Skip to main content Accessibility help


  • Access


      • Send article to Kindle

        To send this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle. Find out more about sending to your Kindle.

        Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

        Find out more about the Kindle Personal Document Service.

        Depressive role impairment and subthreshold depression in older black and white women: race differences in the clinical significance criterion
        Available formats

        Send article to Dropbox

        To send this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Dropbox.

        Depressive role impairment and subthreshold depression in older black and white women: race differences in the clinical significance criterion
        Available formats

        Send article to Google Drive

        To send this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Google Drive.

        Depressive role impairment and subthreshold depression in older black and white women: race differences in the clinical significance criterion
        Available formats
Export citation



We examined race differences in the DSM-IV clinical significance criterion (CSC), an indicator of depressive role impairment, and its impact on assessment outcomes in older white and black women with diagnosed and subthreshold depression.


We conducted a secondary analysis of a community-based interview study, using group comparisons and logistic regression.


Lower-income neighborhoods in a Midwestern city.


411 community-dwelling depressed and non-depressed women ≥ 65 years (45.3% Black; mean age = 75.2, SD = 7.2) recruited through census tract-based telephone screening.


SCID interview for DSM-IV to assess major depression and dysthymia; Center for Epidemiologic Studies-Depression Scale to define subthreshold depression (≥16 points); Mini-Mental State Examination, count of medical conditions, activities of daily living, and mental health treatment to assess health factors.


Black participants were less likely than Whites to endorse the CSC (11.8% vs. 24.1%; p = .002). There were few race differences in depressive symptom type, severity, or count. Blacks with subthreshold depression endorsed more symptoms, though this comparison was not significant after adjustments. Health factors did not account for race differences in CSC endorsement. Disregarding the CSC-eliminated differences in diagnosis rate, race was a significant predictor of CSC endorsement in a logistic regression.


Race differences in CSC endorsement are not due to depressive symptom presentations or health factors. The use of the CSC may lead to underdiagnosis of depression among black older adults. Subthreshold depression among Blacks may be more severe compared to Whites, thus requiring tailored assessment and treatment approaches.


Depression is one of the most common mental disorders and represents a serious public health concern among both black and white older adults (Ford et al., 2007; Haigh et al., 2017). However, comparative studies of prevalence of depression by race have produced mixed results. Epidemiological surveys conducting formal diagnostic assessments using specific criteria (American Psychiatric Association, 2013) have found lower rates of major depressive disorder (MDD) and dysthymia among older Blacks compared to Whites (Blazer et al., 1987; Jimenez et al., 2010). For example, a recent population survey estimated lifetime prevalence of MDD and dysthymia among older adults to be 13.2% and 3.2% among Whites compared to 5.1% and 1.3% among Blacks, respectively (Woodward et al., 2012). In contrast, surveys using self-report measures to assess depressive symptoms among older adults have found higher or equal rates in Blacks compared to Whites (Azar et al., 2005; Blazer et al., 1998; Skarupski et al., 2005).

Another important category within the depression severity continuum is subthreshold depression (SubD), which refers to depressive syndromes that represent significant symptomatology but do not meet diagnostic criteria. Among older adults, SubD is associated with excess role impairment and morbidity at levels similar to that of diagnosed disorder (Chuan et al., 2008; Meeks et al., 2011). Little is known about SubD in older black adults; however, past research suggests that Blacks may be at higher risk for SubD compared to Whites (Adams and Moon, 2009; Judd et al., 1994). Taken together, existing evidence suggests that older black adults experience significant and debilitating depressive symptoms at a rate similar to Whites but are less likely to meet diagnostic criteria for depressive disorders.

Reasons for this race discordance are unclear, but several possibilities can be considered. First, there may be true race-based differences in depressive symptom constellation. Previous work has suggested that somatic complaints are more prominent among Blacks with depression compared to Whites (Adebimpe et al., 1982; Ayalon and Young, 2003). This could result in black elders being less likely to meet diagnostic criteria for depressive disorder, which require the presence of specific symptoms (i.e. depressed mood or anhedonia). Second, cultural differences in stigma related to mental illness may lead to less willingness among Blacks to disclose mood concerns (Conner et al., 2010). For similar reasons, this could lead to lower rates of depression diagnosis among Blacks in clinical and research settings. Third, the discordance may stem from the assessment approach: elderly black adults may interpret their symptoms differently compared to Whites and, thus, respond differently to assessment of diagnostic criteria. In support of this hypothesis, Ward and colleagues (2014) found that older black women view depression as a normal reaction to life circumstances and, thus, may not recognize or report acute severity of symptoms. In sum, the presence of one or more of these phenomena could contribute to differential assessment outcomes and subsequently to fewer Blacks receiving a depression diagnosis. Given known disparities in mental health diagnosis and treatment for Blacks (Jimenez et al., 2013), there is a critical need to improve our understanding of the factors that contribute to racial differences in depression diagnosis.

This study focuses on one diagnostic criterion that has received limited attention in the study of race differences in mental disorder, the clinical significance criterion (CSC). In both the 4th and 5th edition of the Diagnostic and Statistical Manual for Mental Disorders (DSM-IV and DSM-5), diagnosis of major depressive disorder or dysthymia requires the meeting of several criteria. These include Criterion A, the presence of depressed mood or anhedonia; and Criterion B, a minimum total symptom count (see Table S1). A further key criterion is the CSC, which reflects the presence of “clinically significant” distress or impairment in social, occupational, or other important areas of functioning due to depressive symptoms (American Psychiatric Association, 1994, 2013; Spitzer and Wakefield, 1999). Previous work using this dataset showed that black older women were than white older women to endorse the CSC, though there were no group differences in depression scale score or number of depressive symptoms (Heller et al., 2010). That study was limited in that it did not examine race differences in the types of symptoms present (e.g. symptom domains on the depression scale) or in the fulfillment of specific diagnostic criteria (e.g. Criterion A) that may impact diagnosis rates among those with elevated symptoms. Specifically, as noted above, if black older adults with subthreshold depression present with different symptom constellations than their white counterparts, this may result in fewer Blacks receiving a diagnosis — independent of CSC endorsement. Further, the potential impact of other conditions that may underlie race differences in the CSC in our sample, such as medical comorbidities or functional deficits, has not been explored.

As others have noted (Bell and McBride, 2011), an improved understanding of the impact of assessment approach on race-based differences in mental health is an important step in addressing healthcare disparities. To this end, the current study aims to extend previous findings of race differences in the endorsement of the CSC by examining the impact of these differences on diagnosis rates and exploring potential contributing factors. In light of data suggesting that potential race differences in the lived experience of depression may be gender-specific (Black and Woods-Giscombe, 2012; Ward et al., 2014), our study focuses on older black and white women only. In addition, low-income areas were targeted for recruitment to minimize potential confounding between race and socioeconomic status (Blazer et al., 1998).

Figure S1 presents a graphical illustration of our conceptual model and each of the research questions guiding the present study. Given the limited past research on the CSC among older adults, our approach was primarily exploratory. We first undertook a careful examination of the impact of CSC endorsement on diagnosis rates by race by comparing two different diagnostic algorithms. Thus, our first research question was: Does the race difference in endorsement of the CSC impact rates of depressive diagnosis? Next, we explored race differences in depressive symptoms and specific diagnostic criteria to examine clinical presentation as a contributing factor in race discrepancies. The second research question was: Among older women with subthreshold and diagnosed depression, are there differences in the depressive symptom presentations of black and white older women that may underlie race-based discrepancies in diagnosis and CSC endorsement? In a final set of analyses, we sought to examine health and disability factors as potentially confounding variables underlying the race differences in CSC endorsement. There are known race-based disparities in health and disability among women (Belgrave and Abrams, 2016) that could impact the level of impairment reflected in the CSC. Thus, our third research question was: Are there race differences in dimensions of health and disability that could explain the lower rate of CSC endorsement in black older adults? Together, these three related research questions serve to extend previous work and more deeply explore race discrepancies in depression diagnosis and CSC endorsement.


Sample determination

Data were from a community-based, longitudinal study of the impact of social support and coping on depressive symptoms in lower-income older women. The sample, drawn from a large Midwestern city, included community-dwelling white and black women aged ≥ 65 years. The original study restricted recruitment to women, based on known gender differences in social relationships and coping (Blalock and Joiner, 2000) and because depressive symptoms are more prevalent among women (Callahan and Wolinsky, 1994; Kessler et al., 1993). As described elsewhere (Heller et al., 2010; Swindle et al., 2001), the recruitment procedure was designed to locate a diverse sample of older women with depression, who may not have sought treatment and would therefore be underrepresented in clinical settings and may also be unknown to other usual sources of convenience samples of older adults such as senior centers or service agencies. To reach lower-income women, recruitment occurred in census districts with households assessed below the city’s median home value and began with random telephone calling within those districts. Census-based residential data has been shown to be a valid measure of socioeconomic status (Kind et al., 2014; Watson et al., 2012). Persons residing in care facilities were excluded. During the initial telephone contact, all women age ≥ 65 were asked for consent to mail project information. As seen in Figure 1, follow-up telephone screening was completed within one week with 66% (N=1,963) of the older women contacted. There were no race differences in rates of screening completion. Respondents who scored 2 or higher (N=758) on an abbreviated version of the Center for Epidemiological Studies-Depression Scale (CES-D; Andresen et al., 1994), administered during this screening, and a sample of non-depressed participants scoring 0 on this measure (N=99) were invited to continue participation in the study. In the original study, a cut point of 2 had 98% sensitivity and 23% specificity for diagnosis (Heller et al., 2010). The response rate was 50.4% for the primary recruitment process (i.e. women scoring ≥2 on the depression screening who completed the baseline home interview). In-home diagnostic interviews were scheduled within two weeks of the telephone screening. Of the 481 participants who completed this interview, 70 were excluded from the current analyses due to missing data on key variables (n=68) or due to significant cognitive deficits (n=2; see Figure 1). Additional exclusion criteria were current psychosis and substance use disorder; however, no participants were excluded for these reasons. The final analytic sample was 411 (n=186 Black; n=225 White).

Figure 1. Participant recruitment process of the primary study and determination of analytic sample (N=411) for the current study. Short form CES-D (sfCES-D) refers to a 10-item adaptation of the Center for Epidemiologic Studies-Depression Scale (Andresen et al., 1994).

Data collection procedures

All measures used in the current study were collected during the in-home interview. Assessment instruments were administered orally. In virtually all cases, interviews were completed by race-matched interviewers. Most interviewers had completed a professional degree, and all had relevant prior professional experience. Prior to the start of data collection, interviewers completed extensive diagnostic training. They subsequently received regular individual and group supervision from a clinical psychologist with professional training in the diagnostic assessment instrument, the Structured Clinical Interview for DSM-IV, Clinician Version (SCID; First et al., 1996). Interviews were audiotaped in full and reviewed by the supervising psychologist for accuracy. If needed for clinical certainty, additional information or clarification was obtained from participants in a follow-up telephone interview. For diagnoses of MDD and dysthymia, a test of interrater reliability in a subsample of participants (n=24) revealed moderate to substantial agreement (κ = .62).


Demographic variables included age, race (Black or White), marital status (married or not married), and educational attainment (less than high school diploma, high school diploma, or more than high school diploma), which were obtained via self-report.

Assessment of depression and depressive impairment

Clinical significance criterion (depressive role impairment). The CSC was assessed immediately following depression symptom evaluation with the query, “Has it been hard for you to do your work, take care of things at home, or get along with other people?” The interview protocol included additional suggested prompts: e.g. “Have you been affected at all in your capacity to do your work and other activities?” and “What have you actually been doing in work, housework, hobbies, and interests and social life?” In accordance with the SCID protocol, interviewers were trained to use the core query with additional questioning, as needed, and to obtain enough information to complete the clinical assessment.

Depression diagnosis and subthreshold symptoms. The depression module of the SCID (First et al., 1996) was used to determine diagnoses of current MDD and dysthymia using core diagnostic criteria for each disorder (criteria are listed in Table S1). This semi-structured interview provides queries and additional optional prompts for each DSM-IV symptom and diagnostic criterion. The case identification procedure excluded persons with a history of manic episode. Bereavement was not a reason for exclusion.

The 20-item CES-D scale (Radloff, 1977) was used to assess depressive symptoms (α = .86). Each item is rated on a 4-point scale (0–3) of frequency of occurrence over the past week. The standard cut point indicating significant depression is 16 of 60 points. CES-D items were assigned to one of four subscales based on a factor structure confirmed in samples of older adults (Gatz et al., 1993) and in non-white participants (Kim et al., 2011): Depressed Mood (e.g. felt depressed; life was a failure); Somatic Complaints (e.g. everything was an effort; restless sleep; poor appetite); Interpersonal Difficulties (people were unfriendly; people disliked me); and Well-being (reverse-coded; e.g. felt happy; enjoyed life). Subscale means were computed for this study.

Based on SCID and CES-D data, three depressive status groups were created. The diagnosed depression group (DiagD; N=50) comprised participants who met diagnostic criteria for MDD or dysthymia (for the four Whites and three Blacks who met criteria for both, MDD was considered the diagnosis of record). The subthreshold depressed group (SubD; N=65) included participants with CES-D scores ≥16 but who did not meet criteria for depressive diagnosis. Participants with CES-D scores <16 and no diagnosis were placed in the non-depressed comparison group (N=296).

Health and disability factors

Mental health treatment. Lifetime history of mental health treatment was assessed with the question: “Have you ever seen anybody for nerves or psychiatric problems?” A proxy measure of current mental health treatment was obtained with the presence or absence of current antidepressant medication use. Prescription medication use was ascertained via direct observation of medication containers and categorized using a standard drug classification system (American Society of Health-System Pharmacists, 1996) .

Physical health. The Health and Daily Living Form (Moos et al., 1988) assessed medical burden in the past year through a count of medical conditions (possible score 0-14). This measure has previously been used with white and black samples (e.g., Swindle et al., 1989).

Cognitive status. The Mini-Mental Status Examination (MMSE; Folstein et al., 1975) was used to assess global cognition. Mean scores were not adjusted by race, consistent with previous approaches (Tappen et al., 2012), as race-based differences may be most related to education and other socioeconomic factors.

Functional impairment in activities of daily living (ADL). The Older Americans Resources and Services (OARS) Activities of Daily Living Scale (Fillenbaum, 1988) assessed the ability to complete 12 daily tasks such as eating, bathing, and managing medication. This instrument has been widely used, including with older Blacks (Long et al., 1998). ADL total possible scores range from 0 to 24, with higher scores indicating greater functional independence.

Statistical analyses

Bivariate comparisons were performed using tests appropriate for data type (Fisher’s exact test for categorical variables, e.g. presence of depressive impairment; Student’s t-test for continuous variables, e.g. CES-D score). For multiple group comparisons of depressive symptoms and CES-D total scale and subscales, we controlled the false discovery rate through adjustment of the minimum p-value criterion required for statistical significance (reduced from p = .05; Benjamini and Hochberg, 1995). We report adjusted criterion levels. Measures of effect size (Cohen’s d, odds ratio) were calculated for group comparisons, and Pearson point-biserial correlation analyses were used to examine the strength of associations between CSC endorsement and variables of interest. Logistic regression analysis using a backward elimination stepwise likelihood ratio method was used to develop a multivariate model to predict CSC endorsement using race, age, MMSE score, CES-D score, ADL score, and medical conditions count. The backward elimination method is less likely to inappropriately exclude predictors involved in suppressor effects (Field, 2009). We followed the recommendation of Hosmer and Lemeshow (1989) and set conservative criteria for variable entry (p = 0.1) and removal (p = 0.15).


As can be seen in Table 1, there were no significant differences between Blacks and Whites on age or marital status. Black participants had significantly less education (Fisher’s exact p = .000). Depressive status group membership differed significantly by race (Fisher’s exact p = .004), with fewer Blacks than Whites meeting criteria for depressive diagnosis (DiagD; 6.5% and 16.9%, respectively) and more Blacks meeting criteria for SubD status (18.3% and 13.8%, respectively).

Table 1. Demographic characteristics and depressive status group assignment of the study sample (N = 411)


1 Information on marital status was missing for one Black participant, resulting in an analytic sample size of N = 410 for this variable.

**significant at p = .01; ***significant at p = .001.

Research Question #1: CSC endorsement (depressive role impairment)

Across the entire sample, endorsement of the CSC differed by race (11.8% of Black and 24.1% of Whites, Fisher’s exact p = .002). With White race as the reference group, the odds ratio (OR) of CSC endorsement = 0.42, 95% confidence interval (CI) [0.25, 0.73]. Within the SubD group, 14.7% of Blacks and 20% of Whites endorsed the CSC, which was a nonsignificant difference (Fisher’s exact p = .742), OR = 0.69, 95% CI [0.19, 2.5]. To examine the impact of endorsed depressive role impairment on the likelihood of receiving a diagnosis of MDD or dysthymia, we reassigned the depressive status category using only the symptom-based diagnostic criteria A and B (depressed mood or anhedonia and presence of ≥5 symptoms), disregarding the CSC. Under this rubric, 15 additional participants met criteria for DiagD status, of whom 13 were Black, and the statistical difference by race in DiagD status disappeared (p = .59). Nine of the 13 (69%) black participants had CES-D scores of at least 16, placing them initially in the SubD depressive status group.

Research Question #2: Depressive symptom presentations in subthreshold and diagnosed depression

We compared symptom presentations in black and white participants within the SubD and DiagD groups using data from both assessment approaches: the SCID clinical interview items and the CES-D self-report depression summary scale. Unadjusted analyses revealed few race differences in diagnostic criteria A and B and the nine depression symptoms from the SCID (see Table 2). Effect sizes were small, and no comparisons remained significant after adjustments to the significance criterion value to control the false discovery rate. Inspection of symptom frequencies revealed that Blacks with subthreshold depression endorsed a greater number of symptoms (M = 3.18, SD = 1.87 versus M = 2.0, SD = 1.41), with nine black SubD participants (26.5%) endorsing at least five symptoms, compared to just one (3.2%) white SubD participant. Compared to white participants with SubD, rates of endorsement were higher among Blacks for seven of the nine symptoms. These patterns were not seen among diagnosed depressed participants; in this depressive status group, differences in symptom endorsement were smaller and more equally distributed by race.

Table 2. Major depressive episode diagnostic criteria and depressive symptom frequencies compared by race, within subthreshold (N=65) and diagnosed (N=50) depressive status groups.

Notes. MDD = Major Depressive Disorder; SubD = subthreshold depression group; DiagD = diagnosed depression group. Percentages reflect participants endorsing symptoms within that racial group and depressive status group. Adjusted criterion level is the minimum significance criterion p-value after adjustment for multiple comparisons. Bold font indicates statistically significant unadjusted values at p = 05.

1 Diagnosed depression (DiagD) group is comprised of participants meeting criteria for current MDD, dysthymia, or both; thus not all participants met each diagnostic criterion.

2 CES-D refers to Center for Epidemiological Studies-Depression Scale. Four SubD and two DiagD participants were missing one item on the CES-D and were not included in total score comparisons.

3 Odds ratio reference group is White race; CI = 95% confidence interval, unadjusted.

On the CES-D, there were no race differences in total score or the Depressed Mood, Somatic Complaints, and Well-being subscales. Black participants in both SubD and DiagD groups had higher scores on the Interpersonal subscale. However, these differences did not remain statistically significant after adjustment for multiple analyses.

Research Question #3: Health and disability factors

Using the entire sample, we next examined race differences in potential contributors to CSC endorsement. We first tested associations between the CSC and these factors. Not surprisingly, participants with a history of mental health treatment were more likely to endorse the CSC (56.3% versus 43.7%, Fisher’s exact p = .000; OR = 3.32, 95% CI [1.96, 5.64]). CSC endorsement was not associated with ADL total score (r pb = ‐.066, p = .194) nor with MMSE total score (r pb = .080, p = .115), but was positively correlated with number of medical conditions (r pb = .115, p < .05). In comparisons by race, older black women were less likely to have a history of mental health treatment and to be currently prescribed antidepressant medication, an effect of medium size (Table 3). There were no race differences in the count of medical conditions, but black women reported slightly less independence in activities of daily living and had lower scores on the MMSE global cognition measure. These effect sizes were small, and differences in cognitive performance were within education- and race-adjusted normative values for these groups (Wood et al., 2006).

Table 3. Health and disability factors compared by race (N=411)


1 Analytic sample sizes vary due to missing data. When appropriate, the unequal variance (Welch) t-test was used.

2 Higher scores on the Activities of Daily Living total score indicate more functional independence.

*Significant at p=.05; ***Significant at p = .001. P-values are unadjusted. Effect sizes were calculated with odds ratios (OR) with unadjusted 95% confidence intervals (CI) and Cohen’s d.

Logistic regression predicting CSC endorsement

As a final step to examine the relative importance of physical health and disability in predicting CSC endorsement, a logistic regression model was tested. Model 1 included race, age, CES-D total score, MMSE, ADL score, and count of medical conditions as predictors (see Table 4). MMSE was used as a covariate as it correlated strongly with education, a commonly-used covariate (r = .418, p < .001), but more accurately represents current cognitive functioning. Variance inflation factors for each predictor were examined to assess for multicollinearity and were found acceptable (range 1.07– .14). Applying the backward elimination stepwise regression method using five iterations, the final model included only CES-D score and race. Interpretation of the odds ratios in this final model revealed that when depressive symptom severity is equivalent, older white women are over 2.5 times more likely to report depressive role impairment compared with older black women.

Table 4. Logistic regression predicting CSC endorsement (N=351)

Notes. Reference category is White race. Cases with missing data on any variable were excluded from the analysis. Variables excluded from the final model: age, MMSE, medical conditions, and ADL total score. CI = confidence interval, unadjusted.

*Significant at p = .05; **significant at p = .001. P-values are unadjusted.


In this community-dwelling population, older black women with depressive symptoms were less likely than their white counterparts to endorse the clinical significance criterion when depressive disorders were assessed using the SCID for DSM-IV. Differences in depressive symptom presentation, physical health, functional impairment, and cognitive status did not explain the race-based discrepancies in CSC endorsement; in fact, black women were more likely to report dependence in activities of daily living. Importantly, the race difference in endorsement of the CSC appeared to fully account for the disparity in diagnosis rates.

These results are consistent with past epidemiological work suggesting that race differences in the CSC impact diagnostic assessment outcome. Coyne and Marcus (2006) found that Blacks were less likely than Whites to endorse the CSC (defined as either role impairment or need for mental health services), which resulted in race differences in diagnosis rates. Our data demonstrate that this race-based disparity can occur using the SCID interview with community-dwelling older adults. Our findings add to previous work suggesting that use of the CSC in determining diagnosis of depressive disorder should be reconsidered. Some scholars have argued that the CSC does not enhance diagnostic accuracy and represents an unnecessary redundancy in the diagnostic process (Wakefield et al., 2010). Beals and colleagues (2004) based on work conducted in a sample of Native Americans, concluded that the CSC creates an elevated risk of false negatives (i.e. “true” depression not meeting diagnostic criteria) within certain non-white cultures. In the aggregate, we believe that evidence is accumulating in support of a critical examination of depression assessment approaches across cultural subgroups. We cannot confirm that race differences in endorsement of depressive impairment contributed to past findings of lower rates of depressive diagnosis among Blacks (e.g. Blazer et al., 1987). However, our results are consistent with and may help to explain those findings. This should be a focus of future research.

Our findings are also consistent with past work demonstrating race differences in the strength of association between psychiatric symptoms and self-ratings of mental health (Jang et al., 2014) and between objective measures of health and subjective psychological burden (Pinquart and Sorensen, 2005). In general, these associations appear to be weaker among Blacks compared to Whites. Scholars have suggested that pressure to adhere to the “Strong Black Woman” identity can lead black women to deny health symptoms (Black and Peacock, 2011; Black and Woods-Giscombe, 2012). Ward and colleagues reported that older black women experience depression as “normal” or may be unaware of depressive symptoms and tend to focus on maintaining function as a primary coping strategy (Ward et al., 2014). The authors note that this coping approach may mask depressive symptoms by prompting black women to conceptualize and describe symptoms as an ordinary part of life. Other scholars, noting the relatively low participation of older Blacks in depression research studies, have argued that race differences in the view of depressive symptoms may contribute to this disparity (Steffens et al., 1997). Thus, our findings are consistent with past research and with the tenets of the common sense model, which holds that there is variability in beliefs about illness symptoms and that these beliefs ultimately underlie a person’s health appraisals and health behavior (Diefenbach and Leventhal, 1996).

Additionally, our findings of lower rates of antidepressant use among Black women are highly consistent with past studies (Fyffe et al., 2004; Gonzalez et al., 2010; Kales et al., 2013). Less use of medication to treat depression may reflect not only race-based disparities in treatment access, but also differences in treatment-seeking behavior based on illness perceptions and coping preferences.

Our findings add to the limited pool of research on subthreshold depression among black older women. In our exploratory analyses among older women with subthreshold depression, almost 27% of Blacks and less than 4% of Whites endorsed the ≥5 depressive symptoms needed for diagnosis (Criterion B; see Table S1). Blacks reported seven of the nine core symptoms more frequently than Whites. These subgroup analyses were conducted with a limited sample, and differences were not significant after statistical adjustment. Nevertheless, our results suggest that older black women with subthreshold depression may experience more severe pathology than their white counterparts. If confirmed in future, adequately-powered studies, these findings may have significant public health implications for assessment and treatment, informing efforts to develop culturally appropriate assessment approaches to reduce disparities in mental health care access.

The paradoxical race differences in prevalence rates, described above among older adults, are also seen in younger samples (Breslau et al., 2005; Dunlop et al., 2003). The “Strong Black Woman” theory has been applied to younger black female samples (Black and Woods-Giscombe, 2012). Further, prior research demonstrating race-based differences in CSC endorsement was conducted in a general adult population (Coyne and Marcus, 2006). Thus, while we suspect that important age cohort differences in beliefs about depression and role impairment exist, we believe that similar race differences would emerge in studies of younger and middle-aged adults. This is an important area for future research.


Our study has several limitations to consider. First, the sample was restricted to older women, limiting the generalizability of results to other demographic groups. The “Strong Black Woman” hypothesis is conceptualized as being gender-specific (Black and Woods-Giscombe, 2012) and, thus, may have limited applicability, in its current form, to black men. Factors impacting race-based differences in diagnosis among men should be examined in future studies. As always, potential cohort effects should also be considered when interpreting our findings. Our participants may differ from other generations of older adults. Second, although the recruitment procedure of the original study sampled only low-income metropolitan neighborhoods to attain approximate income equality across participants, it remains possible that socioeconomic inequities contributed to race differences (Blazer et al., 1998). Socioeconomic disadvantage is an important factor in mental health and is associated with reduced access to treatment facilities, fewer providers, and more difficulty getting to treatment appointments. Future research should include careful assessment of disadvantage and its impact on assessment outcomes, independent of race. Further, we did not assess additional factors that can have important influences on mental health among people of color in the U.S.A., including level of acculturation, country of origin, or immigration status (Williams et al., 2007). However, given the demographics of this Midwest urban area, it is unlikely that heterogeneity in these factors impacted results. Third, it is possible that race differences in survival affected our results. In other words, black women in this study may be those who have already effectively negotiated hardships to survive to late life and, thus, represent the hardiest and most resilient of their racial group. Ward and colleagues (2014) found high rates of resilience among older black women, which was used in normalizing depressive symptoms in that cohort. High resilience may be associated with less role impairment in the context of depressive symptoms, regardless of race or culture. Future examinations of the association between CSC endorsement and resilience in the context of mental disorder would likely be a useful addition to the literature informing clinical and research use of the CSC.

Finally, it is possible that racial differences among interviewers, who were matched by race to participants, contributed to differences in assessment outcomes. We did not directly assess race-based clinician bias, which is a limitation of our study. Clinician bias can impact diagnostic outcomes, especially when “subjective” judgement is required to draw conclusions. However, it is more likely that the approach of race-matching interviewers – considered best practice in social science research – improved the validity of our measures by supporting increased trust, engagement, and disclosure by black participants (Davis et al., 2013). Combined with the extensive training and supervision provided to interviewers, we believe that unmeasured rater effects were minimized and did not impact the primary results of the current study.


The current study sought to add to our understanding of race differences in endorsement of the clinical significance criterion included in DSM-IV and DSM-5 diagnostic criteria for depressive disorder and to examine the effect of these race differences on rates of diagnosis. We evaluated race differences in potential contributors to CSC endorsement. Finally, we sought to explore subthreshold depressive symptomatology in white and black older women in a community-dwelling sample. Black older women were less likely to endorse the CSC compared to Whites, unrelated to severity of depressive symptoms, medical comorbidity, and functional deficits. Use of the DSM-IV and DSM-5 clinical significance criterion may lead to underdiagnosis of depression among black older adults. Furthermore, subthreshold depressive symptomatology among black older adults may be more severe compared to Whites and, thus, may require culturally tailored assessment and treatment approaches. Our findings represent a significant contribution to the literature on race differences in depression assessment and underscore the urgent need for future research on mental health among older blacks.

Conflicts of interest

The authors have no conflicts of interest to declare.

Description of the authors’ roles

M. Wyman assisted with data collection for the original study, designed the current study, analyzed the data, and drafted the manuscript. E. Jonaitis completed initial dataset preparation, consulted on statistical analyses, and assisted with manuscript preparation. E. Ward, M. Zuelsdorff, and C. Gleason assisted with study design and manuscript preparation. All authors approved of the final publication.


This work was supported by the NIMH under grant #R01 MH049086 (K. Heller, PI); by the NIA under grant #R01 AG054059 (C. Gleason, PI); and by the VA Advanced Fellowship in Women’s Health (M. Wyman). GRECC # 005-2018. This material is in part the result of work supported with resources and the use of facilities at the W.S. Middleton Memorial Veterans Hospital, Madison, Wisconsin, U.S.A. We gratefully acknowledge the key contributions of Kenneth Heller, Ph.D., and Ralph Swindle, Ph.D., in design and data collection for the original study.

Supplementary material

To view supplementary material for this article, please visit


Adams, K. B. and Moon, H. (2009). Subthreshold depression: characteristics and risk factors among vulnerable elders. Aging & Mental Health, 13, 682692. doi: 10.1080/13607860902774501
Adebimpe, V. R., Hedlund, J. L., Cho, D. W. and Wood, J. B. (1982). Symptomatology of depression in black and white patients. Journal of the National Medical Association, 74, 185190.
American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders: DSM-IV (4th ed.). Washington, DC: American Psychiatric Association.
American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders: DSM 5 (5th ed.). Arlington, VA: American Psychiatric Publishing.
American Society of Health-System Pharmacists. (1996). American hospital formulary service drug information. Bethesda, MD: By the authority of the Board of Directors of the American Society of Hospital Pharmacists.
Andresen, E. M., Malmgren, J. A., Carter, W. B. and Patrick, D. L. (1994). Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). American Journal of Preventive Medicine, 10, 7784.
Ayalon, L. and Young, M. A. (2003). A comparison of depressive symptoms in African Americans and Caucasian Americans. Journal of Cross-Cultural Psychology, 34, 111124. doi: 10.1177/0022022102239158
Azar, A. R., Murrell, S. A. and Mast, B. T. (2005). Race and vascular depression risk in community-dwelling older adults. The American Journal of Geriatric Psychiatry, 13, 329332. doi: 10.1176/appi.ajgp.13.4.329
Beals, J. et al. (2004). Challenges in operationalizing the DSM-IV clinical significance criterion. Archives Of General Psychiatry, 61, 11971207. doi: 10.1001/archpsyc.61.12.1197
Belgrave, F. Z. and Abrams, J. A. (2016). Reducing disparities and achieving equity in African American women’s health. American Psychologist, 71, 723733. doi: 10.1037/amp0000081
Bell, C. C. and McBride, D. F. (2011). A commentary for furthering cultural sensitivity within research in geriatric psychiatry. The American Journal of Geriatric Psychiatry, 19, 397402. doi: 10.1097/JGP.0b013e318214ad43
Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological), 57, 289300.
Black, A. R. and Peacock, N. (2011). Pleasing the masses: messages for daily life management in African American women’s popular media sources. American Journal of Public Health, 101, 144150. doi: 10.2105/ajph.2009.167817
Black, A. R. and Woods-Giscombe, C. (2012). Applying the stress and ‘Strength’ hypothesis to Black women’s breast cancer screening delays. Stress Health, 28, 389396. doi: 10.1002/smi.2464
Blalock, A. J. and Joiner, T. E. Jr. (2000). Interaction of cognitive avoidance coping and stress in predicting depression/anxiety. Cognitive Therapy and Research, 24, 4765.
Blazer, D. G., Hughes, D. C. and George, L. K. (1987). The epidemiology of depression in an elderly community population. Gerontologist, 27, 281287. doi: 10.1093/geront/27.3.281
Blazer, D. G., Landerman, L. R., Hays, J. C., Simonsick, E. M. and Saunders, W. B. (1998). Symptoms of depression among community-dwelling elderly African-American and white older adults. Psychological Medicine, 28, 13111320.
Breslau, J., Kendler, K. S., Su, M., Gaxiola-Aguilar, S. and Kessler, R. C. (2005). Lifetime risk and persistence of psychiatric disorders across ethnic groups in the United States. Psychological Medicine, 35, 317327.
Callahan, C. M. and Wolinsky, F. D. (1994). The effect of gender and race on the measurement properties of the CES-D in older adults. Medical Care, 32, 341356.
Chuan, S. K., Kumar, R., Matthew, N., Heok, K. E. and Pin, N. T. (2008). Subsyndromal depression in old age: clinical significance and impact in a multi-ethnic community sample of elderly Singaporeans. International Psychogeriatrics, 20, 188200. doi: 10.1017/S1041610207006187
Conner, K. O. et al. (2010). Barriers to treatment and culturally endorsed coping strategies among depressed African-American older adults. Aging & Mental Health, 14, 971983. doi: 10.1080/13607863.2010.501061
Coyne, J. C. and Marcus, S. C. (2006). Health disparities in care for depression possibly obscured by the clinical significance criterion. The American Journal of Psychiatry, 163, 15771579. doi: 10.1176/ajp.2006.163.9.1577
Davis, R. E. et al. (2013). Ethnic identity, questionnaire content, and the dilemma of race matching in surveys of African Americans by African American interviewers. Field Methods, 25, 142161. doi: 10.1177/1525822x12449709
Diefenbach, M. A. and Leventhal, H. (1996). The common-sense model of illness representation: theoretical and practical considerations. Journal of Social Distress and the Homeless, 5, 11.
Dunlop, D. D., Song, J., Lyons, J. S., Manheim, L. M. and Chang, R. W. (2003). Racial/ethnic differences in rates of depression among preretirement adults. American Journal of Public Health, 93, 19451952.
Field, A. (2009). Discovering Statistics using SPSS (3rd ed.). London: SAGE Publications Ltd.
Fillenbaum, G. G. (1988). Multidimensional functional assessment of older adults: The Duke Older Americans Resources and Services procedures. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
First, M. B., Spitzer, R. L., Gibbon, M. and Williams, J. B. W. (1996). Structured Clinical Interview for DSM-IV Axis I Disorders, Clinician Version (SCID-CV). Washington, DC: American Psychiatric Press, Inc.
Folstein, M. F., Folstein, S. E. and McHugh, P. R. (1975). “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189198.
Ford, B. C., Bullard, K. M., Taylor, R. J., Toler, A. K., Neighbors, H. W. and Jackson, J. S. (2007). Lifetime and 12-month prevalence of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition disorders among older African Americans: findings from the National Survey of American Life. American Journal of Geriatric Psychiatry, 15, 652659. doi: 10.1097/JGP.0b013e3180437d9e
Fyffe, D. C., Sirey, J. A., Heo, M. and Bruce, M. L. (2004). Late-life depression among black and white elderly homecare patients. The American Journal of Geriatric Psychiatry, 12, 531535. doi: 10.1176/appi.ajgp.12.5.531
Gatz, M., Johansson, B., Pedersen, N., Berg, S. and Reynolds, C. (1993). A cross-national self-report measure of depressive symptomatology. International Psychogeriatrics, 5, 147156. doi: 10.1017/S1041610293001486
Gonzalez, H. M., Vega, W. A., Williams, D. R., Tarraf, W., West, B. T. and Neighbors, H. W. (2010). Depression care in the United States: too little for too few. Archives Of General Psychiatry, 67, 3746. doi: 10.1001/archgenpsychiatry.2009.168
Haigh, E. A. P., Bogucki, O. E., Sigmon, S. T. and Blazer, D. G. (2017). Depression among older adults: a 20-year update on five common myths and misconceptions. The American Journal of Geriatric Psychiatry. doi: 10.1016/j.jagp.2017.06.011
Heller, K., Viken, R. J. and Swindle, R. W. Jr. (2010). Screening for depression in African American and Caucasian older women. Aging & Mental Health, 14, 339346.
Hosmer, D. W. and Lemeshow, S. (1989). Applied Logistic Regression. New York: John Wiley & Sons.
Jang, Y., Park, N. S., Kang, S. Y. and Chiriboga, D. A. (2014). Racial/ethnic differences in the association between symptoms of depression and self-rated mental health among older adults. Community Mental Health Journal, 50, 325330. doi: 10.1007/s10597-013-9642-2
Jimenez, D. E., Alegria, M., Chen, C. N., Chan, D. and Laderman, M. (2010). Prevalence of psychiatric illnesses in older ethnic minority adults. Journal of the American Geriatrics Society, 58, 256264. doi:10.1111/j.1532-5415.2009.02685.x
Jimenez, D. E., Cook, B., Bartels, S. J. and Alegria, M. (2013). Disparities in mental health service use of racial and ethnic minority elderly adults. Journal of the American Geriatrics Society, 61, 1825. doi: 10.1111/jgs.12063
Judd, L. L., Rapaport, M. H., Paulus, M. P. and Brown, J. L. (1994). Subsyndromal symptomatic depression: a new mood disorder? The Journal of Clinical Psychiatry, 55(Suppl), 1828.
Kales, H. C. et al. (2013). Racial differences in adherence to antidepressant treatment in later life. American Journal of Geriatric Psychiatry, 21, 9991009. doi: 10.1016/j.jagp.2013.01.046
Kessler, R. C., McGonagle, K. A., Swartz, M., Blazer, D. G. and Nelson, C. B. (1993). Sex and depression in the National Comorbidity Survey. I: lifetime prevalence, chronicity and recurrence. Journal of Affective Disorders, 29, 8596.
Kim, G., Decoster, J., Huang, C. H. and Chiriboga, D. A. (2011). Race/ethnicity and the factor structure of the center for epidemiologic studies depression scale: a meta-analysis. Cultural Diversity and Ethnic Minority Psychology, 17, 381396. doi: 10.1037/a0025434
Kind, A. J. et al. (2014). Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Annals of Internal Medicine, 161, 765774. doi: 10.7326/m13-2946
Long, K., Sudha, S. and Mutran, E. J. (1998). Elder-proxy agreement concerning the functional status and medical history of the older person: the impact of caregiver burden and depressive symptomatology. Journal of the American Geriatrics Society, 46, 11031111.
Meeks, T. W., Vahia, I. V., Lavretsky, H., Kulkarni, G. and Jeste, D. V. (2011). A tune in “a minor” can “b major”: a review of epidemiology, illness course, and public health implications of subthreshold depression in older adults. Journal of Affective Disorders, 129, 126142. doi: 10.1016/j.jad.2010.09.015
Moos, R., Cronkite, R., Billings, A. and Finney, J. (1988). Health and Daily Living Form Manual. Palo Alto, CA: Social Ecology Laboratory, Dept of Psychiatry & Behavioral Sciences, Stanford University Medical Center.
Pinquart, M. and Sorensen, S. (2005). Ethnic differences in stressors, resources, and psychological outcomes of family caregiving: a meta-analysis. Gerontologist, 45, 90106.
Radloff, L. S. (1977). The CES-D Scale: a self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385401. doi: 10.1177/014662167700100306
Skarupski, K. A. et al. (2005). Black-white differences in depressive symptoms among older adults over time. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 60, P136142.
Spitzer, R. L. and Wakefield, J. C. (1999). DSM-IV diagnostic criterion for clinical significance: does it help solve the false positives problem? American Journal of Psychiatry, 156, 18561864.
Steffens, D. C., Artigues, D. L., Ornstein, K. A. and Krishnan, K. R. (1997). A review of racial differences in geriatric depression: implications for care and clinical research. Journal of the National Medical Association, 89, 731736.
Swindle, R. W., Cronkite, R. C. and Moos, R. (1989). Life stressors, social resources, coping, and the 4-year course of unipolar depression. Journal of Abnormal Psychology, 98, 468477.
Swindle, R. W., Heller, K., Alexander, D. B., Allen, S. M. and Wyman, M. F. (2001). Sources of stressor-specific negative transactions and depressive symptoms among White and African American older women. Applied Developmental Science, 5, 98111. doi: 10.1207/S1532480XADS0502_5
Tappen, R. M., Rosselli, M. and Engstrom, G. (2012). Use of the MC-FAQ and MMSE-FAQ in cognitive screening of older African Americans, Hispanic Americans, and European Americans. The American Journal of Geriatric Psychiatry, 20, 955962. doi: 10.1097/JGP.0b013e31825d0935
Wakefield, J. C., Schmitz, M. F. and Baer, J. C. (2010). Does the DSM-IV clinical significance criterion for major depression reduce false positives? Evidence from the National Comorbidity Survey Replication. The American Journal of Psychiatry, 167, 298304. doi: 10.1176/appi.ajp.2009.09040553
Ward, E. C., Mengesha, M. M. and Issa, F. (2014). Older African American women’s lived experiences with depression and coping behaviours. Journal of Psychiatric and Mental Health Nursing, 21, 4659. doi: 10.1111/jpm.12046
Watson, K. T., Roberts, N. M. and Saunders, M. R. (2012). Factors associated with anxiety and depression among African American and White women. ISRN Psychiatry, 2012, 432321. doi: 10.5402/2012/432321
Williams, D. R. 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. Archives of General Psychiatry, 64, 305315. doi: 10.1001/archpsyc.64.3.305
Wood, R. Y., Giuliano, K. K., Bignell, C. U. and Pritham, W. W. (2006). Assessing cognitive ability in research: use of MMSE with minority populations and elderly adults with low education levels. Journal of Gerontological Nursing, 32, 4554.
Woodward, A. T., Taylor, R. J., Bullard, K. M., Aranda, M. P., Lincoln, K. D. and Chatters, L. M. (2012). Prevalence of lifetime DSM-IV affective disorders among older African Americans, Black Caribbeans, Latinos, Asians and non‐Hispanic White people. International Journal of Geriatric Psychiatry, 27, 816.