Key Points
Question
Is medical debt associated with delayed or forgone mental health care?
Findings
This cross-sectional study of 27 651 US adults estimated that between 1 in 4 and 1 in 5 US adults with depression and anxiety carry medical debt. Medical debt was associated with more than a 2-fold increase in delayed or forgone treatment for mental disorders.
Meaning
Medical debt is prevalent among US adults with common mental disorders, and this may contribute to the mental health treatment gap.
Abstract
Importance
Medical debt is common in the US and may hinder timely access to care for mental disorders.
Objective
To estimate the prevalence of medical debt among US adults with depression and anxiety and its association with delayed and forgone mental health care.
Design, Setting, and Participants
Cross-sectional, nationally representative survey study of US adult participants in the 2022 National Health Interview Survey who had current or lifetime diagnoses of depression or anxiety.
Exposures
Self-reported lifetime clinical diagnoses of depression and anxiety; moderate to severe symptoms of current depression (Patient Health Questionnaire–8 score ≥10) and anxiety (Generalized Anxiety Disorder–7 score ≥10) irrespective of lifetime diagnoses; and past-year medical debt.
Main Outcomes and Measures
Self-reported delaying and forgoing mental health care because of cost in the past year.
Results
Among 27 651 adults (15 050 [54.4%] female; mean [SD] age, 52.9 [18.4] years), 5186 (18.2%) reported lifetime depression, 1948 (7.3%) reported current depression, 4834 (17.7%) reported lifetime anxiety, and 1689 (6.6%) reported current anxiety. Medical debt was more common among adults with lifetime depression (19.9% vs 8.6%; adjusted prevalence ratio [aPR], 1.97; 95% CI, 1.96-1.98), lifetime anxiety (19.4% vs 8.8%; aPR, 1.91; 95% CI, 1.91-1.92), current depression (27.3% vs 9.4%; aPR, 2.34; 95% CI, 2.34-2.36), and current anxiety (26.2% vs 9.6%; aPR, 2.24; 95% CI, 2.24-2.26) compared with adults without the respective mental disorders. Medical debt was associated with delayed health care among adults with lifetime depression (29.0% vs 11.6%; aPR, 2.68; 95% CI, 2.62-2.74), lifetime anxiety (28.0% vs 11.5%; aPR, 2.45; 95% CI, 2.40-2.50), current depression (36.9% vs 17.4%; aPR, 2.25; 95% CI, 2.13-2.38), and current anxiety (38.4% vs 16.9%; aPR, 2.48; 95% CI, 2.35-2.66) compared with those without these diagnoses. Medical debt was associated with forgone health care among adults with lifetime depression (29.4% vs 10.6%; aPR, 2.66; 95% CI, 2.61-2.71), lifetime anxiety (28.2% vs 10.7%; aPR, 2.63; 95% CI, 2.57-2.68), current depression (38.0% vs 17.2%; aPR, 2.35; 95% CI, 2.23-2.48), and current anxiety (40.8% vs 17.1%; aPR, 2.57; 95% CI, 2.43-2.75) compared with those without the diagnoses.
Conclusions and Relevance
Medical debt is prevalent among adults with depression and anxiety and may contribute to the mental health treatment gap. In the absence of structural reform, new policies are warranted to protect against this financial barrier to mental health care.
This cross-sectional study of data from the 2022 National Health Interview Survey estimates the prevalence of medical debt among US adults with depression and anxiety and its association with delayed and forgone mental health care.
Introduction
Despite the introduction of new legislation and policies in the past decade, including extending both private and public insurance through the Affordable Care Act (ACA), many financial barriers to health care persist.1,2,3 Among low-income adults, financial barriers remain the most common barrier to receiving mental health care in the US.2 With increasing health care costs, medical debt is a persistent financial barrier to health care that arises from high out-of-pocket costs incurred by both uninsured and underinsured individuals,4 causing patients to use credit cards, loans, or mortgages to pay medical bills.5 When bills are unpaid, the outstanding amount is sent to debt collections,6 with personal debt limiting access to credit and hindering subsequent use of health care services.4,5,6,7 Recent studies using consumer credit reports have found that medical debt is the principal contributor to personal debt,4,6 raising concerns about the impact of medical debt on accessing treatment. Evidence shows that the ACA and related Medicaid expansion may protect against medical debt, but protective effects are concentrated among those with coverage,5,8,9 raising concerns for individuals falling in the coverage gap (ie, those whose income is >138% of the federal poverty level [FPL] for those in Medicaid expansion states but <200% of the FPL to qualify for private insurance subsidies).
Medical debt in the US has soared in recent decades, coinciding with hospitals’ gravitation away from charity care and toward aggressive debt collection tactics, including wage garnishment, bank account seizures, property liens, and lawsuits.10,11,12,13 Past literature suggests that inability to pay medical debt may cause patients to delay or forgo needed care.7,14,15,16,17 This is of particular consequence in psychiatry, where a persistent treatment gap remains; fewer than half of all US adults with mental disorders receive treatment.18 In this study, we estimate the prevalence of medical debt among US adults with depression or anxiety and evaluate its association with delayed and forgone mental health care.
Methods
Participants
This study analyzes 2022 data from the National Health Interview Survey (NHIS), a cross-sectional, nationally representative survey of the civilian, noninstitutionalized US adult population.19 The NHIS uses stratified cluster sampling to select households for interviews, from which 1 adult is randomly selected. Asian, Black, and Hispanic households are oversampled in the NHIS to improve precision of estimates for these groups.20 In 2022, 54.4% of the households completed the household roster interview. The remaining households did not complete the interview for a variety of reasons, including language barriers, no answer after repeated contact attempts, refusal, insufficient data collected during interview, and other reasons. Of eligible adults selected and approached after household interview, 87.6% completed the interview, resulting in a final response rate of 47.7%. A total of 27 651 adults were thus interviewed and provided oral informed consent.
For most measures in the NHIS, unknown or missing values are rare. The percentage of missing responses is less than 5% for all variables, with the exception of family income, requiring use of multiple imputation. These methods have been described elsewhere.21 Complete case analyses were conducted, meaning that participants with missing data for both mental health status (lifetime depression, lifetime anxiety, current depression, and current anxiety) and medical debt were excluded (n = 136), yielding an analytic sample of 27 515 adults.
Analyses were limited to publicly available, deidentified data and did not require institutional board review approval. Reporting adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for observational studies.22
Measures
Mental health status was assessed for depression and anxiety, each of which was measured in 2 ways: (1) lifetime diagnosis and (2) current status. Lifetime depression was measured with the survey item, “Have you ever been told by a doctor or other health professional that you had any type of depression, including major depression (or major depressive disorder), bipolar depression, dysthymia, postpartum depression, and seasonal affective disorder?” Similarly, lifetime anxiety was measured with the survey item, “Have you ever been told by a doctor or other health professional that you had any type of anxiety disorder, including generalized anxiety disorder, social anxiety disorder, panic disorder, posttraumatic stress disorder, obsessive-compulsive disorder, and phobias?”
Current depression and anxiety were ascertained using the Patient Health Questionnaire–8 (PHQ-8) and Generalized Anxiety Disorder–7 (GAD-7) scales, respectively. A cutpoint score of 10 or greater was used on both scales, corresponding to moderate or severe symptoms. Past studies have identified a score of 10 as a cutpoint that optimizes sensitivity and specificity against the gold standard of a research diagnosis of major depressive disorder or generalized anxiety disorder and thus can be used to define current depression and anxiety in population-based studies.23,24
In accordance with prior studies using the NHIS,14,25,26 medical debt was defined as an endorsement of the survey item, “In the past 12 months, did you have problems paying or were unable to pay medical bills?” Delayed mental health care was defined an endorsement of the survey item, “During the past 12 months, have you delayed getting counseling or therapy from a mental health professional because of the cost?” Similarly, forgone mental health care was defined as an endorsement of the survey item, “During the past 12 months, was there any time when you needed counseling or therapy from a mental health professional, but did not get it because of the cost?”
Sociodemographic characteristics, including age, sex, race (American Indian or Alaska Native, Asian, Black or African American, White, and other single and multiple races), ethnicity (Hispanic or Mexican vs not Hispanic), employment status, income, marital status, child(ren) in the home, educational attainment, employment status, health insurance status, geographic region, prescribed medication for anxiety, prescribed medication for depression, and physical comorbidities (from a list of chronic conditions, including asthma, cancer, and diabetes), were self-reported. Income is reported in the NHIS as a ratio of household income to the FPL, scaled by household size.
Statistical Analysis
We first computed descriptive statistics of sociodemographic characteristics across the 4 subgroups: lifetime depression, lifetime anxiety, current depression, and current anxiety. We then fit weighted multivariable Poisson regression models with robust variance to the full analytic sample to assess whether adults with depression and anxiety had higher prevalence of medical debt compared with adults without these mental disorders. Four models were constructed, considering the associations of (1) lifetime depression compared with individuals without a depression diagnosis, (2) lifetime anxiety compared with individuals without an anxiety diagnosis, (3) current depression compared with individuals scoring less than 10 on the PHQ-8, irrespective of lifetime diagnosis, and (4) current anxiety compared with individuals scoring less than 10 on the GAD-7, irrespective of lifetime diagnosis. We adjusted for sociodemographic characteristics (age, sex, race, ethnicity, marital status, children in the home, employment status, income, geographic region of residence, and educational attainment), number of physical comorbidities, and health insurance status in both models.
We next developed Poisson regression models with robust variance to evaluate the association between medical debt and 2 outcomes of interest: delaying mental health care and forgoing mental health care. We stratified each model by mental health status (lifetime depression, lifetime anxiety, current depression, current anxiety) and adjusted for sociodemographic characteristics (age, sex, race, ethnicity, marital status, children in the home, employment status, income, geographic region of residence, and educational attainment), number of physical comorbidities, and health insurance status. Finally, we tested for separate interactions between medical debt and (1) insurance status and (2) income (dichotomized as below median income) to determine whether these variables modify the association between medical debt and forgone mental health care. When interaction terms were statistically significant, we conducted stratified analyses that assessed associations by status of the given moderator.
All statistical tests were 2-sided, with significance defined as P < .05. We performed analyses using the survey package in R version 4.3.1 statistical software (R Foundation for Statistical Computing). Sampling weights and the complex survey design elements of the NHIS were included to account for sampling probabilities and nonresponse. Weighted percentages with corresponding 95% CIs were computed using Taylor series linearization. Weighted percentages reported may not correspond to the raw percentages.
Sensitivity Analysis
As a sensitivity analysis, we raised the PHQ-8 and GAD-7 cutpoint to 15 or higher for current depression and anxiety, respectively. This limited the sample to those participants whose symptoms would be characterized as severe.
Results
Characteristics of Participants With Depression and Anxiety
Among 27 651 adults (15 050 [54.4%] female; mean [SD] age, 52.9 [18.4] years), 5186 (18.2%) reported lifetime depression, 1948 (7.3%) reported current depression, 4834 (17.7%) reported lifetime anxiety, and 1689 (6.6%) reported current anxiety. Across the 4 groups, the majority of participants were female (61.9%-65.3%), were non-Hispanic (83.8%-88.1%), were White (73.9%-80.3%), were 18 to 54 years of age (63.9%-73.4%), were insured (90.5%-93.4%), had at least 1 physical comorbidity (56.0%-64.5%), had household income levels less than 300% of the FPL (51.1%-55.7%), and were employed (46.1%-54.3%). Table 1 presents characteristics of the participants with lifetime depression, lifetime anxiety, current depression, and current anxiety.
Table 1. Sociodemographic and Clinical Characteristics of US Adults With Depression and Anxiety, 2022.
| Characteristic | Depression | Anxiety | ||||||
|---|---|---|---|---|---|---|---|---|
| Current (n = 1948) | Lifetime (n = 5186) | Current (n = 1689) | Lifetime (n = 4834) | |||||
| No. | Weighted % (95% CI) | No. | Weighted % (95% CI) | No. | Weighted % (95% CI) | No. | Weighted % (95% CI) | |
| Age, y | ||||||||
| 18-24 | 195 | 18.0 (15.7-20.0) | 378 | 13.5 (12.1-15.0) | 195 | 19.5 (16.9-22.0) | 404 | 14.7 (13.3-16.0) |
| 25-34 | 306 | 17.5 (15.6-20.0) | 821 | 18.5 (17.2-20.0) | 315 | 21.0 (18.6-24.0) | 897 | 21.0 (19.7-22.0) |
| 35-44 | 280 | 14.7 (13.0-17.0) | 813 | 16.5 (15.4-18.0) | 313 | 18.5 (16.5-21.0) | 909 | 19.4 (18.1-21.0) |
| 45-54 | 299 | 15.5 (13.7-18.0) | 751 | 15.4 (14.3-17.0) | 260 | 14.4 (12.6-16.0) | 681 | 14.8 (13.7-16.0) |
| 55-64 | 388 | 16.8 (15.0-19.0) | 966 | 16.7 (15.5-18.0) | 313 | 15.0 (13.2-17.0) | 812 | 14.4 (13.3-16.0) |
| ≥65 | 480 | 17.5 (15.7-19.0) | 1456 | 19.4 (18.2-21.0) | 293 | 11.6 (10.1-13.0) | 1129 | 15.6 (14.6-17.0) |
| Sex | ||||||||
| Female | 1239 | 61.9 (59.3-65.0) | 3484 | 65.0 (63.4-67.0) | 1089 | 63.3 (60.3-66.0) | 3257 | 65.3 (63.6-67.0) |
| Male | 709 | 38.1 (35.5-41.0) | 1700 | 34.9 (33.3-37.0) | 600 | 36.7 (33.9-40.0) | 1576 | 34.7 (33.1-36.0) |
| Racea | ||||||||
| American Indian or Alaska Native | 69 | 2.8 (2.0-4.0) | 117 | 2.5 (1.8-3.0) | NAb | NAb | 120 | 2.7 (2.1-4.0) |
| Asian | 51 | 2.6 (1.8-4.0) | 124 | 2.5 (2.0-3.0) | 49 | 2.5 (1.8-3.0) | 117 | 2.5 (2.0-3.0) |
| Black or African American | 230 | 12.9 (11.0-15.0) | 459 | 9.3 (8.2-11.0) | 191 | 12.3 (10.3-15.0) | 425 | 9.0 (7.9-10.0) |
| White | 1487 | 73.9 (71.1-76.0) | 4232 | 79.6 (77.9-81.0) | 1295 | 75.1 (72.2-78.0) | 3945 | 80.3 (78.6-82.0) |
| Ethnicitya | ||||||||
| Hispanic | 274 | 16.2 (13.9-19.0) | 577 | 12.8 (11.4-14.0) | 239 | 15.6 (13.2-18.0) | 535 | 11.9 (10.5-13.0) |
| Non-Hispanic | 1674 | 83.8 (81.3-86.0) | 4609 | 87.2 (85.6-89.0) | 1450 | 84.4 (81.7-87.0) | 4299 | 88.1 (86.5-90.0) |
| Family | ||||||||
| Married | 553 | 32.5 (30.1-35.0) | 1806 | 39.8 (38.1-42.0) | 543 | 34.4 (31.8-37.0) | 1754 | 40.5 (38.6-42.0) |
| Unmarried partnership | 166 | 11.5 (9.8-13.0) | 403 | 10.7 (9.7-12.0) | 169 | 13.5 (11.4-16.0) | 394 | 10.9 (9.8-12.0) |
| Single | 1173 | 53.1 (50.5-56.0) | 2789 | 45.7 (44.0-47.0) | 934 | 49.5 (46.6-52.0) | 2489 | 44.3 (42.5-46.0) |
| Child in the home | 422 | 26.8 (24.5-29.0) | 1203 | 28.5 (27.0-30.0) | 458 | 32.1 (29.5-35.0) | 1289 | 31.3 (29.7-33.0) |
| Educational attainment | ||||||||
| <High school | 259 | 15.5 (13.5-18.0) | 445 | 10.1 (9.1-11.0) | 194 | 13.5 (11.6-16.0) | 385 | 9.4 (8.4-11.0) |
| High school or GED | 559 | 30.5 (28.0-33.0) | 1295 | 27.7 (26.2-29.0) | 491 | 30.6 (28.0-33.0) | 1220 | 27.4 (25.8-29.0) |
| Some college | 372 | 20.8 18.6-23.0) | 931 | 19.2 (17.9-21.0) | 316 | 20.4 (18.2-23.0) | 881 | 19.7 (18.3-21.0) |
| Associate degree | 278 | 12.8 (11.2-15.0) | 742 | 14.0 (12.9-15.0) | 231 | 12.5 (10.8-14.0) | 690 | 14.0 (12.9-15.0) |
| Bachelor’s degree | 318 | 14.2 (12.5-16.0) | 1044 | 17.5 (16.3-19.0) | 292 | 15.2 (13.4-17.0) | 1008 | 18.3 (17.0-20.0) |
| Graduate degree | 152 | 5.6 (4.7-7.0) | 707 | 10.9 (10.0-12.0) | 153 | 7.1 (6.0-8.0) | 628 | 10.8 (9.8-12.0) |
| Employment status | ||||||||
| Employed | 819 | 46.1 (43.4-49.0) | 2472 | 51.7 (49.9-53.0) | 816 | 50.9 (47.9-54.0) | 2426 | 54.3 (52.5-56.0) |
| Unemployed | 1069 | 50.8 (48.1-54.0) | 2516 | 44.5 (42.8-46.0) | 827 | 46.4 (43.5-49.0) | 2207 | 41.4 (39.7-43.0) |
| Health insurance | ||||||||
| Insured | 1795 | 90.5 (88.6-92.0) | 4905 | 93.2 (92.3-94.0) | 1553 | 90.5 (88.5-92.0) | 4577 | 93.4 (92.5-94.0) |
| Uninsured | 145 | 9.0 (7.4-11.0) | 262 | 6.4 (6.0-7.0) | 128 | 9.0 (7.4-11.0) | 244 | 6.3 (5.5-7.0) |
| Household income, % of FPL | ||||||||
| <100 | 423 | 14.3 (13.2-15.0) | 800 | 14.2 (13.1-15.0) | 350 | 14.8 (13.6-16.0) | 745 | 13.9 (12.8-15.0) |
| 100-124 | 129 | 6.0 (5.3-7.0) | 292 | 5.5 (4.7-6.0) | 119 | 5.9 (5.1-7.0) | 277 | 5.6 (4.8-6.0) |
| 125-149 | 137 | 7.3 (6.5-8.0) | 326 | 6.4 (5.6-7.0) | 113 | 6.7 (5.8-8.0) | 285 | 5.7 (5.0-7.0) |
| 150-199 | 201 | 10.1 (9.2-11.0) | 501 | 9.3 (8.4-10.0) | 166 | 9.4 (8.4-10.0) | 456 | 8.8 (7.9-10.0) |
| 200-299 | 358 | 18.0 (16.8-19.0) | 874 | 17.9 (16.7-19.0) | 306 | 18.1 (16.8-19.0) | 791 | 17.1 (15.8-18.0) |
| ≥300 | 700 | 44.3 (42.5-46.0) | 2393 | 46.7 (44.9-49.0) | 635 | 45.1 (43.3-47.0) | 2280 | 48.9 (47.0-51.0) |
| Geographic region | ||||||||
| Midwest | 414 | 20.3 (17.9-23.0) | 1278 | 23.7 (22.0-25.0) | 367 | 21.7 (19.2-24.0) | 1139 | 23.2 (21.6-25.0) |
| Northeast | 268 | 15.2 (13.0-18.0) | 799 | 16.2 (14.6-18.0) | 261 | 17.0 (14.5-20.0) | 811 | 17.8 (16.1-20.0) |
| South | 768 | 38.8 (35.8-42.0) | 1843 | 36.7 (34.5-39.0) | 639 | 37.7 (34.5-41.0) | 1788 | 37.1 (34.9-39.0) |
| West | 498 | 25.7 (22.8-29.0) | 1266 | 23.4 (21.3-26.0) | 422 | 23.5 (20.5-27.0) | 1096 | 21.9 (19.9-24.0) |
| Prescribed medications | ||||||||
| Antidepressant | 961 | 46.1 (43.5-49.0) | 2906 | 53.7 (52.0-55.0) | 770 | 42.7 (40.0-45.0) | 2170 | 42.4 (40.7-44.0) |
| Antianxiety | 959 | 46.7 (44.1-49.0) | 2632 | 49.8 (48.2-51.0) | 867 | 48.5 (45.6-51.0) | 2873 | 58.0 (56.2-60.0) |
| Physical comorbidities, No. | ||||||||
| 0 | 615 | 35.4 (32.8-38.0) | 1842 | 39.6 (37.9-41.0) | 653 | 43.1 (40.0-46.0) | 1914 | 44.1 (42.4-46.0) |
| 1 | 576 | 31.3 (28.9-34.0) | 1623 | 31.1 (29.6-33.0) | 490 | 29.2 (26.6-32.0) | 1447 | 29.6 (28.1-31.0) |
| 2 | 399 | 17.9 (16.0-20.0) | 1004 | 17.3 (16.1-19.0) | 296 | 15.8 (13.7-18.0) | 850 | 15.7 (14.5-17.0) |
| ≥3 | 358 | 15.3 (13.7-17.0) | 717 | 12.0 (11.0-13.0) | 250 | 11.9 (10.2-14.0) | 623 | 10.7 (9.8-12.0) |
Abbreviations: GED, General Educational Development; NA, not available; FPL, federal poverty level.
Data are self-reported.
Estimates are suppressed when the sample size is 30 or fewer individuals.
Prevalence of Medical Debt and Its Association With Mental Health Status
Medical debt was common among US adults with mood and anxiety disorders. Among US adults with a lifetime diagnosis of depression (n = 5186), 19.9% (95% CI, 18.6%-21.0%) reported medical debt, compared with 8.6% (95% CI, 8.2%-9.0%) of adults without a lifetime diagnosis (prevalence ratio [PR], 2.31; 95% CI, 2.31-2.32; P < .001; adjusted PR [aPR], 1.97; 95% CI, 1.96-1.98; P < .001). Similarly, among adults with a lifetime diagnosis of anxiety (n = 4834), 19.4% (95% CI, 18.0%-21.0%) reported medical debt, compared with 8.8% (95% CI, 8.3%-9.0%) of adults without a lifetime diagnosis (PR, 2.21; 95% CI, 2.21-2.22; P < .001; aPR, 1.91; 95% CI, 1.91-1.92; P < .001). Associations were more pronounced among adults with current depression (n = 1948) or anxiety (n = 1689): 27.3% (95% CI, 24.9%-30.0%) of adults with current depression had medical debt, compared with 9.4% (95% CI, 8.9%-10.0%) of adults without current depression (PR, 2.97; 95% CI, 2.96-2.97; P < .001; aPR, 2.34; 95% CI, 2.34-2.36; P < .001), and 26.2% (95% CI, 23.7%-29.0%) of adults with current anxiety had medical debt, compared with 9.6% (95% CI, 9.1%-10.0%) of adults without current anxiety (PR, 2.78; 95% CI, 2.78-2.80; P < .001; aPR, 2.24; 95% CI, 2.24-2.26; P < .001).
Medical Debt and Delayed and Forgone Mental Health Care
Medical debt was significantly associated with higher prevalence of delayed mental health care in all 4 subgroups (Table 2). Among adults with a lifetime depression diagnosis, the prevalence of delayed mental health care was 29.0% (95% CI, 25.9%-32.0%) in those with medical debt, compared with 11.6% (95% CI, 10.4%-13.0%) in adults without medical debt (aPR, 2.68; 95% CI, 2.62-2.74; P < .001). Among adults with a lifetime anxiety diagnosis, 28.0% (95% CI, 24.8%-31.0%) of those with medical debt reported delayed mental health care, compared with 11.5% (95% CI, 10.3%-13.0%) of those without medical debt (aPR, 2.45; 95% CI, 2.40-2.50; P < .001). Similar associations were observed when analyses were restricted to adults with a current disorder. Among adults with current depression, the prevalence of delayed mental health care was 36.9% (95% CI, 32.0%-42.0%) in those with medical debt vs 17.4% (95% CI, 15.0%-20.0%) in those without medical debt (aPR, 2.25; 95% CI, 2.13-2.38; P < .001). Among adults with current anxiety, 38.4% (95% CI, 33.0%-44.0%) of those with medical debt reported delayed mental health care, compared with 16.9% (95% CI, 14.6%-19.0%) of those without medical debt (aPR, 2.48; 95% CI, 2.35-2.66; P < .001).
Table 2. Prevalence of Delayed and Forgone Mental Health Care Among US Adults With Depression and Anxiety, 2022.
| Subgroup | Medical debt | Prevalence, weighted % (95% CI) | PR (95% CI) | P value | aPR (95% CI) | P value |
|---|---|---|---|---|---|---|
| Delayed mental health care | ||||||
| Lifetime depression (n = 5186) | No | 11.6 (10.4-13.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 29.0 (25.9-32.0) | 2.50 (2.45-2.57) | 2.68 (2.62-2.74) | |||
| Lifetime anxiety (n = 4834) | No | 11.5 (10.3-13.0) | 1 Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 28.0 (24.8-31.0) | 2.43 (2.38-2.50) | 2.45 (2.40-2.50) | |||
| Current depression (n = 1948) | No | 17.4 (15.0-20.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 36.9 (32.0-42.0) | 2.12 (1.96-2.31) | 2.25 (2.13-2.38) | |||
| Current anxiety (n = 1689) | No | 16.9 (14.6-19.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 38.4 (33.0-44.0) | 2.28 (2.10-2.51) | 2.48 (2.35-2.66) | |||
| Forgone mental health care | ||||||
| Lifetime depression (n = 5186) | No | 10.6 (9.5-12.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 29.4 (26.2-33.0) | 2.78 (2.72-2.85) | 2.66 (2.61-2.71) | |||
| Lifetime anxiety (n = 4834) | No | 10.7 (9.6-12.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 28.2 (24.9-32.0) | 2.63 (2.59-2.70) | 2.63 (2.57-2.68) | |||
| Current depression (n = 1948) | No | 17.2 (14.8-20.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 38.0 (33.2-43.0) | 2.21 (2.06-2.39) | 2.35 (2.23-2.48) | |||
| Current anxiety (n = 1689) | No | 17.1 (14.8-20.0) | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | 40.8 (35.3-47.0) | 2.38 (2.21-2.62) | 2.57 (2.43-2.75) |
Abbreviations: aPR, adjusted prevalence ratio; PR, prevalence ratio.
Similarly, medical debt was associated with increased prevalence of forgoing mental health care among all 4 subgroups (Table 2). The prevalence of forgoing mental health care among those with medical debt compared with those without medical debt was 29.4% (95% CI, 26.2%-33.0%) vs 10.6% (95% CI, 9.5%-12.0%) in participants with lifetime depression (aPR, 2.66; 95% CI, 2.61-2.71; P < .001), 28.2% (95% CI, 24.9%-32.0%) vs 10.7% (95% CI, 9.6%-12.0%) in those with lifetime anxiety (aPR, 2.63; 95% CI, 2.57-2.68; P < .001), 38.0% (95% CI, 33.2%-43.0%) vs 17.2% (95% CI, 14.8%-20.0%) in participants with current depression (aPR, 2.35; 95% CI, 2.23-2.48; P < .001), and 40.8% (95% CI, 35.3%-47.0%) vs 17.1% (95% CI, 14.8%-20.0%) in those with current anxiety (aPR, 2.57; 95% CI, 2.43-2.75; P < .001).
Income did not modify the association between medical debt and forgone mental health care for any group. Insurance status modified the association for adults with lifetime depression and lifetime anxiety but not current depression or current anxiety. The association between medical debt and forgone mental health care was more pronounced among insured adults compared with uninsured adults (aPR, 2.76 [95% CI, 2.70-2.82] vs 1.56 [95% CI, 1.32-1.84] for lifetime depression and 2.70 [95% CI, 2.59-2.73] vs 1.78 [95% CI, 1.46-2.16] for lifetime anxiety, respectively) (eTables 1 and 2 in Supplement 1).
Sensitivity Analysis
No qualitative differences were observed when current depression and anxiety were defined by severe symptoms (PHQ-8 or GAD-7 cutpoint ≥15). Results are presented in eTable 3 in Supplement 1.
Discussion
In this nationally representative sample of US adults with depression and anxiety, we found that medical debt is highly prevalent, with approximately 1 in 4 to 1 in 5 of these adults reporting an inability to pay medical bills. Past studies have shown that financial barriers are the most prominent barrier to mental health care among low-income adults in the US.2 Medical debt, in particular, is associated with both delayed and forgone mental health care, suggesting that medical debt may act as a barrier to timely psychiatric care.
Medical debt has become an increasing concern in the US over the last decade, as it has become a leading source of debt collections and contributes to approximately half of all bankruptcies.6,27 This mounting burden stems from increasing health care costs and inadequate insurance coverage, causing patients to bear high out-of-pocket costs.5,6,9,10 At the same time, health systems have adopted more aggressive tactics in debt collection, engaging in lawsuits, property liens, and foreclosures against patients.10,11,13 Emerging scholarship has linked medical debt to food insecurity, housing instability, and homelessness,28 underscoring medical debt as a key social determinant of health that can potentially be addressed through policy change.4
Evidence shows that policy interventions, namely the ACA and related Medicaid expansion, have been associated with reductions in medical debt, with gains concentrated among those gaining insurance coverage.5,8,9,14 However, the problem of medical debt remains common, even among insured individuals.9,29 This is supported by our own estimates presented herein, as most individuals (>90%) in this study were insured, yet 19.4% to 27.3% reported medical debt over the past year. The association of medical debt with forgone mental health care was more pronounced among insured individuals compared with uninsured individuals. This may be due to differences in service use, ie, if uninsured individuals are already forgoing care,30 they may be less sensitive to the outcomes of medical debt. Prior studies have found that health insurance does not protect against medical debt,16,31 and medical debt is common among privately insured adults with high deductibles.14 Further research is needed to understand the role of insurance, drivers of medical debt in this population, and whether medical debt stems from expenditures for mental health conditions or other comorbidities. Additionally, a recent study using data from 2 randomized clinical trials found that medical debt relief was not associated with improvements in mental health.32 Among those with the highest baseline debt, relief was associated with worsened mental health,32 which may be due to the relief being insufficient to resolve the financial distress, amplifying feelings of distress.33 These findings underscore the complexity of medical debt, catalyzing future studies to inform effective interventions.
Prevalence estimates presented herein are meaningfully higher than national estimates (10.8%-17.8%),5,6 which aligns with our finding that adults with depression or anxiety have higher prevalence of medical debt than adults without depression or anxiety, adjusted for sociodemographic factors, physical comorbidities, and health insurance status. Others have made similar observations, with recent literature noting higher odds of medical debt among individuals with mood disorders and severe psychological distress.34,35 We extend these findings by evaluating the association between medical debt and treatment-seeking patterns for mental health services. The burden of medical debt has not been equally distributed,36 and the heightened burden among adults with psychiatric disorders is likely explained by lesser financial resources among people with common mental disorders. This issue is likely compounded by low rates of psychiatrists’ participation in insurance networks,37,38 causing patients to face high out-of-pocket costs for psychiatric care.
Limitations
This study is subject to several limitations. First, the NHIS is a cross-sectional survey, and we cannot assess the temporal link between mental disorders and medical debt. Economic stressors and financial strain are risk factors for both depression and anxiety,39,40,41 meaning that medical debt may be a risk for poor mental health, but illness and disability are also risk factors for medical debt.5 Medical debt was considered over the past 12 months, whereas mental health status was measured in 2 ways: current symptoms and lifetime diagnosis. We cannot differentiate between those who had depression or anxiety before incurring medical debt and those whose medical debt may have precipitated or exacerbated anxiety and mood disorders. In either case, however, our results point to an association of medical debt with both delayed and forgone care for mental disorders. Second, our measure of lifetime diagnosis is dependent on access to health care professionals, which is associated with outcomes of interest. We sought to account for this in our analysis by adjusting for sociodemographic factors associated with access to care, which did not notably attenuate the association between medical debt and delayed or forgone mental health care. Similar associations were observed between medical debt and delayed or forgone care among those with current depression or anxiety, a measure that relies on current symptoms rather than diagnostic history and thus does not depend on prior engagement in psychiatric care. Third, medical debt disproportionately burdens Black and Latinx individuals in the US,42,43 but the lack of power in this study limited assessment of how this may be associated with racial and ethnic disparities in mental disorders and mental health service use. Fourth, groups are overlapping, as individuals may have comorbid depression and anxiety, so analyses are not fully independent. Fifth, this analysis was limited to depression and anxiety, as these were the only psychiatric disorders included in the NHIS. Medical debt prevalence may be even higher among those with severe and persistent mental and behavioral disorders.
Conclusions
Medical debt appears to contribute to the mental health treatment gap, suggesting that aggressive debt collection practices have negative consequences for population mental health. In the absence of structural health care reform, many states have enacted laws to protect patients from medical debt,11,44 but regulations have been heterogeneous across states. Further studies are needed to evaluate policies to inform protections against medical debt at the federal level. Such protections may aid in addressing barriers to treatment for individuals with mental and behavioral disorders.
eTable 1. Assessment of Insurance Status and Income as Effect Modifiers of Medical Debt on Forgone Mental Healthcare
eTable 2. Association Between Medical Debt and Forgone Mental Healthcare Among U.S. Adults Reporting a Lifetime Diagnosis of Depression or Anxiety, by Insurance Status
eTable 3. Association of Delaying and Forgoing Mental Healthcare With Medical Debt Among U.S. Adults With Severe Current Depression (PHQ-8 ≥ 15) and Severe Current Anxiety (GAD-7 ≥ 15), 2022
Data Sharing Statement
References
- 1.Tomko C, Olfson M, Mojtabai R. Gaps and barriers in drug and alcohol treatment following implementation of the Affordable Care Act. Drug Alcohol Depend Rep. 2022;5:100115. doi: 10.1016/j.dadr.2022.100115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mojtabai R. U.S. health care reform and enduring barriers to mental health care among low-income adults with psychological distress. Psychiatr Serv. 2021;72(3):338-342. doi: 10.1176/appi.ps.202000194 [DOI] [PubMed] [Google Scholar]
- 3.Kilchenstein D, Banta JE, Oh J, Grohar A. Cost barriers to health services in U.S. adults before and after the implementation of the Affordable Care Act. Cureus. 2022;14(2):e21905. doi: 10.7759/cureus.21905 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mendes de Leon CF, Griggs JJ. Medical debt as a social determinant of health. JAMA. 2021;326(3):228-229. doi: 10.1001/jama.2021.9011 [DOI] [PubMed] [Google Scholar]
- 5.Himmelstein DU, Dickman SL, McCormick D, Bor DH, Gaffney A, Woolhandler S. Prevalence and risk factors for medical debt and subsequent changes in social determinants of health in the US. JAMA Netw Open. 2022;5(9):e2231898. doi: 10.1001/jamanetworkopen.2022.31898 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kluender R, Mahoney N, Wong F, Yin W. Medical debt in the US, 2009-2020. JAMA. 2021;326(3):250-256. doi: 10.1001/jama.2021.8694 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kalousova L, Burgard SA. Debt and foregone medical care. J Health Soc Behav. 2013;54(2):204-220. doi: 10.1177/0022146513483772 [DOI] [PubMed] [Google Scholar]
- 8.Callison K, Walker B. Medicaid expansion and medical debt: evidence from Louisiana, 2014-2019. Am J Public Health. 2021;111(8):1523-1529. doi: 10.2105/AJPH.2021.306316 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Himmelstein DU, Lawless RM, Thorne D, Foohey P, Woolhandler S. Medical bankruptcy: still common despite the Affordable Care Act. Am J Public Health. 2019;109(3):431-433. doi: 10.2105/AJPH.2018.304901 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Messac L. Debt collection in American medicine—a history. N Engl J Med. 2023;389(17):1621-1625. doi: 10.1056/NEJMms2308571 [DOI] [PubMed] [Google Scholar]
- 11.Messac L. Your Money or Your Life. Oxford University Press; 2024. doi: 10.1093/oso/9780197676639.001.0001. [DOI] [Google Scholar]
- 12.Hashim F, Hennayake S, Walsh CM, et al. Characteristics of US hospitals using extraordinary collections actions against patients for unpaid medical bills: a cross-sectional study. BMJ Open. 2022;12(7):e060501. doi: 10.1136/bmjopen-2021-060501 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Richman BD, Greene SS, Chen S, Havlak J. Hospitals suing patients: how hospitals use North Carolina courts to collect medical debt. SSRN. 2023:1-30. doi: 10.2139/ssrn.4540657 [DOI] [Google Scholar]
- 14.Rabin DL, Jetty A, Petterson S, Froehlich A. Under the ACA higher deductibles and medical debt cause those most vulnerable to defer needed care. J Health Care Poor Underserved. 2020;31(1):424-440. doi: 10.1353/hpu.2020.0031 [DOI] [PubMed] [Google Scholar]
- 15.O’Toole TP, Arbelaez JJ, Lawrence RS; Baltimore Community Health Consortium . Medical debt and aggressive debt restitution practices: predatory billing among the urban poor. J Gen Intern Med. 2004;19(7):772-778. doi: 10.1111/j.1525-1497.2004.30099.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Herman PM, Rissi JJ, Walsh ME. Health insurance status, medical debt, and their impact on access to care in Arizona. Am J Public Health. 2011;101(8):1437-1443. doi: 10.2105/AJPH.2010.300080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Choi S. Experiencing financial hardship associated with medical bills and its effects on health care behavior: a 2-year panel study. Health Educ Behav. 2018;45(4):616-624. doi: 10.1177/1090198117739671 [DOI] [PubMed] [Google Scholar]
- 18.McGinty EE, Eisenberg MD. Mental health treatment gap—the implementation problem as a research problem. JAMA Psychiatry. 2022;79(8):746-747. doi: 10.1001/jamapsychiatry.2022.1468 [DOI] [PubMed] [Google Scholar]
- 19.National Center for Health Statistics . National Health Interview Survey. 2022. Accessed December 13, 2023. https://www.cdc.gov/nchs/nhis/data-questionnaires-documentation.htm
- 20.Blewett LA, Dahlen HM, Spencer D, Rivera Drew JA, Lukanen E. Changes to the design of the National Health Interview Survey to support enhanced monitoring of health reform impacts at the state level. Am J Public Health. 2016;106(11):1961-1966. doi: 10.2105/AJPH.2016.303430 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Centers for Disease Control and Prevention . Multiple Imputation of Family Income in 2022 National Health Interview Survey: Methods. Centers for Disease Control and Prevention; 2023. [Google Scholar]
- 22.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. doi: 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 23.Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092-1097. doi: 10.1001/archinte.166.10.1092 [DOI] [PubMed] [Google Scholar]
- 24.Kroenke K, Strine TW, Spitzer RL, Williams JBW, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1-3):163-173. doi: 10.1016/j.jad.2008.06.026 [DOI] [PubMed] [Google Scholar]
- 25.Cahn J, Sundaram A, Balachandar R, et al. The association of childbirth with medical debt in the USA, 2019-2020. J Gen Intern Med. 2023;38(10):2340-2346. doi: 10.1007/s11606-023-08214-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yabroff KR, Zhao J, Han X, Zheng Z. Prevalence and correlates of medical financial hardship in the USA. J Gen Intern Med. 2019;34(8):1494-1502. doi: 10.1007/s11606-019-05002-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Himmelstein DU, Thorne D, Warren E, Woolhandler S. Medical bankruptcy in the United States, 2007: results of a national study. Am J Med. 2009;122(8):741-746. doi: 10.1016/j.amjmed.2009.04.012 [DOI] [PubMed] [Google Scholar]
- 28.Bielenberg JE, Futrell M, Stover B, Hagopian A. Presence of any medical debt associated with two additional years of homelessness in a Seattle sample. Inquiry. 2020;57:46958020923535. doi: 10.1177/0046958020923535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Uppal N, Woolhandler S, Himmelstein DU. Alleviating medical debt in the United States. N Engl J Med. 2023;389(10):871-873. doi: 10.1056/NEJMp2306942 [DOI] [PubMed] [Google Scholar]
- 30.Zhou RA, Baicker K, Taubman S, Finkelstein AN. The uninsured do not use the emergency department more—they use other care less. Health Aff (Millwood). 2017;36(12):2115-2122. doi: 10.1377/hlthaff.2017.0218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Seifert RW, Rukavina M. Bankruptcy is the tip of a medical-debt iceberg. Health Aff (Millwood). 2006;25(2)(suppl 1):w89-w92. doi: 10.1377/hlthaff.25.w89 [DOI] [PubMed] [Google Scholar]
- 32.Kluender R, Mahoney N, Wong F, Yin W. The effects of medical debt relief: evidence from two randomized experiments. National Bureau of Economic Research working paper 32315. April 2024. doi: 10.3386/w32315 [DOI]
- 33.Jaroszewicz A, Jachimowicz J, Hauser O, Jamison J. Cash can make its absence felt: randomizing unconditional cash transfer amounts in the US. SSRN. 2022:1-112. doi: 10.2139/ssrn.4154000 [DOI] [Google Scholar]
- 34.Novak PJ, Ali MM, Sanmartin MX. Disparities in medical debt among U.S. adults with serious psychological distress. Health Equity. 2020;4(1):549-555. doi: 10.1089/heq.2020.0090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Grafova IB, Clifford PR, Hudson SV, et al. Disease and debt: findings from the 2019 Panel Study Of Income Dynamics in the United States. Prev Med. 2022;164:107248. doi: 10.1016/j.ypmed.2022.107248 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wiltshire JC, Elder K, Kiefe C, Allison JJ. Medical debt and related financial consequences among older African American and White adults. Am J Public Health. 2016;106(6):1086-1091. doi: 10.2105/AJPH.2016.303137 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Cummings JR. Rates of psychiatrists’ participation in health insurance networks. JAMA. 2015;313(2):190-191. doi: 10.1001/jama.2014.12472 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Benson NM, Myong C, Newhouse JP, Fung V, Hsu J. Psychiatrist participation in private health insurance markets: paucity in the land of plenty. Psychiatr Serv. 2020;71(12):1232-1238. doi: 10.1176/appi.ps.202000022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Roy-Byrne PP, Joesch JM, Wang PS, Kessler RC. Low socioeconomic status and mental health care use among respondents with anxiety and depression in the NCS-R. Psychiatr Serv. 2009;60(9):1190-1197. doi: 10.1176/ps.2009.60.9.1190 [DOI] [PubMed] [Google Scholar]
- 40.Gilman SE, Trinh NH, Smoller JW, Fava M, Murphy JM, Breslau J. Psychosocial stressors and the prognosis of major depression: a test of Axis IV. Psychol Med. 2013;43(2):303-316. doi: 10.1017/S0033291712001080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ettman CK, Cohen GH, Galea S. Is wealth associated with depressive symptoms in the United States? Ann Epidemiol. 2020;43:25-31.e1. doi: 10.1016/j.annepidem.2020.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kumar WM, Adashi EY. The medical debt burden: overdue federal action. J Gen Intern Med. 2023;38(5):1291-1292. doi: 10.1007/s11606-022-08003-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Stephenson J. Medical debt burdens millions of US adults. JAMA Health Forum. 2022;3(3):e220910. doi: 10.1001/jamahealthforum.2022.0910 [DOI] [PubMed] [Google Scholar]
- 44.Robertson CT, Rukavina M, Fuse Brown EC. New state consumer protections against medical debt. JAMA. 2022;327(2):121-122. doi: 10.1001/jama.2021.23061 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Assessment of Insurance Status and Income as Effect Modifiers of Medical Debt on Forgone Mental Healthcare
eTable 2. Association Between Medical Debt and Forgone Mental Healthcare Among U.S. Adults Reporting a Lifetime Diagnosis of Depression or Anxiety, by Insurance Status
eTable 3. Association of Delaying and Forgoing Mental Healthcare With Medical Debt Among U.S. Adults With Severe Current Depression (PHQ-8 ≥ 15) and Severe Current Anxiety (GAD-7 ≥ 15), 2022
Data Sharing Statement