Antidepressant use among American adults in a 50-state survey

Abstract

Background

Antidepressants are among the most prescribed medications in the USA, yet challenges in access to mental health treatment persist.

Objective

To assess current and lifetime antidepressant and psychotherapy use among American adults, and examine attitudes towards potential federal restrictions on antidepressant prescribing.

Methods

We conducted a cross-sectional survey study using data from a national non-probability internet-based panel weighted to approximate national demographics (age, gender, race and ethnicity, education, US census region, and urbanicity) based on 2020 US Census data. Data were collected between 10 April and 27 May 2025 from 30 810 adults residing in the USA. The primary outcomes were self-reported current and past antidepressant and psychotherapy use, and support for or opposition to potential federal restrictions on antidepressant prescribing. Logistic regression models estimated demographic and treatment-related features associated with these outcomes.

Findings

Among 30 115 respondents with complete antidepressant data, 16.6% reported current antidepressant use, and of 30 098 respondents with psychotherapy data, 10.4% reported current psychotherapy. Use of both treatments was significantly greater among White respondents compared with all other racial groups. When asked about potential federal restrictions on doctors prescribing antidepressants, 16.4% of respondents supported and 48.0% opposed such regulation, with lesser opposition among those of male gender (OR 0.69, 95% CI 0.65 to 0.73), and greater opposition among those with lifetime antidepressant treatment (OR 2.37, 95% CI 2.21 to 2.54).

Conclusions

Antidepressant and psychotherapy use remains unevenly distributed across demographic groups. A significant proportion of adults in every US state oppose efforts to restrict access to antidepressant prescribing, reflecting broad public support for maintaining access to treatment.

Clinical implications

Findings from this study suggest that restrictive policies on antidepressant prescribing are unlikely to align with public sentiment and may risk exacerbating existing inequities in care.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Antidepressants are safe and effective treatments for major depression and anxiety, yet prior studies show persistent disparities in their use across racial, ethnic and gender groups. Recent public and political discourse has questioned antidepressant safety and raised the prospect of federal prescribing restrictions.

WHAT THIS STUDY ADDS

• In a national survey of 30 810 adults across all 50 US states, current antidepressant use was reported by one in six adults and psychotherapy by one in 10, with marked disparities by race, ethnicity and gender. Three times as many respondents opposed versus supported government restrictions on antidepressant prescribing.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Findings highlight the need to address factors contributing to uneven treatment use and underscore that new restrictions on antidepressant use do not align with public sentiment.

Background

There is abundant and compelling evidence that antidepressants are effective for the treatment of moderate or severe major depression, as well as multiple anxiety disorders.¹ They are among the most widely prescribed medications in the USA, with 13.2% of adults reporting antidepressant use in the past 30 days during 2015–2018.² Recent analysis of commercial prescribing data demonstrates that antidepressants remain among the most frequently prescribed therapeutic medication classes in the USA.³

However, some studies suggest marked disparities in antidepressant prescribing, including lower prescription rates and delayed treatment initiation among some groups. Of adults with past-year depressive disorder, 63.7% of individuals identifying as Latino, 68.7% of those identifying as Asian and 58.8% of those identifying as African American did not access mental health treatment in the last year, compared with 40.2% of non-Latino White adults.⁴ In particular, the prevalence of antidepressant use in the past 30 days was highest among non-Hispanic White adults (16.6%) compared with non-Hispanic Black (7.8%), Hispanic (6.5%) and non-Hispanic Asian (2.8%) adults.² Non-white individuals also experience longer delays on average from first major depressive episode to first antidepressant medication treatment.⁵

Data about antidepressant utilisation are often drawn from large-scale claims or electronic health records data, but these estimates may be less reliable in the context of mental health, as many patients opt not to use their insurance coverage, access care outside of their primary care network⁶ and fill prescriptions at non-affiliated pharmacies.⁷ Prior work using such data has indicated an increase in antidepressant prescriptions associated with telehealth visits,⁸ particularly during the COVID-19 pandemic,⁹ consistent with self-reports.¹⁰ To understand differences in antidepressant use in the USA, we drew on a large 50-state internet survey that collected a representative state-by-state sample of US adults. We examined differences in current and past antidepressant use, and for comparison evaluated current and past psychotherapy use, focusing on variability in use between sociodemographic groups. In light of recent misinformation about antidepressants spread by some US government leaders,¹¹ we also examined attitudes towards federal restrictions of antidepressant use, and whether such attitudes vary by state, sociodemographic characteristics or political affiliation.

Objective

The objective of this study was to characterise patterns of current and lifetime antidepressant and psychotherapy use among US adults and to identify sociodemographic predictors of treatment utilisation. Using data from a 50-state internet-based survey weighted to approximate national demographics, we also examined public attitudes towards proposed federal restrictions on antidepressant prescribing.

Methods

Study design

We used data from a web-based Qualtrics survey fielded on behalf of an academic consortium by a commercial survey panel aggregator, Pure Spectrum, with waves approximately every 12 weeks since spring 2020. This platform allows individuals to opt in to survey participation in return for incentives that vary based on panel and participant choice, but could include cash payments, airline miles, gift cards, points redeemable for mobile game play or entrance into a sweepstakes. The study employed a non-probability design¹² with weights for age, gender, race and ethnicity, education, US census region, and urbanicity. in each of the 50 US states and the District of Columbia, in order to yield samples representative of adults at the state and national levels. Participants were eligible if they were 18 and older and resided in the USA. As the survey addressed multiple topics, it was described as a general survey of opinions in lieu of beliefs about a particular subject. The present analysis used data from survey wave 35, which included questions related to antidepressant and psychotherapy use. For details of the Civic Health and Institutions Project survey, see https://www.chip50.org.

Measures

We assessed current and prior antidepressant use by asking, ‘Have you ever used a prescription antidepressant medication like Prozac, Zoloft, or Lexapro?’ Possible responses were ‘Yes, currently taking’, ‘Yes, but not currently’, ‘No, but considering it’, ‘No, not considering it’ or ‘Prefer not to answer’. Likewise, we assessed use of psychotherapy by asking, ‘Have you ever received psychotherapy (talk therapy)?’ To explore attitudes towards antidepressant medication more generally, we also asked, ‘Do you approve or disapprove of federal policies that would make it harder for doctors to prescribe antidepressants?’, with a 5-point response scale from ‘Strongly approve’ to ‘Strongly disapprove’. To facilitate interpretation of this question in a political context, we asked respondents whether they identified most as Republican, Democrat, independent or other. Depressive symptoms were measured with the 9-item Patient Health Questionnaire (PHQ-9), which assesses Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition depression criteria over the past 2 weeks on a 0–3 scale. Scores ≥10 indicated moderate to severe depressive symptoms.¹³

To enable analyses of differential treatment use and attitudes by population subgroup, and to confirm representativeness of the US population and facilitate survey weighting, we also collected sociodemographic data. We asked participants to self-identify race from a list including African American or Black, Asian American, Native American, Pacific Islander, White or other; in the last category, they could also provide a free text self-description. To facilitate inclusion of smaller groups, we collapsed Native American, Pacific Islander and another race into a single category for analysis. We also asked individuals whether or not they identified as Hispanic. We asked participants to self-report their gender identity by selecting from the options: male, female, genderqueer or other. For the purposes of analysis, responses of ‘genderqueer’ and ‘other’ were combined into a single category labelled ‘transgender and nonbinary (TGNB)’.

Statistical analysis

We initially examined rates of current and past antidepressant use and, in parallel, psychotherapy use by sociodemographic group in a univariate analysis. We then fit a series of multiple logistic regression models to estimate the association between individual sociodemographic features and antidepressant use, as well as psychotherapy use. In all regression models, covariates included age category (to allow for non-linear effects), gender, education (categorised as graduate, undergraduate, some college, high school graduate, some high school or less), annual household income (categorised as <$25 000, $25 000 to <$50 000, $50 000 to <$100 000, >$100 000), race and ethnicity, and rural, suburban or urban setting.

We also examined attitudes towards regulation of antidepressant prescribing, fitting a series of logistic regression models. The attitudes were dichotomised to ‘Strongly disapprove’ and ‘disapprove’ versus other categories, as a means of understanding rates of disapproval. (Secondarily, we examined individual response categories using ordinal logistic regression, after confirming proportionality of risks.) First, we examined support for restrictions on antidepressants based on sociodemographic features alone. Next, we added antidepressant and psychotherapy use to these models to determine whether current or lifetime use influenced these attitudes. Finally, we added political party affiliation to understand the extent to which political beliefs associated with support for restrictions.

As recommended for non-probability samples,¹⁴ we applied survey weights to estimate national distributions via the R survey package (version 4.2-1).¹⁵ Interlocking poststratification national weights for age at survey completion, gender, race and ethnicity, as well as education and region, drew on 2020 US Census American Community Survey data.¹⁶ All analyses used R 4.3.2,¹⁷ with two-sided uncorrected p<0.05 as the threshold for statistical significance. Because rates of missing sociodemographic data were extremely low (eg, missing race or ethnicity data for seven individuals), we applied listwise deletion rather than imputation.

Findings

The survey included 30 810 unique respondents who participated between 10 April and 27 May 2025. Analyses of antidepressant utilisation were conducted on a final sample of 30 115 individuals after excluding 664 participants who selected ‘prefer not to answer’ and 31 who did not respond to the question on current antidepressant use. In all, 30 098 respondents answered the question about psychotherapy use. Of the survey respondents who answered treatment utilisation questions, mean age was 48.2 years (SD 17.1); 15 815 (52.5%) identified as female, 14 096 (46.8%) as male and 198 (0.7%) as TGNB. A total of 1305 (4.3%) identified as Asian American, 4821 (16.0%) as Black, 3607 (12.0%) as Hispanic, 1976 (6.6%) as another race or ethnicity, and 22 013 (73.1%) as White. Additional characteristics of the cohort, stratified by current antidepressant use, are summarised in table 1. Cohort characteristics stratified by current psychotherapy use are summarised in online supplemental table 1.

Table 1. Cohort characteristics stratified by current antidepressant use.

	No (n=25 113)	Yes (n=5002)	Total (n=30 115)	P value
Age category				<0.001
18–24	2028 (8.1%)	401 (8.0%)	2429 (8.1%)
25–34	4465 (17.8%)	799 (16.0%)	5264 (17.5%)
35–44	5115 (20.4%)	1016 (20.3%)	6131 (20.4%)
45–54	4119 (16.4%)	1027 (20.5%)	5146 (17.1%)
55–64	3749 (14.9%)	899 (18.0%)	4648 (15.4%)
65 and over	5637 (22.4%)	860 (17.2%)	6497 (21.6%)
Gender				<0.001
n missing	6	0	6
Female	12 383 (49.3%)	3432 (68.6%)	15 815 (52.5%)
Male	12 603 (50.2%)	1493 (29.8%)	14 096 (46.8%)
Non-binary	121 (0.5%)	77 (1.5%)	198 (0.7%)
Race				<0.001
Asian American	1189 (4.7%)	116 (2.3%)	1305 (4.3%)
Black	4425 (17.6%)	396 (7.9%)	4821 (16.0%)
Other	1717 (6.8%)	259 (5.2%)	1976 (6.6%)
White	17 782 (70.8%)	4231 (84.6%)	22 013 (73.1%)
Ethnicity				<0.001
n missing	1	0	1
Hispanic	3137 (12.5%)	470 (9.4%)	3607 (12.0%)
Non-Hispanic	21 975 (87.5%)	4532 (90.6%)	26 507 (88.0%)
Education				<0.001
Some high school or less	888 (3.5%)	182 (3.6%)	1070 (3.6%)
High school graduate	6946 (27.7%)	1213 (24.3%)	8159 (27.1%)
Some college	6038 (24.0%)	1335 (26.7%)	7373 (24.5%)
College degree	8328 (33.2%)	1702 (34.0%)	10 030 (33.3%)
Graduate degree	2913 (11.6%)	570 (11.4%)	3483 (11.6%)
Income				<0.001
n missing	3	0	3
<25 000	5572 (22.2%)	1263 (25.2%)	6835 (22.7%)
25 000 to <50 000	6387 (25.4%)	1371 (27.4%)	7758 (25.8%)
50 000 to <100 000	8029 (32.0%)	1492 (29.8%)	9521 (31.6%)
100 000 and over	5122 (20.4%)	876 (17.5%)	5998 (19.9%)
Urban/rural				<0.001
Rural	4840 (19.3%)	1145 (22.9%)	5985 (19.9%)
Suburban	13 536 (53.9%)	2834 (56.7%)	16 370 (54.4%)
Urban	6737 (26.8%)	1023 (20.5%)	7760 (25.8%)

In the reweighted sample, 16.6% of individuals reported current antidepressant treatment, and an additional 17.3% reported past (but not current) treatment. For psychotherapy, 10.4% of participants reported current use, and an additional 24.2% reported past use. Figure 1 illustrates rates of current antidepressant use (top panel) and current psychotherapy use (bottom panel) by state; any lifetime (ie, current or past) use is illustrated in online supplemental figure 1.

Figure 1. Current antidepressant use (A) and current psychotherapy use (B) by US state.

We first examined differential rates of current antidepressant use by population subgroup (online supplemental table 2). The greatest rates of current antidepressant use were observed among those aged 45–54 years (20.0%), individuals identifying as White (19.2%) and TGNB individuals (38.9%). Lowest prevalence of current antidepressant use was evident in Asian American (8.9%) and Black (8.2%) individuals, as well as those identifying as male (10.6%). In logistic regression models, sociodemographic features significantly associated with higher odds of current use included being White (OR 2.66, 95% CI 2.12 to 3.34) and non-Hispanic ethnicity (OR 1.43, 95% CI 1.25 to 1.64). Compared with female respondents, those who identified as male were less likely to be receiving antidepressants (OR 0.45, 95% CI 0.41 to 0.48), while those identifying as TGNB were more likely to be receiving antidepressants (OR 2.60, 95% CI 1.82 to 3.71) (figure 2). In secondary analyses incorporating self-reported Medicaid (9271; 30.8%) and Medicare (10 363; 34.4%) receipt, associations were of similar magnitude (online supplemental figure 2). Features associated with lifetime use followed a similar pattern (online supplemental figures 3 and 4).

Figure 2. Features associated with current antidepressant use.

For comparison, we repeated these analyses examining current and lifetime use of psychotherapy (online supplemental figures 5 and 6). Compared with antidepressant use, current psychotherapy use varied less across subgroups (online supplemental table 2). However, features associated with differential odds of current psychotherapy use were similar to those associated with current antidepressant use. Compared with those identifying as female, psychotherapy use was lower among male individuals (OR 0.66, 95% CI 0.60 to 0.72), and greater among those identifying as TGNB (OR 2.49, 95% CI 1.74 to 3.57). Current psychotherapy use was also more likely among those identifying as White race (OR 1.63, 95% CI 1.28 to 2.07). Associations for lifetime psychotherapy use were similar, as were those incorporating Medicare and Medicaid status (online supplemental figures 7 and 8).

When analysis was restricted to individuals who indicated moderate to severe current depressive symptom severity (score ≥10), treatment use likewise varied substantially by demographic subgroup (online supplemental table 2). Rates were highest among White participants (28.5% currently on antidepressants, 16.8% currently in therapy) and lowest among Black (13.8%, 14.7%) and Asian American respondents (14.6%, 14.6%). In this subset of respondents, individuals identifying as female more often reported treatment compared with those who identified as male (29.5% vs 18.2% for current antidepressant use; 17.6% vs 14.2% for current therapy).

We also investigated attitudes towards federal restriction of antidepressant use. Nationally, applying survey weights, 16.4% supported such restrictions; 23.8% strongly disapprove, 24.2% disapprove, 35.6% neither approve nor disapprove, 10.9% approve and 5.5% strongly approve. Online supplemental figure 9 and online supplemental table 3 illustrate state-by-state rates of opposition. In logistic regression models incorporating sociodemographic features and treatment history, opposition was lower among those identifying as male gender compared with female (OR 0.69, 95% CI 0.65 to 0.73), and greater among those who had received lifetime antidepressant treatment (OR 2.37, 95% CI 2.21 to 2.54) or lifetime psychotherapy (OR 1.44, 95% CI 1.34 to 1.54) (figure 3). Opposition was also associated with political party affiliation: individuals who described themselves as Republican (OR 0.35, 95% CI 0.33 to 0.38) or independent (OR 0.56, 95% CI 0.52 to 0.60) were less likely than Democrats to oppose such restrictions. In secondary analyses applying ordinal logistic regression to evaluate individual categories, effects were qualitatively similar (online supplemental figure 10).

Figure 3. Binary disapproval of restrictions (logistic regression).

Discussion

In this nationally representative sample of 30 810 US adults, current mental health treatments were common: 16.6% of respondents were currently taking an antidepressant medication and 10.4% were receiving psychotherapy, while more than one in three reported at least some lifetime exposure to either treatment. Our antidepressant prevalence estimate somewhat exceeds the 13.2% observed in the 2015–2018 National Health and Nutrition Examination Survey² and the 11.4% observed in 2023 data from the National Health Interview Survey (NHIS).¹⁸ The latter shows rates of any mental health treatment rising from 9.8% in 2019 to 11.4% in 2023, with the steepest gains among individuals identifying as female, so it is possible our results simply reflect continued increase through and after the pandemic. Increases were steepest among those identifying as female, paralleling our finding that female-identifying individuals are more likely than males to receive treatment.

We also identified substantial variation in treatment use by sociodemographic groups. Rates were highest among individuals identifying as female or non-binary as well as mid-life adults and White or non-Hispanic respondents. Conversely, rates of current antidepressant use were the lowest among Asian American and Black individuals as well as those identifying as male. This variation was generally similar when analyses were restricted to individuals with moderate to severe depressive symptoms. Among individuals with PHQ-9 scores ≥10, indicating moderate to severe depressive symptoms, treatment gaps persisted across subgroups, with only 13–15% of Black and Asian American respondents currently on antidepressants, compared with 27.3% of White respondents.

Our findings highlight substantial variability in receipt of mental health treatment which is not explained by differences in prevalence of depressive disorder diagnoses.¹⁹ Race and ethnicity utilisation patterns are consistent with earlier NHIS briefs that found greatest use of medication for mental health in White adults and lowest in Black and Asian adults.¹⁸ They are also aligned with evidence that non-Hispanic Black and Hispanic individuals experience longer delays from first major depressive episode to first antidepressant medication treatment compared with non-Hispanic White participants.⁵ Among Medicaid beneficiaries with depression, Black participants are half as likely as hite participants to receive any depression treatment, and Hispanic participants are a third as likely.²⁰

These gaps could reflect biases in prescribing patterns, inequitable insurance design and other challenges in accessing or trusting healthcare systems.21,23 Lower medication uptake among minority groups could also reflect variation in community beliefs about mental illness, preferences for non-pharmacological interventions and concerns about confidentiality.^{24 25} In support of a contribution of differential access to care, Centers for Disease Control and Prevention data indicate that unmet mental health needs because of cost remain higher among women than men and among uninsured compared with privately insured adults.²⁶

For gender-minority groups, population-level data are scarce. Our survey shows substantially greater rates of treatment utilisation in TGNB respondents, echoing the Trevor Project’s 2024 national poll in which 50% of LGBTQ+ youth reported trying to access mental healthcare.²⁷ Taken together, these findings suggest that access, rather than willingness to seek care, may remain a barrier for TGNB populations as well.

We also examined attitudes towards antidepressant regulation. Three times as many participants were opposed compared with supportive of federal restrictions on antidepressant access, and opposition was strongest among those with personal treatment experience. Some US health leaders have made repeated false claims about the risks associated with antidepressant use,¹¹ raising concern that these medications could become more highly regulated in spite of evidence of safety and efficacy that is similar to, or greater than, many other Food and Drug Administration (FDA)-approved medications used for chronic illness. In this survey, 48.0% of respondents opposed federal limits on antidepressant prescribing—a proportion comparable to public reactions when Medicare Part D restrictions were proposed and later withdrawn in 2014 in response to public outcry.²⁸ Personal experience with antidepressants was associated with a 137% increase in odds of opposing restrictions. Political ideology was also associated with differential views, in line with prior evidence that conservative adults engage less with mental health services and report better self-rated mental health compared with liberals.²⁹ In light of evidence of adverse outcomes following past regulatory actions targeting antidepressants (eg, the 2004–2005 FDA black box warning and subsequent dip in adolescent treatment coupled with rising suicides³⁰), public attitudes may reflect an awareness of the importance of mental health treatment and the consequences of denying access to care.

Limitations

This study has multiple limitations, including reliance on self-report, lack of diagnostic verification and cross-sectional design. Although any internet-based, non-probability panel could undersample digitally disconnected adults, recent Pew data indicate that few older, low-income or low-education adults lack internet access.³¹ Prior studies comparing results from this survey design to those of administrative data or probability samples further suggest the validity of this sampling approach in this context.^{32 33} Still, we recognise that lower levels of trust in research and healthcare institutions among certain communities may influence both survey participation in general and responses related to treatment use in particular.³⁴ Thus, while we apply quotas and reweighting to maximise representativeness, differential trust or other factors may still skew results for some subgroups. We also do not distinguish where individuals seek or receive treatment, such that any challenges in accessing care could reflect lack of access to mental health specialty care, primary care or both of these. Finally, as with any hypothetical query, the wording of our question about antidepressant regulation likely impacted responses. Alternate framings (eg, incorporating language regarding safety, despite abundant evidence of such safety) might alter support for such regulation.

Conclusions

Antidepressant and psychotherapy use is common among US adults but remains unevenly distributed across demographic groups, both in terms of states and individual subpopulations defined by gender, race and ethnicity. Despite recent criticisms of pharmacological mental health treatment, substantially more adults oppose than support federal restrictions on antidepressants, with opposition especially likely among those with personal treatment experience. Taken together, these findings underscore persistent inequities in mental health treatment and highlight the importance of understanding both access and attitudes when addressing gaps in care.

Clinical implications

Efforts to improve mental health outcomes could address equitable access to treatment and barriers that limit care for under-represented populations. Restrictive prescribing policies are unlikely to reflect public sentiment and may further exacerbate existing gaps. As policymakers develop strategies to address such disparities, it will be important to consider public sentiment, with more adults opposing than supporting greater regulation of depression treatment in all 50 US states.

The sponsors did not have any role in the design and conduct of the study; collection, management, analysis and interpretation of data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication.

Data availability statement

Data may be obtained from a third party and are not publicly available.