a therapy with a medication, or one drug with another, even when the original studies used different scales. How it appears: Used to compare whether CBT or medication has a larger overall impact on anxiety reduction. In more detail: Trials measure anxiety in different ways and on different scales. An effect size puts results on a common ruler. A widely used one is Cohen's d. As a rough guide, around 0.2 is a small effect, 0.5 is moderate, and 0.8 or more is large. Another common form is the standardized mean difference, used in meta-analyses. Example: A meta-analysis reports that exposure therapy for specific phobia has an effect size of about 1.1 compared with waitlist. That's a large effect. What to watch for: Effect size labels are rough guides, not strict rules. A small effect can still matter at the population level, and a large effect against a weak comparison group can shrink against an active one. Check what the effect was measured against. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Number needed to treat (NNT) URL: https://anxietyresearch.org/glossary/number-needed-to-treat Category: Statistics The number needed to treat, or NNT, is how many people would need to receive a treatment for one extra person to benefit. Why it matters: NNT is one of the most useful ways to understand what a trial result means in real life. It's more concrete than an effect size or a p-value. It also makes clear that no treatment helps everyone. Even a good treatment has an NNT above 1. How it appears: Used in clinical summaries of antidepressant and psychotherapy efficacy. In more detail: NNT turns a trial result into a practical number. If a treatment has an NNT of 5, then on average 5 people need to take it for 1 more person to improve than would have improved anyway. A lower NNT means a more effective treatment. The best possible NNT is 1, where everyone treated benefits. Example: An anxiety medication has an NNT of 6 for a meaningful response. For every 6 people who take it, about 1 more responds than would have on placebo. What to watch for: NNT depends on the comparison, the outcome chosen, and the length of the study, so the same treatment can have different NNTs in different trials. It's most useful read next to the number needed to harm, which counts how many people are harmed by side effects. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Number needed to harm (NNH) URL: https://anxietyresearch.org/glossary/number-needed-to-harm Category: Statistics The number needed to harm, or NNH, is how many people would need to receive a treatment for one extra person to experience a particular harm, such as a side effect. Why it matters: Every anxiety treatment has trade-offs. Medications can cause side effects. NNH puts a number on how often. Comparing NNT and NNH helps a clinician and patient weigh whether a treatment's likely benefit is worth its likely harm. How it appears: Used in benzodiazepine, SSRI, and antipsychotic safety analyses. In more detail: NNH is the mirror image of the number needed to treat. NNT counts benefit. NNH counts harm. A high NNH is good, because it means harm is rare. A low NNH is concerning, because it means harm is common. Reading the two together gives a balanced picture of a treatment. Example: A medication has an NNT of 6 for a meaningful response and an NNH of 30 for a bothersome side effect. Benefit is more common than this particular harm, which supports its use for many patients, though not all. What to watch for: NNH depends entirely on which harm is being counted. A drug can have a high NNH for a rare serious harm and a low NNH for a common mild one. Check which harm the number describes. A single NNH doesn't capture every risk. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Heterogeneity URL: https://anxietyresearch.org/glossary/heterogeneity Category: Statistics Heterogeneity is the amount that study results differ from each other when they're combined in a meta-analysis. High heterogeneity means the studies disagreed. Why it matters: Anxiety trials vary in who they enroll, how they deliver treatment, and how they measure outcomes. That variety can produce real heterogeneity. When a meta-analysis of an anxiety treatment reports high heterogeneity, its average effect size deserves more caution. How it appears: Reported as I-squared or Q statistics in systematic reviews. In more detail: When a meta-analysis pools several trials, the trials rarely give identical results. A little spread is normal. A lot of spread is a problem. Researchers measure it with a statistic often called I-squared, written as a percentage. A high value means the studies are telling different stories, and the single pooled number may hide more than it shows. Example: A meta-analysis of mindfulness for anxiety pools 20 trials. The trials used different programs, different lengths, and different populations. The reported heterogeneity is high, which signals that "mindfulness" isn't one single, consistent thing across the studies. What to watch for: A pooled number from a high-heterogeneity meta-analysis can mislead. It may average together treatments or populations that shouldn't be averaged. When heterogeneity is high, look for whether the researchers explored why, often by analyzing subgroups. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Comorbidity URL: https://anxietyresearch.org/glossary/comorbidity Category: Clinical outcomes Comorbidity means having two or more health conditions at the same time. Why it matters: Comorbidity shapes how anxiety is studied and treated. Many anxiety trials exclude people with significant comorbid conditions, which makes the trials cleaner but less like real life. Comorbid anxiety and depression in particular is linked to greater impairment and a longer path to recovery than either condition alone. How it appears: Often discussed regarding the high rate of individuals experiencing both anxiety and depression. In more detail: Conditions often travel together. When a person has both an anxiety disorder and another condition, such as depression or a substance use disorder, those conditions are comorbid. In mental health, comorbidity is common; it isn't the exception. Example: A person with generalized anxiety disorder who also meets criteria for major depression has comorbid anxiety and depression. Treatment usually has to account for both rather than one at a time. What to watch for: When you read that a treatment works for anxiety, check whether the trial enrolled people with comorbid conditions or screened them out. A treatment proven in a comorbidity-free trial may behave differently in the many real patients who carry more than one condition. Source: Kessler RC, et al. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005. PMID: 15939837. (https://pubmed.ncbi.nlm.nih.gov/15939837/) ## Remission URL: https://anxietyresearch.org/glossary/remission Category: Clinical outcomes Remission means a condition's symptoms have dropped low enough that the person no longer meets the criteria for the diagnosis. The condition is quiet, though not necessarily gone for good. Why it matters: Remission is one of the most meaningful outcomes a trial can report, because it's closer to what patients actually want, which is to feel well, not just less unwell. A treatment can produce a good response rate but a lower remission rate. The gap between the two is worth noticing. How it appears: Studies report the percentage of patients achieving full remission after a course of therapy. In more detail: Remission is a higher bar than response. Response means symptoms improved by a meaningful amount. Remission means symptoms have fallen close to a normal range. Remission can be partial, where some symptoms remain, or full. It is usually defined by a score below a set cutoff on a symptom scale. Example: In an anxiety trial, remission might be defined as a GAD-7 score below 5. A trial could report that 60 percent of patients responded but only 35 percent reached remission. What to watch for: Different trials define remission with different cutoffs, so remission rates aren't always comparable across studies. Remission is also not the same as cure. Symptoms can return, which is called relapse. Check how long the trial followed people after remission. Source: APA Dictionary of Psychology, "remission." (https://dictionary.apa.org/) ## Response rate URL: https://anxietyresearch.org/glossary/response-rate Category: Clinical outcomes The response rate is the share of people in a study who improved by a set amount, often a 50 percent drop in symptoms. Why it matters: Response rate is a standard way trials report results. It's easy to picture: out of every 100 people treated, how many improved by a meaningful amount. Comparing the response rate in the treatment group with the response rate in the control group shows what the treatment added. How it appears: Used to compare the broad effectiveness of different SSRIs in clinical trials. In more detail: Response is a defined threshold, not just any improvement. A common definition is a 50 percent reduction on a symptom scale from where the person started. The response rate is the percentage of the group that crossed that line. It is a different and lower bar than remission. Example: A trial reports a 55 percent response rate for an SSRI and a 35 percent response rate for placebo. The treatment's added benefit is the 20-point gap, not the full 55 percent. What to watch for: A response rate on its own can mislead, because some people improve no matter what, as the placebo response rate shows. Always compare the treatment's response rate against the control group's. The gap is the real story. Also check how response was defined, since the threshold can vary. Source: APA Dictionary of Psychology, "treatment response." (https://dictionary.apa.org/) ## Relapse URL: https://anxietyresearch.org/glossary/relapse Category: Clinical outcomes Relapse is the return of a condition's symptoms after a period of improvement or remission. Why it matters: Relapse is why follow-up length matters so much in anxiety trials. A study that ends at twelve weeks shows whether a treatment works in the short term. It says nothing about whether the benefit lasts. Research that follows people for a year or more is far more informative about relapse. How it appears: Research evaluates strategies to prevent relapse following the completion of active CBT. In more detail: Getting better once isn't the end of the story. Anxiety disorders can come back. Relapse means symptoms have returned to the point of meeting the diagnosis again, after the person had improved. It is different from a brief bad day or a short rough patch. Example: A trial finds that many people reach remission with medication. After the trial, some stop the medication and their symptoms return. That return is relapse, and it's a common reason clinicians discuss continuing treatment past the point of feeling better. What to watch for: A short trial can't speak to relapse at all. When a treatment is described as durable or lasting, check how long the study actually followed people. Relapse rates also depend on whether treatment continued or stopped. Source: APA Dictionary of Psychology, "relapse." (https://dictionary.apa.org/) ## Treatment adherence URL: https://anxietyresearch.org/glossary/treatment-adherence Category: Clinical outcomes Treatment adherence is how closely a person follows a treatment as it was planned, such as taking a medication regularly or attending therapy sessions. Why it matters: Adherence shapes both research and real-world results. In trials, low adherence can make an effective treatment look weak. In daily life, a treatment that works in theory does nothing if a person can't stay on it. This is one reason real-world results often fall short of trial results. How it appears: Studied to understand why patients discontinue medications early or drop out of therapy. In more detail: A treatment can only work if it's actually used. Adherence covers taking the right dose at the right time, attending sessions, and continuing for the planned length. Non-adherence is common and has many causes, including side effects, cost, feeling better and stopping early, and simply forgetting. Example: An SSRI may need four to six weeks to take effect. A person who stops after two weeks because they feel no benefit yet has low adherence, and they may never learn whether the medication would have helped. What to watch for: When a treatment underperforms in practice, poor adherence is a common reason, separate from whether the treatment works. Trials that use an intention-to-treat analysis handle non-adherence more honestly than trials that count only the people who completed treatment. Source: World Health Organization. Adherence to Long-Term Therapies: Evidence for Action. 2003. (https://www.who.int/) ## Minimal clinically important difference URL: https://anxietyresearch.org/glossary/minimal-clinically-important-difference Category: Statistics The minimal clinically important difference, or MCID, is the smallest change on a symptom scale that a patient would actually notice and consider meaningful. Why it matters: The MCID is the tool that connects a trial result to a patient's experience. It is how you tell whether a significant improvement is also a meaningful one. For the GAD-7, a common anxiety scale, research points to a change of roughly 4 points as the rough mark of a noticeable improvement. How it appears: Used to interpret GAD-7 and PHQ-9 score changes during treatment. In more detail: Symptom scales turn how a person feels into a number. Not every change in that number matters. The MCID is the threshold where a change becomes large enough to feel real. A change below the MCID may be statistically detectable yet too small to matter in daily life. Example: A trial reports that a treatment improved GAD-7 scores by 2 points more than placebo. That's below the rough 4-point mark, so the average patient might not notice the extra benefit, even if the result is statistically significant. What to watch for: The MCID is an estimate, not a hard line, and it differs by scale and by study. Still, it's one of the best questions to ask of any trial: did the treatment beat placebo by enough to matter, not just by enough to reach statistical significance. Source: Spitzer RL, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006. PMID: 16717171. (https://pubmed.ncbi.nlm.nih.gov/16717171/) ## Dropout URL: https://anxietyresearch.org/glossary/dropout Category: Clinical outcomes Dropout is when a participant leaves a study before it ends, or a patient stops a treatment before completing it. Why it matters: Dropout rates are a quiet but important part of any anxiety trial. A treatment with strong results but very high dropout may be hard to tolerate in practice. Comparing dropout rates between the treatment group and the control group can reveal a side-effect burden that the headline result hides. How it appears: Routinely reported in psychotherapy and pharmacotherapy trials for anxiety. In more detail: Dropout is more than a missing data point. It often carries information. People frequently leave a study or stop a treatment because it isn't helping, or because the side effects are hard to tolerate. A high dropout rate is itself a finding about a treatment. Example: A medication trial reports good symptom improvement but a 30 percent dropout rate in the drug group versus 10 percent in the placebo group. The gap suggests side effects drove people out, which matters for real-world use. What to watch for: Watch how a trial handled dropouts. An intention-to-treat analysis keeps them in the count and gives a more honest result. A per-protocol analysis that drops them can make a hard-to-tolerate treatment look better than it is. Source: Cochrane Handbook for Systematic Reviews of Interventions, current edition. (https://training.cochrane.org/handbook) ## Dose-response URL: https://anxietyresearch.org/glossary/dose-response Category: Statistics A dose-response relationship means that more of something is linked to more of an effect. As the dose goes up, the response changes in step. Why it matters: Dose-response thinking helps judge both medication and therapy. For therapy, researchers ask whether more sessions produce more improvement, and whether there is a point where extra sessions stop adding much. For medication, they ask whether higher doses help more, or only add side effects. How it appears: Discussed for SSRIs, SNRIs, pregabalin, and structured CBT exposure dose. In more detail: Dose-response is a classic sign that a link might be causal. If a small dose produces a small effect, a medium dose a medium effect, and a large dose a large effect, that pattern is harder to explain away as coincidence. The dose doesn't have to be a drug. It can be the number of therapy sessions or the length of a program. Example: A study of CBT for anxiety finds that people who completed more sessions had larger improvements, up to a point, after which extra sessions added little. That's a dose-response pattern with a plateau. What to watch for: A dose-response pattern supports a causal link but doesn't prove one. People who attend more therapy sessions may differ in other ways, such as motivation or access. And more isn't always better. Many treatments reach a ceiling, and higher medication doses often raise side effects faster than benefit. Source: CDC, Principles of Epidemiology in Public Health Practice. (https://www.cdc.gov/csels/dsepd/ss1978/) ## Discontinuation syndrome URL: https://anxietyresearch.org/glossary/discontinuation-syndrome Category: Clinical outcomes Discontinuation syndrome is a set of symptoms that can appear when a person stops or quickly lowers an antidepressant, such as an SSRI or SNRI. Why it matters: SSRIs and SNRIs are first-line medications for anxiety, so discontinuation syndrome is a practical issue for many patients. It is a key reason clinicians lower these medications slowly, with a taper, instead of stopping all at once. It can also be mistaken for a return of anxiety, which can confuse decisions about treatment. How it appears: Most prominently described with short-half-life SSRIs and with venlafaxine. In more detail: After the body adjusts to an antidepressant, stopping it suddenly can cause temporary symptoms. These can include dizziness, flu-like feelings, sleep trouble, irritability, and brief electric-shock sensations. The symptoms are usually short-lived. They're more likely with medications that leave the body quickly, such as paroxetine and venlafaxine. Example: A person stops an SNRI abruptly and feels dizzy and unwell for a week. This is discontinuation syndrome, not necessarily a sign that their anxiety has come back. What to watch for: Discontinuation syndrome isn't the same as addiction in the way that word is usually meant. It's a predictable physical adjustment. It is also a reason not to stop an antidepressant on your own. A clinician can plan a taper that lowers the risk. Source: U.S. Food and Drug Administration product labeling for SSRI and SNRI antidepressants. (https://www.fda.gov/) ## Sustained response URL: https://anxietyresearch.org/glossary/sustained-response Category: Clinical outcomes A sustained response means an improvement that holds up over time, rather than fading after a few weeks. Why it matters: Patients don't want to feel better only briefly. Sustained response is closer to the goal of treatment than a short-term response. For therapies like CBT, one argued advantage is that the skills may keep working after treatment ends, which would show up as a more sustained response compared with stopping a medication. How it appears: Reported in maintenance trials and long-term CBT follow-up studies. In more detail: A treatment can produce a fast improvement that doesn't last. Sustained response is the opposite. It describes a benefit that is still present months later. It is measured by following people after they improve and checking whether they stay improved. Example: A trial follows people for a year after they responded to treatment. The share who are still well at one year is the sustained response rate. What to watch for: Sustained response can only be measured with long follow-up, and long follow-up is expensive, so many trials skip it. The absence of sustained-response data isn't proof a treatment fails to last. It often just means no one studied it long enough. Source: APA Dictionary of Psychology, "treatment response." (https://dictionary.apa.org/) ## Generalized anxiety disorder URL: https://anxietyresearch.org/glossary/generalized-anxiety-disorder-term Category: Treatment research A chronic condition characterized by excessive, uncontrollable worry about everyday issues. Why it matters: It is one of the most common mental health disorders, significantly impacting daily functioning. How it appears: The focus of studies examining long-term pharmacological and psychological management. ## Panic disorder URL: https://anxietyresearch.org/glossary/panic-disorder Category: Conditions Panic disorder is an anxiety disorder marked by repeated, unexpected panic attacks and ongoing worry about having more of them. Why it matters: Panic disorder is a common reason people end up in emergency rooms, because the physical symptoms can feel like a heart problem. Research on panic focuses on treatments that reduce how often and how hard the attacks hit, mainly CBT with interoceptive exposure, and SSRIs. How it appears: Research assesses interventions designed to reduce the frequency and severity of attacks. In more detail: A panic attack is a sudden surge of intense fear with strong physical symptoms, such as a pounding heart, shortness of breath, chest tightness, dizziness, and a sense of dread. In panic disorder, the attacks recur, and the person starts to fear the attacks themselves. That fear can lead to avoiding places or situations linked to past attacks, which can develop into agoraphobia. Example: A person has several panic attacks with no clear trigger, then begins avoiding the grocery store where one attack happened, for fear it will happen again. What to watch for: New panic-like symptoms should be evaluated, not assumed to be anxiety, especially the first time. Cardiac problems, thyroid conditions, and the effects of caffeine, stimulants, or alcohol withdrawal can mimic panic. A medical check helps rule those out before settling on a panic disorder explanation. Source: National Institute of Mental Health, Panic Disorder: When Fear Overwhelms. (https://www.nimh.nih.gov/health/publications/panic-disorder-when-fear-overwhelms) ## Social anxiety disorder URL: https://anxietyresearch.org/glossary/social-anxiety-disorder-term Category: Treatment research An intense, persistent fear of being watched and judged by others in social situations. Why it matters: It causes profound impairment in educational, occupational, and interpersonal domains. How it appears: Evaluated in trials testing specialized cognitive-behavioral protocols. ## Cognitive behavioral therapy URL: https://anxietyresearch.org/glossary/cognitive-behavioral-therapy Category: Treatment research Cognitive behavioral therapy, or CBT, is a structured, time-limited talk therapy. It works on the links between thoughts, feelings, and behavior. Why it matters: CBT is the most studied talk therapy for anxiety disorders. Across generalized anxiety disorder, social anxiety, panic disorder, and specific phobias, meta-analyses report moderate to large effects compared with waitlist conditions. Guidelines from the APA and NICE place it among first-line treatments. How it appears: Delivered as individual, group, or internet-supported protocols. In more detail: CBT rests on a simple idea. The way a person thinks about a situation, and what they do in response, both shape how they feel. CBT helps a person notice unhelpful thought patterns and avoidance behaviors, then test and change them. It's usually short, often around 12 to 20 sessions, and it includes practice between sessions. For anxiety, CBT often includes exposure, which means facing feared situations in a planned, gradual way. Example: A person with panic disorder works with a therapist to understand their panic cycle, then practices facing the body sensations they fear, such as a fast heartbeat, until those sensations feel less dangerous. What to watch for: CBT isn't one fixed thing. Protocols differ by condition, and the quality of delivery varies. Effects also look larger against a waitlist than against another active treatment. CBT helps many people, not all, and relapse can still happen. Source: Carpenter JK, et al. Cognitive behavioral therapy for anxiety and related disorders: a meta-analysis of randomized placebo-controlled trials. Depress Anxiety. 2018. PMID: 29451967. (https://pubmed.ncbi.nlm.nih.gov/29451967/) ## Exposure therapy URL: https://anxietyresearch.org/glossary/exposure-therapy Category: Treatment research A behavioral treatment in which patients repeatedly and systematically engage with feared stimuli, situations, or sensations to weaken fear responses. Why it matters: It is the core active ingredient of CBT for specific phobias, panic, and social anxiety. How it appears: In vivo, imaginal, interoceptive, and virtual reality formats are all used. ## Interoceptive exposure URL: https://anxietyresearch.org/glossary/interoceptive-exposure Category: Treatment research Interoceptive exposure is a CBT technique that has a person bring on the body sensations they fear, on purpose and in a safe setting, so those sensations feel less threatening over time. Why it matters: Interoceptive exposure is a core, well-supported part of CBT for panic disorder. It targets the specific fear that drives the panic cycle, the fear of the body's own alarm signals. How it appears: Common exercises include breath holding, hyperventilation, and brief CO2 inhalation in controlled settings. In more detail: People with panic disorder often fear the physical feelings of anxiety itself, such as a racing heart, dizziness, or shortness of breath. They read those feelings as dangerous. Interoceptive exposure has the person trigger the sensations deliberately, for example by breathing fast or spinning, then sit with them. With repetition, the brain learns the sensations are uncomfortable but not dangerous. Example: A therapist asks a person to breathe quickly for a short time to bring on lightheadedness, then guides them to notice that the feeling passes and causes no harm. What to watch for: Interoceptive exposure should be done with guidance, especially at first. It's uncomfortable by design, and that discomfort is the point. It isn't a relaxation exercise. It works by teaching the brain a new lesson about the sensations, not by making them go away. Source: Craske MG, et al. Maximizing exposure therapy: an inhibitory learning approach. Behav Res Ther. 2014. PMID: 24864005. (https://pubmed.ncbi.nlm.nih.gov/24864005/) ## SSRI URL: https://anxietyresearch.org/glossary/ssri Category: Treatment research An SSRI, or selective serotonin reuptake inhibitor, is a class of antidepressant medication. SSRIs are also a first-line medication for most adult anxiety disorders. Why it matters: SSRIs are first-line drug treatment for generalized anxiety disorder, social anxiety disorder, and panic disorder in US and international guidelines. Meta-analyses find small to moderate effects compared with placebo. No single SSRI has been shown to be clearly better than the others for anxiety overall, so the choice often depends on side effects and prior response. How it appears: Used as monotherapy or in combination with CBT. In more detail: SSRIs act on serotonin, a chemical messenger in the brain, by slowing its reabsorption. Common SSRIs include sertraline, escitalopram, fluoxetine, and paroxetine. Despite the name antidepressant, they're well studied and widely used for anxiety. They usually take four to six weeks to show a clear effect, and sometimes longer for a full response. Example: A person with generalized anxiety disorder starts a low dose of an SSRI. The dose is raised gradually, and the clinician reassesses after several weeks, because the medication needs time to work. What to watch for: SSRIs carry an FDA boxed warning about a raised risk of suicidal thoughts in patients under 25, and close monitoring is standard early in treatment. They can cause side effects, including sexual side effects, which are a common reason people stop. Stopping suddenly can cause discontinuation syndrome, so they're lowered with a taper. Source: Bandelow B, et al. Efficacy of treatments for anxiety disorders: a meta-analysis. Int Clin Psychopharmacol. 2015. PMID: 25932596. (https://pubmed.ncbi.nlm.nih.gov/25932596/) ## SNRI URL: https://anxietyresearch.org/glossary/snri Category: Treatment research An SNRI, or serotonin-norepinephrine reuptake inhibitor, is a class of antidepressant medication. SNRIs are also used as a first-line medication for several anxiety disorders. Why it matters: SNRIs are a first-line option for generalized anxiety disorder and are also used for social anxiety disorder and panic disorder. Meta-analyses place their effect size in a similar small-to-moderate range as SSRIs. They're often considered when an SSRI hasn't worked or wasn't tolerated, though either class can be a first choice. How it appears: Often considered when SSRI response or tolerability is limited. In more detail: SNRIs act on two brain messengers, serotonin and norepinephrine. Common SNRIs include venlafaxine and duloxetine. Like SSRIs, they're antidepressants that are well studied for anxiety, and they take several weeks to reach a clear effect. Example: A person whose anxiety didn't improve on an SSRI is switched to an SNRI, which acts on an additional brain messenger. What to watch for: SNRIs carry the same FDA boxed warning about suicidal thoughts in patients under 25 as other antidepressants. Venlafaxine in particular leaves the body quickly, which makes discontinuation syndrome more likely if it's stopped abruptly. SNRIs can also raise blood pressure at higher doses. Source: Bandelow B, et al. Efficacy of treatments for anxiety disorders: a meta-analysis. Int Clin Psychopharmacol. 2015. PMID: 25932596. (https://pubmed.ncbi.nlm.nih.gov/25932596/) ## Benzodiazepine URL: https://anxietyresearch.org/glossary/benzodiazepine Category: Treatment research A benzodiazepine is a class of medication that calms the nervous system quickly. Benzodiazepines reduce acute anxiety fast, but long-term use carries real risks. Why it matters: Guidelines generally recommend benzodiazepines for short-term or occasional use, not as a long-term solo treatment for anxiety. The evidence on short-term relief is clear. The evidence on harm with long-term use is also clear. How it appears: Used short-term or situationally under cautious clinical judgment. In more detail: Benzodiazepines include lorazepam, clonazepam, alprazolam, and diazepam. They work within minutes to an hour, which makes them different from SSRIs and SNRIs, which take weeks. That speed is useful in some situations and a problem in others. With regular use, the body adjusts, so the same dose does less, which is called tolerance, and stopping can cause withdrawal. Example: A clinician might use a benzodiazepine briefly while waiting for an SSRI to take effect, then taper it off once the SSRI is working. What to watch for: Benzodiazepines carry an FDA boxed warning about the danger of combining them with opioids, which can cause severe sedation and slowed breathing. They're listed as potentially inappropriate for older adults in the AGS Beers Criteria, because of fall and confusion risk. Stopping abruptly after regular use can be medically serious, including a risk of seizures, so any taper should be planned with a clinician. Source: U.S. Food and Drug Administration Drug Safety Communications on benzodiazepines, 2016 and 2020. (https://www.fda.gov/) ## Patient-reported outcome URL: https://anxietyresearch.org/glossary/patient-reported-outcome Category: Clinical outcomes A patient-reported outcome is a result measured by what the patient says about their own health, rather than by a lab test or a clinician's rating. Why it matters: Almost all anxiety research depends on patient-reported outcomes. That's appropriate, because the patient's experience is the point. It also means anxiety results can be influenced by expectation, which is why blinding and a control group matter so much in anxiety trials. How it appears: Self-report surveys measuring anxiety severity before and after an intervention. In more detail: Some health outcomes can be measured from the outside, like blood pressure. Anxiety mostly can't. How anxious a person feels is something only they can report. A patient-reported outcome captures that directly, usually through a questionnaire. The GAD-7 is a common patient-reported outcome measure for anxiety. Example: In a trial, each person fills out a GAD-7 questionnaire every few weeks. The change in their scores is a patient-reported outcome and is the main result of the study. What to watch for: Patient-reported outcomes are valuable and also sensitive to expectation. A person who knows they received the active treatment may report more improvement. This isn't dishonesty; it's how expectation works. It is one reason an unblinded anxiety trial deserves more caution. Source: U.S. Food and Drug Administration, patient-reported outcome guidance for industry. (https://www.fda.gov/) ## Screening tool URL: https://anxietyresearch.org/glossary/screening-tool Category: Tools and instruments A screening tool is a short, standardized questionnaire that flags whether a person may have a condition and should be assessed further. It doesn't make a diagnosis. Why it matters: Screening tools shape anxiety statistics. Large surveys that use short screeners tend to report higher numbers than studies that use a full diagnostic interview. Neither is wrong. They measure different things. A screener captures symptoms; a diagnostic interview captures disorders. How it appears: Brief questionnaires administered during routine medical visits to flag potential anxiety disorders. In more detail: Screening tools are designed to be quick and easy to score, so they can be used widely, for example in a primary care visit. For anxiety, the GAD-7 is a common screening tool. A score above a cutoff is a signal to look closer. It isn't the same as a diagnosis, which needs a fuller clinical assessment. Example: A primary care clinic gives every adult patient a GAD-7. Patients who score above the cutoff are scheduled for a longer assessment with a clinician. What to watch for: A screening result is a flag, not a finding. A high score means look into this, not you have this. Screening tools are also built to err toward catching possible cases, so they can produce false positives. The follow-up assessment is what sorts that out. Source: U.S. Preventive Services Task Force, recommendations on screening for anxiety and depression in adults. (https://www.uspreventiveservicestaskforce.org/) ## GAD-7 URL: https://anxietyresearch.org/glossary/gad-7 Category: Tools and instruments The GAD-7 is a short, seven-question form used to screen for and measure generalized anxiety. The letters stand for Generalized Anxiety Disorder 7-item scale. Why it matters: The GAD-7 is one of the most widely used anxiety measures in both research and ordinary clinics. It's brief, free, and validated. Many anxiety trials use the change in GAD-7 score as a main outcome, and measurement-based care uses it to track progress over time. How it appears: Used as a primary outcome measure in numerous clinical trials to track treatment progress. In more detail: The GAD-7 asks how often, over the last two weeks, a person was bothered by symptoms such as feeling nervous, not being able to stop worrying, and being easily annoyed. Each item is scored 0 to 3, for a total of 0 to 21. Higher scores mean more anxiety. Rough bands are often used: 5 for mild, 10 for moderate, and 15 for severe. Example: A clinic has a patient complete the GAD-7 at each visit. A drop from 16 to 8 over several visits shows a clear, trackable improvement. What to watch for: The GAD-7 is a screening and tracking tool, not a diagnosis on its own. A high score is a reason for a fuller assessment, not a verdict. It was designed with generalized anxiety in mind, though it's often used more broadly. Research points to a change of roughly 4 points as the rough mark of a noticeable improvement. Source: Spitzer RL, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006. PMID: 16717171. (https://pubmed.ncbi.nlm.nih.gov/16717171/) ## PHQ-9 URL: https://anxietyresearch.org/glossary/phq-9 Category: Tools and instruments The PHQ-9 is a short, nine-question form used to screen for and measure depression. The letters stand for Patient Health Questionnaire 9-item scale. Why it matters: Anxiety and depression overlap often, so anxiety research and anxiety clinics frequently track depression at the same time. Pairing the PHQ-9 with the GAD-7 gives a fuller picture than either alone. The PHQ-9 also includes a question about thoughts of self-harm, which makes it useful for safety screening. How it appears: Administered to assess comorbidity and track mood changes during anxiety treatment. In more detail: The PHQ-9 asks how often, over the last two weeks, a person was bothered by symptoms of depression, such as low mood, loss of interest, sleep changes, and low energy. Each item is scored 0 to 3, for a total of 0 to 27. It's the depression counterpart to the GAD-7, and the two are often used together. Example: A clinic treating a person for anxiety also has them complete the PHQ-9, because comorbid depression would change the treatment plan. What to watch for: Like the GAD-7, the PHQ-9 is a screening and tracking tool, not a diagnosis. The self-harm question should always be reviewed by a clinician when a person endorses it. A score is a starting point for a conversation, not an endpoint. Source: Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001. PMID: 11556941. (https://pubmed.ncbi.nlm.nih.gov/11556941/) ## DSM-5-TR URL: https://anxietyresearch.org/glossary/dsm-5-tr Category: Tools and instruments The DSM-5-TR is the handbook US clinicians use to name and define mental health conditions. The letters stand for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision. Why it matters: Most US anxiety research uses DSM criteria to decide who counts as having a disorder. A shared definition lets studies be compared. It also means that when definitions change between editions, prevalence numbers can shift for reasons that have nothing to do with how common a condition really is. How it appears: Defines the exact diagnostic criteria study participants must meet for inclusion in a clinical trial. In more detail: The DSM-5-TR, published by the American Psychiatric Association, sets out the criteria for each diagnosis, including the anxiety disorders. The "TR" means text revision, an update released in 2022 that refreshed the text and some details of the 2013 fifth edition. When a clinician diagnoses generalized anxiety disorder or panic disorder, they're matching a person's experience against DSM-5-TR criteria. Example: To diagnose generalized anxiety disorder, the DSM-5-TR requires excessive worry on most days for at least six months, along with specific physical and cognitive symptoms. What to watch for: The DSM is a clinical tool, not a self-test. Matching a few criteria from a list isn't the same as a diagnosis. Diagnosis also depends on distress and on how much daily life is affected. The wider world uses a different system, the ICD-11. Source: American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision. (https://www.psychiatry.org/psychiatrists/practice/dsm) ## ICD-11 URL: https://anxietyresearch.org/glossary/icd-11 Category: Tools and instruments The ICD-11 is the global system for classifying diseases and health conditions, including mental health conditions. ICD stands for International Classification of Diseases, and 11 is the current edition. Why it matters: Research and health data from outside the United States often use ICD definitions. When comparing anxiety statistics across countries, part of any difference can come from the two systems defining a disorder slightly differently, rather than from a true difference in how common it is. How it appears: Used in international public health data to track global anxiety prevalence. In more detail: The ICD is published by the World Health Organization and is used worldwide for health records, statistics, and billing. The 11th edition came into effect in 2022. It includes definitions for anxiety and related disorders. The ICD and the DSM cover much of the same ground and broadly agree, but they're organized differently and don't match on every detail. Example: A global health report on anxiety prevalence would typically use ICD-11 categories, while a US study would more often use DSM-5-TR. What to watch for: When comparing numbers across countries or sources, check which system was used. ICD-11 and DSM-5-TR are close, not identical. Small differences in criteria can produce real differences in counts. Source: World Health Organization, International Classification of Diseases, 11th Revision. (https://icd.who.int/)