Mental health behaviors and suicide risk among White and Black adults with Opioid Use Disorder.

Mental health behaviors and suicide risk among White and Black adults with Opioid Use Disorder - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Soc Sci Med . Author manuscript; available in PMC: 2026 Jun 2. Published in final edited form as: Soc Sci Med. 2025 Jun 2;381:118283. doi: 10.1016/j.socscimed.2025.118283 Mental health behaviors and suicide risk among White and Black adults with Opioid Use Disorder Hannah S Szlyk a, * , Daphne Lew b , Xiao Li a , Lindsey M Filiatreau a , JaNiene Peoples c , Erin Kasson a , Patricia Cavazos-Rehg a Author information Article notes Copyright and License information a Department of Psychiatry, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8314, St. Louis, MO, 63130, USA b Institute for Informatics, Data Science, and Biostatistics, Washington University in St. Louis School of Medicine, Campus Box 8132, St. Louis, MO, 63110, USA c Brown School Washington University in St. Louis, 1 Brookings Drive, Campus Box 1196, St. Louis, MO, 63130, USA * Corresponding author. szlyk@wustl.edu (H.S. Szlyk). Issue date 2025 Sep. PMC Copyright notice PMCID: PMC12758180  NIHMSID: NIHMS2115919  PMID: 40472645 The publisher's version of this article is available at Soc Sci Med Abstract This study explores differences in depression and anxiety symptoms and suicidal behaviors by symptom severity and race among White and Black adults with Opioid Use Disorder (OUD). We examine the viability of using PHQ-9 and GAD-7 symptoms scores to indicate past suicidal behavior by race. We analyze baseline data among individuals who were enrolled in a supplemental digital intervention intended to support adults with OUD in recovery. White (n = 345) or Black (n = 76) individuals were included in the analytic sample. Participants completed an online survey on depression (PHQ-9) and anxiety (GAD-7) symptoms, past 30-day (P30D) suicidal ideation (yes/no), and lifetime suicide attempts (yes/no). Descriptive statistics were calculated; mean symptom scores were compared by race. Race-stratified precision-recall (PR) curves were used to determine the extent to which mental health scores indicated P30D suicidal ideation and the optimal cut points for each measure. Whites had significantly higher depression and anxiety scores than Black participants. When comparing past suicidal behaviors, there were no significant differences by race. A larger proportion of Black participants with moderate or severe depression and anxiety endorsed past suicidal behavior compared to Whites. PR curves and area under the curve values suggested that optimal cut-points for mental health scores as an indicator for suicidal ideation differed by race. Alone, mental health symptom severity and total scores may not be ideal indicators of suicide risk. Mental health screenings plus suicide risk evaluations are recommended for OUD patients, especially Black Americans. Keywords: Opioid use disorder, Black Americans, Mental health, Suicidal ideation, Suicide attempt 1. Introduction The United States (U.S.) has been amid an opioid epidemic since the late 1990s. Over two decades later, opioids continue to drive increases in the age-adjusted mortality rates from drug overdoses in the U.S., (from 1.0 per 100,000 in 2013 to 21.8 per 100,000 in 2021).( Garnet and Minino, 2024 ) The number of opioid-related overdose deaths among all ages more than doubled from 2010 to 2019 (21,089 to over 47, 600) followed by alarming increases during the COVID-19 pandemic (68,630 reported deaths in 2020 and 80,411 reported deaths in 2021) ( Drug Overdose Death Rates, 2023 ). The intersection of a global COVID-19 pandemic with the continued opioid epidemic has garnered renewed interest in and careful attention to identifying concurrent risk factors contributing to the opioid crisis in the U.S. Opioid use disorder (OUD) can be a destabilizing health condition and is associated with considerable morbidity, mortality, and economic burden ( Ayoo et al., 2020 ). Suicide has been referred to as the “silent contributor” to drug-related overdose deaths (including opioids) with data indicating that at least 20 % of drug-related overdose deaths may be intentional ( Johnson et al., 2013 ; Oquendo and Volkow, 2018 ). Individuals with OUD are at heightened risk of dying by suicide and often report histories of suicidal behaviors that are higher than the general population. For instance, Ali and Dubenitz analyzed 2015–2018 data from the National Survey on Drug Use and Health (NSDUH) and found the rate of suicidal behavior (ideation and/or attempt) among those with OUD and without OUD to be 22 % and 4 %, respectively ( Ali and Dubenitz, 2021 ). Greater empirical attention is needed to better understand the mechanisms underlying suicide risk among patients with OUD. Psychiatric comorbidities such as depression and anxiety are established critical risk factors for opioid misuse ( Conway et al., 2006 ), overdose ( Fendrich et al., 2019 ; van Draanen et al., 2020 ), and suicide, including suicidal ideation and suicidal attempts ( Bentley et al., 2016 ). Providers need to continue assessing the combined impact of co-occurring psychiatric disorders on suicide risk due to the increased risk in mortality ( Bohnert et al., 2017 ), and among those receiving treatment for OUDs. Suicidal behavior can increase the burden on healthcare utilization ( Shepard et al., 2016 ). Patients with OUD, depression and anxiety, and heightened suicide risk may represent a critical high-risk group requiring close monitoring and continued empirical focus. Until recently, the opioid epidemic was broadly mislabeled as an issue solely impacting White Americans ( Hansen and Netherland, 2016 ; Bechteler and Kane-Willis, 2017 ). Yet, Black Americans are disproportionately impacted by the opioid crisis ( Townsend et al., 2022 ), demonstrated by a 575 % increase in opioid-related mortality compared to a 184 % increase among White Americans between 2007 and 2019. Simultaneously, during this time, there have been reported increases in suicide rates among Black Americans ( Horigian et al., 2022 ). Specifically, this trend has been observed among adolescents and emerging adult Black males ( Han et al., 2022 ). Similar to the opioid crisis, suicide has long been considered a problem specific to older White males in the U.S. Reasons contributing to this false narrative include beliefs that Black Americans are “resilient” to factors that impact suicide risk (e.g., psychiatric symptomatology, life crises), stigma around the topic of suicide, presumptions about the under-reporting of Black American suicidal behaviors and deaths, and inappropriate measurement of risk factors and symptoms ( Woodward et al., 2013 ; Bailey et al., 2019 ). Due to the underestimation of the impact of the opioid and suicide crises on Black Americans, social and structural factors such as fewer economic resources ( Artiga et al., 2021 ), health education inequities ( Eversman, 2015 ), poorer access to healthcare ( Edelman et al., 2016 ), and exposure to stigma and discrimination ( Stevens-Watkins, 2020 ) often prevent Black Americans from accessing and benefiting from life-saving care for substance misuse and mental health risk factors for suicide. Despite the tremendous increases in opioid-related deaths and suicidal behaviors among Black Americans (i.e., overall drug overdose death rates are now highest among Black older men ( Han et al., 2022 ), with increasing rates among ages 15–24 largely attributed to fentanyl ( Kariisa et al., 2022 ), inconsistent findings limit our understanding of differences in depression and anxiety symptoms and suicidal behaviors among White and Black Americans with OUD. For example, some studies report higher prevalence of lifetime Major Depression Disorder among non-Hispanic Whites than Black Americans ( Williams et al., 2007 ; Assari et al., 2015 ), while other studies report increases in or substantial proportions of suicidal ideation among both Black and White patients with OUD ( Na et al., 2022 ), but no significant differences between groups ( Kuramoto et al., 2012 ). Available literature also suggests that White Americans are more likely to be diagnosed with anxiety disorders than Black Americans ( Asnaani et al., 2010 ). There are potential methodological reasons for this incomplete understanding of how mental health compares among White and Black Americans. Notably, high impact survey studies, including those cited above ( Williams et al., 2007 ; Assari et al., 2015 ; Na et al., 2022 ; Kuramoto et al., 2012 ; Asnaani et al., 2010 ), tend to utilize highly structured sampling protocols and study materials that are designed specifically for their projects. In turn, studies may not be designed to accurately capture the socioecological experience of Black participants with OUD ( Jalali et al., 2020 ), as well as other minoritized racial and ethnic groups, in contrast to White Americans. For example, one study may only survey households, likely omitting certain subgroups from participating (e.g., those who have unstable housing situations or who are incarcerated); another study may misclassify a participant’s racial identity depending on how questions were asked, and response options were designed. Additionally, participants may not be asked certain questions about their mental health depending on the skip logic of the respective survey (e.g., not asked about past year suicide attempts if the person answers “no” to past year suicidal ideation) ( Bommersbach et al., 2023 ). Overall, it can be challenging to compare findings even across nationally representative survey studies. Item-bias (how items that comprise a measurement scale are systematically related to various exogenous variables ( Cole et al., 2000 ) is another potential reason for observed inconsistencies in depression and anxiety diagnoses and suicide risk detection across races. For example, there is mixed evidence that widely used standardized measures have the same function and predictive validity across racial groups ( McClure et al., 2023 ; Keum et al., 2018 ), especially among Black Americans ( Keum et al., 2018 ; Harry et al., 2021 ). Even well-established measurement instruments for depression and anxiety have demonstrated non-equivalent results across groups. The Patient Health Questionnaire 9 (PHQ-9) has demonstrated mixed results in tests of measurement invariance across racial and ethnic groups in the US ( Harry et al., 2021 ), including among Black Americans. More often, two factor models have demonstrated better fits versus the standard one-factor model across diverse racial and ethnic groups ( Keum et al., 2018 ). Yet, some studies have found both one- and two-factor models to be good fits (( Patel et al., 2019 ), while others have found a good fit for the one-factor model ( Huang et al., 2006 ). In contrast, the measurement invariance of the Generalized Anxiety Disorder 7 (GAD-7) has been less studied among subgroups, especially across racial and ethnic groups ( Riglea et al., 2025 ). Like the PHQ-9, the available literature also includes varied support for one, two, or both factor models of the GAD-7 ( Sriken et al., 2022 ; White and Karr, 2023 ). Lastly, clinician bias may also influence the assessment of symptoms ( Bao et al., 2008 ). Thus, it is important to gain a better understanding of the viability of depression and anxiety classification levels and of suicide risk screening between Black and White patients ( Asnaani et al., 2010 ; Jalali et al., 2020 ; McClure et al., 2023 ), as research findings can have life-saving implications for screening, prevention, and treatment for patients with OUD. To address these gaps in the literature, we used data from a community-based sample of White and Black participants with OUD who reside in Missouri, predominately within the St. Louis metropolitan area, to explore differences in depression and anxiety symptoms and suicidal behaviors. Next, we examined differences in suicidal behaviors stratified by race and severity of depression and anxiety symptoms (minimal or mild, moderate or severe). Finally, we leveraged the PHQ-9 and the GAD-7 to examine the viability of using depression and anxiety symptoms scores to indicate past 30-day suicidal and behavior by race. We hypothesized the following: 1) There would be significant differences in depression and anxiety symptoms and suicidal ideation and behavior by race, with Black participants reporting greater symptoms than their White peers; 2) the traditional cut-off scores for clinical symptom severity for the PHQ-9 and GAD-7 would be less viable for Black participants than White participants. Importantly, drawing on a sample of Black Americans with OUD from the geographic location of St. Louis, Missouri responds to urgent calls for more research in medium metropolitan areas, including St. Louis, which are experiencing the steepest increases in opioid overdose deaths among Black Americans ( Banks et al., 2021 ; Lippold et al., 2019 ). 2. Methods 2.1. Study design and recruitment In total, 681 individuals with OUD were recruited between January 2020 and October 2021 for a service project to provide free, supplemental support to people in all stages of substance use disorder (SUD) recovery. After consent was obtained, 475 participants completed the baseline survey. Fifty-four were excluded from the analytic dataset because they were missing key measures (n = 21) or not the population of interest (e.g., other races/ethnicities, n = 33). Our final sample included 421 participants (see Fig. 1 ). The present study analyzes baseline data from these individuals who were enrolled in a supplemental digital intervention ( uMAT-R) (Reference blinded for peer), which aimed to support the recovery of adults with OUD. The study sample was recruited from several settings located in a mid-size, Midwestern city (St. Louis, Missouri). St. Louis is like other medium sized cities as drug overdose death rates continue to increase at the highest rate among Black Americans ( Banks et al., 2021 ), and overdose deaths are largely connected to fentanyl. The uMAT-R app included content on evidence-based treatment recovery options taken from SAMHSA’s online decision support tool entitled “Decisions in Recovery: Treatment for Opioid Use Disorder.” At the time of the study’s data collection, the app included over 40 distinct education courses on recovery-related topics such psychoeducation about substance use and mental health, basic information about Medication for Opioid Use Disorder (MOUD), prevention of returning to use, management of withdrawal symptoms, and the increased risk of recurrence associated with polysubstance use. Participants also had access to a directory of community resources with the ability to access an e-coach to support app usage and provide psychological support. Additional app features included goal setting, appointment and medication reminders, sober day tracking, and daily motivational quotes which appeared on the homepage. Fig. 1. Open in a new tab Flow chart for participants selected into the analytic sample (N = 421). At the eligible settings, members of the research team presented information about the uMAT-R intervention and left study flyers. During the initial months of the COVID-19 pandemic, providers at the settings would share study flyers with eligible clients. Less often, some participants were recruited via snowball sampling methods, a strategy used to access hard-to-reach populations ( Wohl et al., 2017 ) including individuals with an SUD ( Barati et al., 2014 ). Eligibility criteria included 1) used opioids at some point in their life; 2) 18 years or older; 3) U.S. resident, 4) fluent in English, and 5) owned a smartphone with an iOS or Android operating system. Interested individuals would then contact the research team either via phone or email. All study responses were self-reported using the online Research Electronic Data Capture (REDCap) system, which is a secure, web-based application that can be used either on a computer or mobile device. Research staff reviewed the informed consent document with eligible participants via phone or in-person and participants provided verbal or written consent to join the study. All participants provided informed consent for this study, and to have their data used in publication. The study was performed in line with the principles of the Declaration of Helsinki. This study was approved by the university’s Institutional Review Board (blinded for peer review). 2.2. Measures 2.2.1. Mental health symptoms Lifetime suicide attempt. One item, “have you ever, in your whole life, tried to kill yourself or made a suicide attempt,” was used to assess if participants had ever attempted suicide (yes/no) ( Kuramoto et al., 2012 ; Santaella-Tenorio et al., 2022 ; Wright and Ramirez, 2021 ). Past 30-day (P30D) suicidal ideation. One item asked if participants had experienced suicidal ideation in the past 30 days of completing the survey (yes/no): “Has there been a time in the past month when you have had serious thoughts about ending your life?” Previous studies specific to people with OUD have also used one-item screeners to assess past suicidal ideation ( Kuramoto et al., 2012 ; Darke et al., 2004 ). Depression. The Patient Health Questionnaire-9 (PHQ-9 ( Kroenke et al., 2001 )) was used to assess symptoms of depression in the past two weeks. The nine-item standardized measure is used to screen for depression severity based on DSM-IV depression criteria. Participants were asked to identify the frequency of symptoms, from “0” (not at all) to “3” (nearly every day). The PHQ-9 has demonstrated excellent internal reliability (e.g., a = 0.89 in primary care, a = 0.91 in outpatient substance use treatment) ( Dum et al., 2008 ), test-retest reliability, validity, and feasibility ( Kroenke et al., 2001 ; Wulsin et al., 2002 ). The total scale score was generated by summing all item scores, which range from 0 to 27, theoretically. In the current study, we applied a score of 10 as the cut-off point for minimal to mild (<10) and moderate to severe depression (≥10) ( Kroenke et al., 2001 ). Sensitivity analyses excluding the question asking about “Thoughts that you would be better off dead or of hurting yourself in some way” in the calculation of the PHQ score were also performed. Anxiety. The Generalized Anxiety Disorder-7 (GAD-7 ( Spitzer et al., 2006 )) is a 7-item measure used to assess symptoms of anxiety. Item scores range from “0” (not at all) to “3” (nearly every day). The GAD-7 has been validated for use in primary care settings ( Spitzer et al., 2006 ), outpatient drug treatment ( Delgadillo et al., 2012 ), and in the general population ( Löwe et al., 2008 ). To generate a total scale score, item scores are summed such that potential scores range from 0 to 21 with scores equal to or greater than 10 indicative of moderate or severe anxiety symptoms ( Spitzer et al., 2006 ). Lifetime and P30D substance use. Among all participants, substance use status was assessed for tobacco, cannabis, alcohol, opioids, stimulants, and other (non-opioid) illicit drugs at lifetime and within P30D, respectively. The assessment of opioids included heroin, methadone, morphine, OxyContin, codeine, oxycodone, fentanyl, hydrocodone, dilaudid, and other opioids. Stimulants were evaluated based on cocaine use (including injection and snorting, mixed use), crack cocaine, amphetamines, benzodiazepines, and other hallucinogens. Binary variables were created to represent the presence or absence of each substance use, indicating a “yes” or “no” response. Demographic covariates. Demographic variables, including age (18–30, 31–50, 51 or above), sex (male, female, others), race and ethnicity (dichotomized to White vs. Black), current employment status (unemployed, employed (part-time included)), and insurance coverage (no, yes) were captured via self-report. 2.3. Statistical analysis Descriptive statistics were calculated as means and standard deviations (SDs) for continuous variables and frequencies and percentages for categorical variables. Skewness (Sk) and kurtosis (K) statistics were also calculated for mental health symptom scores. Comparisons of means were conducted using t-tests whereas comparisons of proportions were conducted using Chi-Square or Fisher’s Exact tests, as appropriate. Reliability of scales in the full sample and stratified by race were assessed by calculating Cronbach’s alpha. Logistic regression models were used to examine associations between mental health symptoms and past 30-day suicidal ideation after controlling for relevant covariates, in the full sample and stratified by race. To determine the extent to which mental health scores reliably indicate past 30-day suicidal ideation, precision-recall (PR) curves were used. Separate curves were calculated for PHQ-9 and GAD-7, and these PR curves were stratified by race to determine whether the predictive ability of the mental health scores differ across racial groups. The area under the PR curve (AUC-PR) was calculated as a measure of predictive ability, with values greater than the prevalence of past 30-day suicidal ideation indicating better prediction. Optimal cut points based on the PR curves were also identified using the maximum F1 score ( Goutte and Gaussier, 2005 ). Analyses were performed in SAS for Windows version 9.4 (SAS Institute, Cary, NC) and α = 0.05 was used to determine statistical significance. 3. Results Most participants were between the ages of 31–50 years old (68.4 %), female (56.8 %), had completed high school or greater (80.5 %), were unemployed (61.7 %), had stable housing (69 %), and had some type of insurance coverage (55.1 %) ( Table 1 ). There were meaningful differences between participants included in the analytic sample versus those who were not with regards to age and race ( Supplemental Table 1 ). When comparing characteristics by race, a larger proportion of White participants had stable housing (72 %) than Blacks (56.6 %; χ 2 (1, N = 419) = 6.95, p = 0.01); a larger proportion of Black participants were ages 51 (21.1 %; χ 2 (2, N = 421) = 14.62, p < 0.001) and up than Whites (8.7 %); a higher proportion of White participants had completed high school or higher education (82.3 %) compared to Black participants (72.4 %; χ 2 (1, N = 421) = 3.93; p = 0.05). During the past 30 days, most participants had used tobacco products (85.9 %) ( Table 2 ). Regarding lifetime use, most participants endorsed use of all the substances assessed. A larger proportion of Black participants reported opioid use in the past 30 days (39.2 %) than Whites (25.2 %; χ 2 (1, N = 408) = 5.96, p = 0.02). With the exception of tobacco and opioids, a larger proportion of White participants reported lifetime use of all substances assessed than Blacks. Table 1. Demographic characteristics by race (N = 421). Overall (N = 421) White (N = 345) Black (N = 76) N (%) χ 2 df P-value Age 18–30 87 (20.7) 80 (23.2) 7 (9.2) 14.62 2 <0.001 31–50 288 (68.4) 235 (68.1) 53 (69.7) 51+ 46 (10.9) 30 (8.7) 16 (21.1) Gender Male 181 (43.0) 154 (44.6) 27 (35.5) 2.39 2 0.16 Female 239 (56.8) 190 (55.1) 49 (64.5) Others 1 (0.2) 1 (0.3) 0 Education attainment High school below 82 (19.5) 61 (17.7) 21 (27.6) 3.93 1 0.05 High school completed or above 339 (80.5) 284 (82.3) 55 (72.4) Employment status Unemployed 259 (61.7) 206 (59.7) 53 (70.7) 3.12 1 0.08 Employed (part-time included) 161 (38.3) 139 (40.3) 22 (29.3) Missing 1 0 1 Housing/Living situation Unstable housing a 129 (30.8) 96 (28.0) 33 (43.4) 6.95 1 0.01 Stable housing 290 (69.2) 247 (72.0) 43 (56.6) Missing 2 2 0 Insurance coverage b No 189 (44.9) 159 (46.1) 30 (39.5) 1.10 1 0.29 Yes 232 (55.1) 186 (53.9) 46 (60.5) Open in a new tab NA, Not Applicable. Bold indicates statistically significant. a The assessment of unstable housing includes living in shelters, streets, treatment facilities, and medical centers. b The assessment of insurance coverage includes Medicaid, TRICARE, CHAMPUS, CHAMPVA, the VA, military health care, and private health insurance. Table 2. P30D and lifetime substance use by race (N = 421). Overall (N = 421) White (N = 345) Black (N = 76) N (%) χ 2 (with df = 1) P-value P30D use Tobacco No 57 (14.1) 46 (13.9) 11 (15.1) 0.07 0.8 Yes 347 (85.9) 285 (86.1) 62 (84.9) Missing 17 14 3 Marijuana No 287 (71.0) 231 (69.6) 56 (77.8) 1.93 0.16 Yes 117 (29.0) 101 (30.4) 16 (22.2) Missing 17 13 4 Alcohol No 308 (76.2) 255 (76.8) 53 (73.6) 0.33 0.56 Yes 96 (23.8) 77 (23.2) 19 (26.4) Missing 17 13 4 Opioids 1 No 295 (72.3) 250 (74.9) 45 (60.8) 5.96 0.02 Yes 113 (27.7) 84 (25.2) 29 (39.2) Missing 13 11 2 Stimulants 2 No 258 (75.9) 215 (76.2) 43 (74.1) 0.12 0.73 Yes 82 (24.1) 67 (23.8) 15 (25.9) Missing 81 63 18 Other illicit drugs No 323 (80.6) 265 (80.3) 58 (81.7) 0.07 0.79 Yes 78 (19.5) 65 (19.7) 13 (18.3) Missing 20 15 5 Lifetime use Tobacco No 19 (4.8) 13 (4.0) 6 (8.5) 2.57 0.11 Yes 379 (95.2) 314 (96.0) 65 (91.5) Missing 23 18 5 Marijuana No 42 (10.7) 26 (8.0) 16 (23.5) 14.21 <0.001 Yes 351 (89.3) 299 (92.0) 52 (76.5) Missing 28 20 8 Alcohol No 49 (12.3) 33 (10.1) 16 (22.5) 8.30 0.01 Yes 348 (87.7) 293 (89.9) 55 (77.5) Missing 24 19 5 Opioids 1 No 17 (4.2) 13 (3.9) 4 (5.4) 0.32 0.53 Yes 387 (95.8) 317 (96.1) 70 (94.6) Missing 17 15 2 Stimulants 2 No 23 (6.9) 15 (5.4) 8 (13.8) 5.31 0.02 Yes 313 (93.1) 263 (94.6) 50 (86.2) Missing 85 67 18 Other illicit drugs No 100 (25.1) 74 (22.6) 26 (37.1) 6.52 0.01 Yes 298 (74.9) 254 (77.4) 44 (62.9) Missing 23 17 6 Open in a new tab P30D, Past 30-Day. The assessment of opioids includes heroin, methadone, morphine, oxycontin, codeine, oxycodone, fentanyl, hydrocodone, dilaudid, and other opioids. The assessment of stimulants includes cocaine (injected and snorted, mixed included), crack, rock, amphetamines, benzodiazepines, or other hallucinogens. Bold indicates statistically significant. The overall Cronbach’s alpha for the PHQ-9 in this sample was 0.90, and reliability did not meaningfully differ across the two racial groups (α = 0.90 among Whites and α = 0.92 among Blacks). Whites reported a mean PHQ-9 score of 9.1 (SD: 6.6, Sk: 0.5, K: −0.5), which was significantly higher than the mean score reported by Blacks (6.8, SD: 6.4, Sk: 0.9, K: 0.07; t (106.71) = 2.79, p < 0.01). A larger proportion of Whites reported moderate or severe depression (42 %) compared to Blacks (29 %; χ 2 (1, N = 415) = 4.28, p = 0.04). For the GAD-7, the overall Cronbach’s alpha was 0.94 in the full sample and when stratified by race. Similar results were found with regards to comparisons of GAD-7 scores, with Whites reporting a significantly higher mean score (8.6, SD: 6.4, Sk: 0.5, K: −0.8) compared to Blacks (6.5, SD: 5.9, Sk: 0.9, K: 0.2; t (116.23) = 2.77, p < 0.01). A significantly larger proportion of Whites also reported moderate or severe anxiety (39 %) compared to Blacks (25 %; χ 2 (1, N = 412) = 5.06, p = 0.03). In contrast, a higher proportion of Black individuals (18 %) reported past 30-day suicidal ideation compared to White individuals (13 %), though the difference was not statistically significant (χ 2 (1, N = 420) = 1.66, p = 0.20). There was also no meaningful difference in the proportions reporting lifetime suicide attempts among White (41 %) versus Black participants (43 %; χ 2 (1, N = 421) = 0.10, p = 0.75). When stratifying by symptom severity, we did not observe statistically significant differences in P30D suicidal ideation or lifetime suicide attempts by race, though there were differences in the magnitude of proportions with P30D suicidal ideation. Specifically, a larger proportion of Blacks with moderate or severe depression reported P30D suicidal ideation (38 %) compared to Whites (23 %; Table 3 ). Similarly, Blacks with mild or moderate anxiety reported higher P30D suicidal ideation (37 %) compared to Whites (23 %). These findings persisted when examining lifetime suicide attempts, with Blacks with mild or moderate depression or anxiety reporting larger proportions (67 % and 58 %, respectively) than Whites (45 % for both depression and anxiety). Table 3. Prevalence of suicidality by mental health severity and race. Depression No to mild (N = 251) Moderate to severe (N = 164) White Black White Black N (%) χ 2 (with df = 1) p-value N (%) χ 2 (with df = 1) p-value P30D Suicidal thoughts No 189 (95.0) 47 (90.4) 1.55 0.21 109 (76.8) 13 (61.9) 2.14 0.14 Yes 10 (5.0) 5 (9.6) 33 (23.2) 8 (38.1) Lifetime Suicide attempts No 121 (60.8) 35 (67.3) 0.74 0.39 79 (55.2) 7 (33.3) 3.52 0.06 Yes 78 (39.2) 17 (32.7) 64 (44.8) 14 (66.7) Anxiety No to mild (N = 261) Moderate to severe (N = 151) White Black White Black N (%) χ 2 (with df = 1) p-value N (%) χ 2 (with df = 1) p-value P30D Suicidal thoughts No 193 (94.2) 49 (87.5) 2.88 0.09 101 (67.3) 12 (8.0) 1.74 0.19 Yes 12 (5.8) 7 (12.5) 30 (20.0) 7 (4.7) Lifetime Suicide attempts No 124 (60.5) 35 (62.5) 0.07 0.78 73 (55.3) 8 (42.1) 1.16 0.28 Yes 81 (39.5) 21 (37.5) 59 (44.7) 11 (57.9) Open in a new tab A cutoff at 9 was applied to define no to mild (0–9) vs. moderate to severe (10 or above). Since there were statistically significant differences in ages, housing status, and past 30-day opioid use by race, we conducted analyses that examine the relationship between mental health symptoms and past 30-day suicidal ideation in the full sample and stratified by race, including the respective covariates (see Supplemental Table 2 ). Mental health symptom scores were significantly associated with suicidality (OR: 1.17, 95 % CI: 1.11, 1.23 for PHQ-9; OR: 1.13, 95 % CI: 1.08, 1.18 for GAD-7), but none of the covariates (age, past 30-day opioid use, and housing status) were meaningfully associated with past 30-day suicidal ideation. Thus, these covariates were not included in the models used to generate the PR curves to determine the predictive ability of the mental health scores and the optimal cut points for PHQ-9 and GAD-7 by race. PR curves indicated that both PHQ-9 and GAD-7 had better predictive ability for identifying P30D suicidal ideation than the null model ( Fig. 2 ), though AUC-PR values were still fairly low. There were also slight differences in AUC-PR values by race, with higher AUC-PR values observed among Black individuals than Whites. Specifically, the AUC-PR for PHQ-9 was 0.34 for White individuals and 0.47 for Blacks, while the AUC for GAD-7 was 0.25 for White individuals and 0.42 for Blacks. The optimal cut-points identified by the PR curves did differ by race and were generally higher than the traditionally recommended cut-points for both measures. For the PHQ-9, the optimal cut-point among Black participants was 14, whereas among White participants the optimal cut-point was 16. For the GAD-7, the optimal cut-point among Blacks was 13, whereas among Whites the optimal cut-point was 17. Sensitivity analyses excluding the self-harm question from the PHQ score yielded no meaningful differences in findings, with an AUC-PR of 0.29 and an optimal cut-point of 17 for White participants versus an AUC-PR of 0.48 and optimal cut-point of 16 for Black participants ( Supplemental Fig. 1 ). Fig. 2. Open in a new tab Precision-recall (PR) analysis of area under the curve (AUC-PR) and optimal cut points of depression and anxiety by race. For the full sample (N = 421). 4. Discussion In this study, we explored differences in depression and anxiety symptoms and suicidal behaviors among a community-based sample of White and Black participants with OUD who reside in a medium metropolitan area (St. Louis, Missouri). Next, we examined differences in suicidal behaviors stratified by race and severity of depression and anxiety symptoms (minimal or mild, moderate or severe). Lastly, we examined the viability of using PHQ-9 and GAD-7 symptoms scores to indicate past 30-day suicidal ideation by race. We found significant differences in moderate and severe depression and anxiety symptoms when using traditionally recommended cut-off scores for the respective standardized instruments (PHQ-9 and GAD-7), with White participants being more likely to report these symptoms in the past two weeks than Black peers. We did not find significant differences in past 30-day suicidal ideation and lifetime suicide attempts by race. When stratifying prevalence of suicidal behavior by level of depression and anxiety symptom severity, we found no significant differences by race. This study contributes to the dearth of literature on differences in mental health symptoms and suicidal behaviors among White and Black individuals with OUD – two racial groups that experienced high opioid-related mortality over the past decade. It also adds to the critical conversation on race, mental health, behavioral health symptoms, and measurement ( Woodward et al., 2013 ). Our finding that there were no significant differences in suicidal ideation and suicide attempts by race is in line with studies that have used samples from nationwide surveys ( Na et al., 2022 ; Kuramoto et al., 2012 ), where the proportion of Black and White participants with OUD who endorse suicidal behavior is very similar. As such, suicidal behavior and suicide risk among people with OUD is not exclusive to White patients, and this should be reflected in the care that Black patients receive. Specifically, our and others’ findings underscore that prevention of suicide among Black patients with OUD needs to be considered an urgent and prioritized public health issue. It is important to note that the prevalence of suicidal ideation and behavior among our sample is higher than some studies on people with opioid use in the literature; yet the range in ideation and attempts reported in the literature widely varies (e.g., lifetime attempts reported as 27 %, 48 % respectively ( Lent et al., 2023 ; Trémeau et al., 2008 )). Reasons for elevated prevalence may be that many of our participants reported factors that could imply greater psychosocial challenges and, thus, heightened suicide risk (e.g., fewer years of formal education, unemployment, insured by Medicaid) ( Hillary et al., 2019 ). Additionally, our data was collected more recently than many newly published studies and most studies differ in their measurement of suicidal behaviors, (e.g., past 12 months, lifetime) ( Na et al., 2022 ; Lent et al., 2023 ) and by type of sample (e.g., treatment seekers, nationally representative data) ( Trémeau et al., 2008 ; Hillary et al., 2019 ), which makes it hard to compare findings. The high prevalence of suicidal ideation and behavior may reflect the toll of COVID-19 on people with OUD, especially during the worst years of the pandemic. Investigation on prevalence of suicidal ideation and behaviors in this population post-pandemic is warranted. Despite not finding statistically significant difference in past 30-day suicidal ideation and lifetime suicide attempts by race, we did find that the proportion of Black participants with moderate or above depressive and anxiety symptoms who also endorsed suicidal behaviors was higher than Whites. The stark difference in endorsement of past 30-day suicidal ideation in the presence of at least moderate depressive and anxiety symptoms between Black and Whites (38 % vs. 23 %, 37 % vs. 23 % respectively) is alarming, especially considering that epidemiological trends show increases in suicide rates for Black Americans and decreasing for White individuals ( Ramchand et al., 2021 ). Therefore, our findings help to dispel past false narratives that have downplayed the mental health risks among Black Americans in comparison to Whites, and, thus, contributed to disparities in available and appropriate preventive care and treatment for this group. Because self-report mental health instruments are commonly used as screening tools in a variety of care settings (e.g., mental health settings, primary care settings, substance use treatment settings ( Mulvaney-Day et al., 2018 )), the potential for under capturing mental health problems among patients with OUD can consequently decrease the likelihood of these individuals being treated for their mental health symptoms. Thus, we examined the viability of using depression (PHQ-9) and anxiety (GAD-7) symptom scores to indicate past 30-day suicidal behavior by race. Results of the ROC analyses suggest that the current classification level for moderate or severe depression and anxiety symptoms may be too low to accurately classify White participants reporting past 30-day suicidal ideation. Among our sample, a larger proportion of Whites reported moderate or severe depression or anxiety but no suicidal ideation. In contrast, the classification level for moderate or severe anxiety symptoms may be too high for Black individuals; a larger proportion of Black participants not classified as having moderate or severe anxiety reported past 30-day suicidal ideation. These findings contribute to existing literature highlighting racial and ethnic disparities in the under-identification and underreporting of mental health symptoms among those with substance use disorders ( Priester et al., 2016 ), including studies that suggest well-established mental health screening instruments (e.g., Center for Epidemiological Studies-Depression Scale) may underestimate symptoms in Black Americans ( Jalali et al., 2020 ; Improving Cultural Competence, 2014 ). For example, Coley and colleagues ( Coley et al., 2021 ) found that death by suicide risk prediction models, which included PHQ-9 item 9 (often used as a screener for suicide risk), performed poorly in Black and American Indian/Alaskan Native patients as compared to White, Asian, and Hispanic patients. Overall, as supported by our study, future research requires larger sample sizes (especially subgroup samples), additional sensitivity analysis methods, and qualitative exploration of symptom expression to inform best practices for evaluating the clinical validity of and use of mental health measures across groups ( Vaughn-Coaxum et al., 2016 ; Pahwa et al., 2021 ). Providers should assess for mental health symptoms in addition to suicidal behaviors, as alone, one symptom or behavior may not offer an accurate profile of potential suicide risk, particularly for patients with OUD. For Black patients with OUD, the presence of other factors (e.g., unemployment, lower income, fewer years of education, perceived impact of discrimination and racism on wellbeing) may help to identify those at greater risk for mental health symptoms ( Shim et al., 2012 ; Assari and Caldwell, 2017 ). Inadvertently, lowering symptom thresholds can increase the number of false positives ( Sheldrick et al., 2015 ), which could result in increased distress for the patient, misallocation of provider time and resources, and, potentially, iatrogenic hospitalizations ( Ward-Ciesielski and Rizvi, 2021 ; Olfson et al., 2016 ). Strategies including multiple gate screening (i.e., successive rounds of screening to rule out false positives), prioritizing non-emergent resources (e.g., community health workers, telehealth, referrals), and differentiating between acute and non-acute cases of suicidality may help to prevent potential harms ( Scott et al., 2010 ; Horowitz et al., 2023 ). Tools that can help to produce a valid evidence-based assessment, like the DSM-5 Cultural Formulation Interview (CFI ( American Psychiatric Association, 2013 )) may also assist clinicians with better understanding their patient’s conceptualization and perspective of the identified problem ( Turner et al., 2016 ). Furthermore, future research is needed to test the viability of different cut-off scores in indicating suicidal behaviors and with larger samples of Black patients with OUD. Lastly, providers can recommend Medication for Opioid Use Disorder (MOUD) as a part of an effective treatment plan for OUD patients ( Wakeman et al., 2020 ). MOUD may also help to address co-occurring mental health symptoms and prevent suicidal behavior. Specifically, buprenorphine has shown promise for reducing depressive symptoms and suicidal ideation during OUD treatment ( Latif et al., 2019 ) and preventing future overdose ( Mattick et al., 2014 ). Yet, most OUD patients do not receive MOUD treatment ( Jones et al., 2015 ), especially patients who represent racialized and minoritized groups ( Barnett et al., 2023 ). Historically, buprenorphine – which is considered a safer MOUD that allows for dosage outside of a regulated clinic – has been less accessible in areas where lower income, non-privately insured, and non-White patients live ( Goedel et al., 2020 ; Hansen et al., 2016 ). Therefore, Black patients may require additional support with access to MOUD options beyond methadone, including financial assistance. Providers may also advocate for expanded access to buprenorphine and other MOUDs in established treatment settings and the publicly funded mental health system ( Olfson et al., 2021 ). It is important to note limitations of the present study. First, our study uses cross-sectional baseline data, thus limiting our ability to draw causal inferences between our variables. Longitudinal cohort studies are needed to inform causal effects. Second, data used in this study is based on participant self-report and may be subject to recall bias, although self-report tools have been shown to encourage honest disclosure of mental health and substance use ( Benjamin et al., 2011 ; McNeely et al., 2014 ). Also, the subsamples of Black and White participants varied in age and housing status, factors that should be considered when interpreting the results. Further, there were meaningful differences in age and race for participants included in the analytic sample versus those excluded, though the proportion of individuals excluded from the sample was small. The generalizability of the findings may be limited as this population of individuals was relatively highly educated and recruited from only one midwestern metropolitan area in the U.S. Still, multi-method approaches to the identification of depression, anxiety, and suicidal behaviors are needed. Our small sample size is also noted as a limitation and we acknowledge that our analyses may be underpowered, particularly among the subset of Black OUD patients. While smartphone ownership is ubiquitous in the U.S. (98 %) ( Pew Research Center, 2024 ), certain subpopulations often have lower rates of smartphone ownership, including Black Americans. Thus, some individuals may have been excluded from the study due to lack of access to a digital device. The smaller subsample of Black OUD participants may also be reflective of the many barriers to treatment initiation for Black residents in St. Louis and, thus, mirror the consequential underrepresentation of this population in many local treatment facilities. Future studies should replicate our study’s findings with larger, random and representative samples, and consider strategies to mitigate potential participants’ lack of or limited access to necessary digital tools (e.g., budgeting for refurbished devices in a grant). Approximately 30 % of our recruited sample with OUD did not log into the app and/or complete the baseline survey. Available research with individuals with OUD highlights that engaging and retaining such participants in studies, including digital interventions can be a challenge. This population may especially feel ambivalent about treatment and research opportunities due to concerns about privacy and stigma ( Harris et al., 2022 ); Additionally, struggles with meeting basic needs ( Williams et al., 2017 ) which impacts consistencies in communication (e.g., disconnected cell phone numbers or lack of mailing permanent address) ( Hawk et al., 2021 )), and an elevated risk of premature mortality ( Harris et al., 2022 ) contribute to challenges regarding engagement and retention. Notably, our study’s rates of attrition and survey completion are like those of other mobile health studies with this population ( Hawk et al., 2021 ; Hodges et al., 2022 ). Despite notable limitations, there were several strengths of this study that should be mentioned. First, this study drew upon a community-based sample of adult patients with OUD in a geographical area where Black Americans are witnessing unprecedented increases in opioid overdose deaths ( Lippold et al., 2019 ). Therefore, our study responds to a public health crisis on a national and regional level. To our knowledge, our study is one of the first to examine Black-White differences in suicide risk by mental health symptom severity and Black-White differences in the viability of using two well-established screening instruments (PHQ-9 and GAD-7) as indicators of suicidal behavior in the same study, particularly among patients with OUD. 5. Conclusion Our findings indicate that suicide risk among adults with OUD is not exclusive to White patients and that better identification, prevention, and treatment efforts are needed to support Black patients with OUD. Findings also highlight a health equity issue and suggest that among Black and White adults engaged in OUD treatment, Black Americans’ mental health symptoms may be significantly under-diagnosed compared to Whites, which has important research, early identification, prevention, and treatment implications. Alone, mental health symptom severity and total scores may not be ideal indicators of suicide risk. Mental health screenings plus suicide risk evaluations are recommended for OUD patients, especially Black Americans. Future research may address study limitations by testing the viability of different cut-off scores in indicating suicidal behaviors, assessing potential implications of using consistent cut-off scores across heterogeneous populations and examining larger samples of racially diverse patients. Supplementary Material Supplement NIHMS2115919-supplement-Supplement.docx (173.3KB, docx) Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.socscimed.2025.118283 . Funding This study was supported by the National Institutes of Health [grant number K02 DA043657]; the Substance Abuse and Mental Health Services Administration [grant numbers 1H79TIO80271 and 1H79TI081697]; the National Institute on Drug Abuse of the National Institutes of Health[grant numbers K12 DA041449 and T32DA015035]; the National Institute on Minority Health & Health Disparities under Award Number T37MD014218–04 The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse or the National Institutes of Health. Footnotes CRediT authorship contribution statement Hannah S. Szlyk: Writing – review & editing, Writing – original draft, Project administration, Methodology, Conceptualization. Daphne Lew: Writing – review & editing, Writing – original draft, Validation, Formal analysis, Data curation. Xiao Li: Writing – review & editing, Writing – original draft, Formal analysis. Lindsey M. Filiatreau: Writing – review & editing, Writing – original draft, Validation. JaNiene Peoples: Writing – review & editing, Writing – original draft. Erin Kasson: Writing – review & editing, Writing – original draft. Patricia Cavazos-Rehg: Writing – review & editing, Writing – original draft, Funding acquisition. Ethics approval This study was approved by the Washington University School of Medicine (WUSM) Institutional Review Board (study # 201910161, first approved on 12/17/2019). All procedures were performed in compliance with the relevant laws and institutional guidelines of the WUSM IRB. Declaration of competing interest The first author is a consultant for Google Health. The senior author is a consultant for Rissana, LLC, and PredictView. Data availability Data will be made available on request. References Ali MM, Dubenitz J, 2021. Suicidal behavior, opioid use disorder, and behavioral health treatment: prevalence and correlates among adults in the United States 2015–2018. J. Subst. Abuse Treat 10.1016/j.jsat.2021.108413. [ DOI ] [ PubMed ] [ Google Scholar ] American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders: DSM-5, 5. American Psychiatric Association, Washington, DC. [ Google Scholar ] Artiga S, Hill L, Orgera K, Damico A, 2021. Health Coverage by Race and Ethnicity-2010–2019. Kaiser Family Foundation. https://www.kff.org/racial-equity-and-health-policy/issue-brief/health-coverage-by-race-and-ethnicity/ . (Accessed 20 July 2022). [ Google Scholar ] Asnaani A, Richey JA, Dimaite R, Hinton DE, Hofmann SG, 2010. A cross-ethnic comparison of lifetime prevalence rates of anxiety disorders. J. Nerv. Ment. Dis 198 (8), 551–555. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Assari S, Caldwell CH, 2017. The link between mastery and depression among black adolescents; ethnic and gender differences. Behav. Sci 7. Pii–E32. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Assari S, Burgard SA, Zivin K, 2015. Long term reciprocal associations between depression and chronic medical conditions; longitudinal support for black-white health paradox. J. Racial Ethn. Health Dispar 2 (4), 589–597. [ DOI ] [ PubMed ] [ Google Scholar ] Ayoo K, Mikhaeil J, Huang A, Wąsowicz M, 2020. The opioid crisis in North America: facts and future lessons for Europe. Anaesthesiol. Intensive Ther 52, 139–147. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bailey RK, Mokonogho J, Kumar A, 2019. Racial and ethnic differences in depression: current perspectives. Neuropsychiatric Dis. Treat 603–609. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Banks DE, Carpenter RW, Wood CA, Winograd RP, 2021. Commentary on Furr-Holden et al. : As Opioid Overdose Deaths Accelerate among Black Americans, COVID-19 Widens Inequities-A Critical Need to Invest in Community-based Approaches. Addict 116, 686. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bao Y, Fisher J, Studnicki J, 2008. Racial differences in behavioral inpatient diagnosis: examining the mechanisms using the 2004 Florida inpatient discharge data. J. Behav. Health Serv. Res 35, 347–357. [ DOI ] [ PubMed ] [ Google Scholar ] Barati M, Ahmadpanah M, Soltanian A, 2014. Prevalence and factors associated with methamphetamine use among adult substance abusers. J. Res. Health Sci 14, 221–226. [ PubMed ] [ Google Scholar ] Barnett ML, Meara E, Lewinson T, Hardy B, Chyn D, Onsando M, et al. , 2023. Morden NE. Racial inequality in receipt of medications for opioid use disorder. N. Engl. J. Med 388, 1779–1789. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bechteler SS, Kane-Willis K, 2017. Whitewashed: the African American opioid epidemic. The Chicago Urban League. https://chiul.org/wp-content/uploads/2019/01/Whitewashed-AA-Opioid-Crisis-11-15-17_EMBARGOED_-FINAL.pdf . (Accessed 3 October 2023). [ Google Scholar ] Benjamin S, Herr NR, McDuffie J, Williams JW Jr, Nagi A, Wing L, 2011. Performance characteristics of self-report instruments for diagnosing generalized anxiety and panic disorders in primary care: a systematic review. In: Health Services Research and Department Services. (Accessed 3 October 2023). [ PubMed ] [ Google Scholar ] Bentley KH, Franklin JC, Ribeiro JD, Kleiman EM, Fox KR, Nock MK, 2016. Anxiety and its disorders as risk factors for suicidal thoughts and behaviors: a meta-analytic review. Clin. Psychol. Rev 43, 30–46. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bohnert KM, Ilgen MA, Louzon S, McCarthy JF, Katz IR, 2017. Substance use disorders and the risk of suicide mortality among men and women in the US Veterans Health Administration. Addict 112, 1193–1201. [ DOI ] [ PubMed ] [ Google Scholar ] Bommersbach TJ, Rosenheck RA, Rhee TG, 2023. Racial and ethnic differences in suicidal behavior and mental health service use among US adults, 2009–2020. Medicine 53 (12), 5592–5602. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Cole SR, Kawachi I, Maller SJ, Berkman LF, 2000. Test of item-response bias in the CES-D scale. experience from the New Haven EPESE study. J. Clin. Epidemiol 53 (3), 285–289. [ DOI ] [ PubMed ] [ Google Scholar ] Coley RY, Johnson E, Simon GE, Cruz M, Shortreed SM, 2021. Racial/ethnic disparities in the performance of prediction models for death by suicide after mental health visits. JAMA Psychiatry 78 (7), 726–734. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Conway KP, Compton W, Stinson FS, Grant BF, 2006. Lifetime comorbidity of DSM-IV mood and anxiety disorders and specific drug use disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J. Clin. Psychiatry 67, 247–257. [ DOI ] [ PubMed ] [ Google Scholar ] Darke S, Ross J, Lynskey M, Teesson M, 2004. Attempted suicide among entrants to three treatment modalities for heroin dependence in the Australian Treatment Outcome Study (ATOS): prevalence and risk factors. Drug Alcohol Depend 73, 1–10. [ DOI ] [ PubMed ] [ Google Scholar ] Delgadillo J, Payne S, Gilbody S, Godfrey C, Gore S, Jessop D, Dale V, 2012. Brief casefinding tools for anxiety disorders: validation of GAD-7 and GAD-2 in addictions treatment. Drug Alcohol Depend 125, 37–42. [ DOI ] [ PubMed ] [ Google Scholar ] Drug Overdose Death Rates, 2023. National Institute of drug Abuse. https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates . (Accessed 3 October 2023). [ Google Scholar ] Dum M, Pickren J, Sobell LC, Sobell MB, 2008. Comparing the BDI-II and the PHQ-9 with outpatient substance abusers. Addict. Behav 33, 381–387. [ DOI ] [ PubMed ] [ Google Scholar ] Edelman EJ, Hansen NB, Cutter CJ, et al. , 2016. Implementation of integrated stepped care for unhealthy alcohol use in HIV clinics. Addiction Sci. Clin. Pract 11 (1), 1–14. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Eversman MH, 2015. “We want a living solution”: views of harm reduction programs in black U.S. Communities. J. Ethn. Subst. Abuse 14 (2), 187–207. [ DOI ] [ PubMed ] [ Google Scholar ] Fendrich M, Becker J, Hernandez-Meier J, 2019. Psychiatric symptoms and recent overdose among people who use heroin or other opioids: results from a secondary analysis of an intervention study. Addict Behav Rep 10.1016/j.abrep.2019.100212. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Garnet MF, Minino AM, 2024. Drug overdose deaths in the United States, 2003–2023. NCHS Data Brief, no. 522. National Center for Health Statistics, Hyattsville, MD. [ Google Scholar ] Goedel WC, Shapiro A, Cerda M, Tsai JW, Hadland SE, Marshall BDL, 2020. Association of racial/ethnic segregation with treatment capacity for opioid use disorder in counties in the United States. JAMA Netw. Open 3 (4), e203711. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Goutte C, Gaussier E, 2005. A probabilistic interpretation of precision, recall and F-score, with implication for evaluation. In: European Conference on Information Retrieval. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 345–359. [ Google Scholar ] Han B, Einstein EB, Jones CM, Cotto J, Compton WM, Volkow ND, 2022. Racial and ethnic disparities in drug overdose deaths in the US during the COVID-19 pandemic. JAMA Netw. Open. 10.1001/jamanetworkopen.2022.32314. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hansen H, Netherland J, 2016. Is the prescription opioid epidemic a white problem? Am. J. Publ. Health 106, 2127–2129. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hansen H, Siegel C, Wanderling J, Dirocco D, 2016. Buprenorphine and methadone treatment for opioid dependence by income, ethnicity and race of neighborhoods in New York City. Drug Alcohol Depend 164, 14–21. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Harris R, Rosecrans A, Zoltick M, Willman C, Saxton R, Cotterell M, Bell J, Blackwell I, Page KR, 2022. Utilizing telemedicine during COVID-19 pandemic for a low-threshold, street-based buprenorphine program. Drug Alcohol Depend 10.1016/2Fj.drugalcdep.2021.109187. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Harry ML, Coley RY, Waring SC, Simon GE, 2021. Evaluating the cross-cultural measurement invariance of the PHQ-9 between American Indian/Alaska Native adults and diverse racial and ethnic groups. J. Affect. Disord. Rep 4, 100121. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hawk K, Malicki C, Kinsman J, D’Onfrio G, Taylor A, Venkatesh A, 2021. Feasibility and acceptability of electronic administration of patient reported outcomes using mHealth platform in emergency department patients with non-medical opioid use. Addiction Sci. Clin. Pract 16, 66. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hillary S, Stuart EA, Olfson M, 2019. Opioid use and misuse and suicidal behaviors in a nationally representative sample of US adults. Am. J. Epidemiol 188 (7), 245–1253. 10.1093/aje/kwz061. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hodges J, Waselewski M, Harrington W, Franklin T, Schorling K, Huynh J, Tabackman A, Otero K, Ingersoll K, Tiouririne NAD, Flickinger T, 2022. Six-month outcomes of the HOPE smartphone application designed to support treatment with medications for opioid use disorder and piloted during an early statewide COVID-19 lockdown. Addiction Sci. Clin. Pract 17, 16. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Horigian VE, Schmidt RD, Shmueli-Blumberg D, Hefner K, Feinberg J, Kondapaka R, Tross S, 2022. Suicidality as a predictor of overdose among patients with substance use disorders. J. Clin. Med 11, 6400. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Horowitz LM, Ryan PC, Wei AX, Boudreaux ED, Ackerman JP, Bridge JA, 2023. Screening and assessing suicide risk in medical settings: feasible strategies for early detection. Focus 21 (2), 145–151. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Huang FY, Chung H, Kroenke K, Delucchi KL, Spitzer RL, 2006. Using the patient health questionnaire-9 to measure depression among racially and ethnically diverse primary care patients. J. Gen. Intern. Med 21 (6), 547–552. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Improving cultural competence. In: National Library of Medicine. Center for Substance Abuse Treatment, 2014. https://pubmed.ncbi.nlm.nih.gov/25356450/ . (Accessed 3 October 2023). [ Google Scholar ] Jalali MS, Botticelli M, Hwang RC, Koh HK, McHugh RK, 2020. The opioid crisis: a contextual, social-ecological framework. Health Res. Pol. Syst 18, 1–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Johnson EM, Lanier WA, Merrill RM, Crook J, Porucznik CA, Rolfs RT, Sauer B, 2013. Unintentional prescription opioid-related overdose deaths: description of decedents by next of kin or best contact, Utah, 2008–2009. J. General Intern. Med [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Jones CM, Campopiano M, Baldwin G, McCance-Katz E, 2015. National and state treatment need and capacity for opioid agonist medication-assisted treatment. Am. J. Publ. Health 105 (8), e55–e63. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Kariisa M, Davis NL, Kumar S, et al. , 2022. Vital signs: drug overdose deaths, by selected sociodemographic and social determinants of health characteristics — 25 States and the District of Columbia, 2019–2020. MMWR Morb. Mortal. Wkly. Rep 71, 940–947. 10.15585/mmwr.mm7129e2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Keum BT, Miller MJ, Inkelas KK, 2018. Testing the factor structure and measurement invariance of the PHQ-9 across racially diverse US college students. Psychol. Assess 30, 1096–1106, 2018. [ DOI ] [ PubMed ] [ Google Scholar ] Kroenke K, Spitzer RL, Williams JB, 2001. The PHQ-9: validity of a brief depression severity measure. J. Gen. Intern. Med 16, 606–613. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Kuramoto SJ, Chilcoat HD, Ko J, Martins SS, 2012. Suicidal ideation and suicide attempt across stages of nonmedical prescription opioid use and presence of prescription opioid disorders among US adults. J. Stud. Alcohol Drugs 73, 178–184. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Latif ZEH, Benth Jš, Solli KK, Opheim A, Kunoe N, Krajci P, Sharma-Haase K, Tanum L, 2019. Anxiety, depression, and insomnia among adults with opioid dependence treated with extended-release naltrexone vs buprenorphine-naloxone: a randomized clinical trial and follow-up study. JAMA Psychiatry 76, 127–134. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lent MR, Dugosh KL, Hurstak E, et al. , 2023. Prevalence and predictors of suicidality among adults initiating office-based buprenorphine. Addiction Sci. Clin. Pract 18. 10.1186/s13722-023-00393-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lippold KM, Jones CM, Olsen EOM, Giroir BP, 2019. Racial/ethnic and age group differences in opioid and synthetic opioid–involved overdose deaths among adults aged≥ 18 years in metropolitan areas—United States, 2015–2017. MMWR Morb. Mortal. Wkly. Rep 68, 967. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Löwe B, Decker O, Müller S, Brähler E, Schellberg D, Herzog W, Herzberg PY, 2008. Validation and standardization of the generalized anxiety disorder screener (GAD-7) in the general population. Med. Care 266–274. [ DOI ] [ PubMed ] [ Google Scholar ] Mattick RP, Breen C, Kimber J, Davoli M, 2014. Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database Syst. Rev 2. 10.1002/14651858.cd002207.pub2. [ DOI ] [ PubMed ] [ Google Scholar ] McClure K, Bell KA, Jacobucci R, Ammerman BA, 2023. Measurement invariance and response consistency of single-item assessments for suicidal thoughts and behaviors. Psychol. Assess 35, 830–841. 10.1037/pas0001268. [ DOI ] [ PubMed ] [ Google Scholar ] McNeely J, Strauss SM, Wright S, Rotrosen J, Khan R, Lee JD, Gourevitch MN, 2014. Test–retest reliability of a self-administered alcohol, smoking and substance involvement screening test (ASSIST) in primary care patients. J. Subst. Abuse Treat 47, 93–101, 2014. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Mulvaney-Day N, Marshall T, Downey Piscopo K, Korsen N, Lynch S, Karnell LH, et al. , 2018. Screening for behavioral health conditions in primary care settings: a systematic review of the literature. J. Gen. Intern. Med 33, 335–346. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Na PJ, Bommersbach TJ, Petrakis IL, Rhee TG, 2022. National trends of suicidal ideation and mental health services use among US adults with opioid use disorder, 2009–2020. eClinicalMedicine 54, 101696. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Olfson M, Wall M, Wang S, Crystal S, Liu SM, Gerhard T, Blanco C, 2016. Short-term suicide risk after psychiatric hospital discharge. JAMA Psychiatry 73 (11), 1119–1126. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Olfson M, Zhang VS, King M, Mojtabai R, 2021. Changes in buprenorphine treatment after Medicaid expansion. Psychiatr. Serv 72 (6), 633–640. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Oquendo MA, Volkow ND, 2018. Suicide: a silent contributor to opioid-overdose deaths. N. Engl. J. Med 10.1056/NEJMp1801417. [ DOI ] [ PubMed ] [ Google Scholar ] Pahwa R, Brekke JS, Jaccard J, 2021. Cross-ethnic invariance of BSI-18 in individuals with serious mental illness: new perspectives on measurement invariance testing. J. Soc. Soc. Work. Res 12 (4), 755–787. [ Google Scholar ] Patel JS, Oh Y, Rand KL, Wu W, Cyders MA, Kroenke K, Stewart JC, 2019. Measurement invariance of the Patient Health Questionnaire-9 (PHQ-9) depression screener in US adults across sex, race/ethnicity, and education level: NHANES 2005–2016. Depress. Anxiety 36 (9), 813–823. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Pew Research Center, 2024. Mobile fact sheet. Available from: https://www.pewresearch.org/internet/fact-sheet/mobile/ . (Accessed 6 March 2025). [ Google Scholar ] Priester MA, Browne T, Iachini A, Clone S, DeHart D, Seay KD, 2016. Treatment access barriers and disparities among individuals with co-occurring mental health and substance use disorders: an integrative literature review. J. Subst. Abuse Treat 61, 47–59. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Ramchand R, Gordon JA, Pearson JL, 2021. Trends in suicide rates by race and ethnicity in the United States. JAMA Netw. Open 4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Reference Blinded for Peer-Review. [ Google Scholar ] Riglea T, Wellman RJ, Sylvestre MP, Sabiston C, O’Loughlin J, 2025. Factor structure and measurement invariance of the GAD-7 across time, sex, and language in young adults. J. Affect. Disord 375, 359–365. [ DOI ] [ PubMed ] [ Google Scholar ] Santaella-Tenorio J, Martins S, Cerdá M, Olfson M, Keyes K, 2022. Suicidal ideation and attempts following nonmedical use of prescription opioids and related disorder. Psychol. Med 10.1017/S0033291720002160. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Scott M, Wilcox H, Huo Y, Turner JB, Fisher P, Shaffer D, 2010. School-based screening for suicide risk: balancing costs and benefits. Am. J. Publ. Health 100 (9), 1648–1652. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Sheldrick RC, Benneyan JC, Kiss IG, Briggs-Gowan MJ, Copeland W, Carter AS, 2015. Thresholds and accuracy in screening tools for early detection of psychopathology. J. Child Psychol. Psychiatry 56 (9), 936–948. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shepard DS, Gurewich D, Lwin AK, Reed GA Jr., Silverman MM, 2016. Suicide and suicidal attempts in the United States: costs and policy implications. Suicide Life Threat. Beyond Behav 46, 352–362. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shim RS, Ye J, Baltrus P, Fry-Johnson Y, Daniels E, Rust G, 2012. Racial/ethnic disparities, social support, and depression: examining a social determinant of mental health. Ethn. Dis 22, 15–20. [ PMC free article ] [ PubMed ] [ Google Scholar ] Spitzer RL, Kroenke K, Williams JB, Löwe B, 2006. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch. Intern. Med 166, 1092–1097. [ DOI ] [ PubMed ] [ Google Scholar ] Sriken J, Johnsen ST, Smith H, Sherman MF, Erford BT, 2022. Testing the factorial validity and measurement invariance of college student scores on the Generalized Anxiety Disorder (GAD-7) scale across gender and race. Meas. Eval. Counsel. Dev 55 (1), 1–6. [ Google Scholar ] Stevens-Watkins D, 2020. Opioid-related overdose deaths among African Americans: implications for research, practice and policy. Drug Alcohol Rev 39 (7), 857–861. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Townsend T, Kline D, Rivera-Aguirre A, et al. , 2022. Racial/ethnic and geographic trends in combined stimulant/opioid overdoses, 2007–2019. Am. J. Epidemiol 19, 599–612. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Trémeau F, Darreye A, Staner L, Corrêa H, Weibel H, Khidichian F, Macher JP, 2008. Suicidality in opioid-dependent subjects. Am. J. Addict 17 (3), 187–194. [ DOI ] [ PubMed ] [ Google Scholar ] Turner EA, Mills CJ, 2016. Culturally relevant diagnosis and assessment of mental illness. In: Breland-Noble A, Al-Mateen C, Singh N (Eds.), Handbook of Mental Health in African American Youth. Springer Series on Child and Family Studies. Springer, Cham. 10.1007/978-3-319-25501-9_2. [ DOI ] [ Google Scholar ] van Draanen J, Tsang C, Mitra S, et al. , 2020. Mental disorder and opioid overdose: a systematic review. Soc. Psychiatr. Psychiatr. Epidemiol 57, 647–671. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Vaughn-Coaxum RA, Mair P, Weisz JR, 2016. Racial/ethnic differences in youth depression indicators: an item response theory analysis of symptoms reported by White, Black, Asian, and Latino youths. Clin. Psychol. Sci 4 (2), 239–253. 10.1177/2167702615591768. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wakeman SE, Larochelle MR, Ameli O, et al. , 2020. Comparative effectiveness of different treatment pathways for opioid use disorder. JAMA Netw. Open 3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Ward-Ciesielski EF, Rizvi SL, 2021. The potential iatrogenic effects of psychiatric hospitalization for suicidal behavior: a critical review and recommendations for research. Clin. Psychol. Sci. Pract 28 (1), 60. [ Google Scholar ] White AE, Karr JE, 2023. Psychometric properties of the GAD-7 among college students: reliability, validity, factor structure, and measurement invariance. Transl. Issues Psychol. Sci [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Williams DR, González HM, Neighbors H, Nesse R, Abelson JM, Sweetman J, et al. , 2007. Prevalence and distribution of major depressive disorder in African Americans, Caribbean Blacks, and non-Hispanic whites: results from the national survey of American life. Arch. Gen. Psychiatry 64 (3), 305–315. [ DOI ] [ PubMed ] [ Google Scholar ] Williams AR, Barbieri V, Mishlen K, Levin FR, Nunes EV, Mariani JJ, Bisaga A, 2017. Long-term follow-up study of community-based patients receiving XR-NTX for opioid use disorders. Am. J. Addict 26, 319–325. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wohl AR, Ludwig-Barron N, Dierst-Davies R, Kulkarni S, Bendetson J, Jordan W, et al. , 2017. Project engage: snowball sampling and direct recruitment to identify and link hard-to-reach HIV-infected persons who are out of care. JAIDS J. Acquir. Immune Defic. Syndr 75, 190–197. [ DOI ] [ PubMed ] [ Google Scholar ] Woodward AT, Taylor RJ, Abelson JM, Matusko N, 2013. Major depressive disorder among older African Americans, Caribbean Blacks, and non-Hispanic whites: secondary analysis of the National Survey of American Life. Depress. Anxiety 30, 589–597. [ DOI ] [ PubMed ] [ Google Scholar ] Wright N, Ramirez MR, 2021. A cross sectional study of non-medical use of prescription opioids and suicidal behaviors among adolescents. Inj. Epidemiol 10.1186/s40621-021-00329-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wulsin L, Somoza E, Heck J, 2002. The feasibility of using the Spanish PHQ-9 to screen for depression in primary care in Honduras. Prim. Care Companion CNS Disord 4, 24967. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplement NIHMS2115919-supplement-Supplement.docx (173.3KB, docx) Data Availability Statement Data will be made available on request. ACTIONS View on publisher site PDF (695.6 KB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Cannabis Use Among Individuals Treated with Medication for Opioid Use Disorder: Correlates, Patterns, and Motivations for Use.#236099
• Opioid Use Disorder Practice by Licensure Category in North Carolina.#371650
• Racial Inequality in Receipt of Medications for Opioid Use Disorder in Nationwide Outpatient Settings-A Cross-sectional Study.#630942
• Differential Changes in Use of Medications for Opioid Use Disorder by Race-Ethnicity: Effects of a Hub-and-Spoke Model.#312591
• 42 U.S.C. § 290ee-3a — Grant program for State and Tribal response to opioid use disorders#595084