Measurement suitability of the Center for Epidemiological Studies-Depression Scale among older urban black men.

This study investigated the measurement suitability of the Center for Epidemiological Studies-Depression Scale (CED-S) when it was administered to older urban black men. This study used two samples of older black men in the Harlem Men's Health Study (clinic, N = 404; community, N = 319). Three factors emerged in both the clinic and the community samples instead of four factors originally found in Radloff's study. Depressive and somatic factors merged as one factor instead of as separate distinct factors to explain about 31% of the variance in the clinic sample and 37% of the variance in the community sample. Interpersonal distress measured differently in the two Harlem samples. The "crying" item (depressive affect item) loaded as interpersonal distress in both the clinic and the community samples. These results showed that the four-factor model of CES-D may not be applicable to older urban black men; therefore, caution should be exercised when interpreting CES-D results on populations of older black men.

Keywords: older black men, measurement suitability, factor analyses, depression, depression scale, Center for Epidemiological Studies-Depression Scale (CES-D)

It has been stated that a survey instrument that purports to measure depression or any other mental health disorder must assess the actual symptoms in individuals and groups (Foley, Reed, Mutran, & DeVellis, 2002). Many of these survey instruments have been validated on middle-aged white populations and are then often applied to minority populations across the lifespan. However, social and cultural differences in minority populations as well as gender differences may lead to differential measures of symptoms, either underestimating or overestimating a mental health disorder such as depression. As the population of the United States is getting older, more information about the psychometric properties of survey instruments when applied to minority populations is needed to better serve the growing older minority population.

CENTER FOR EPIDEMIOLOGICAL STUDIES-DEPRESSION SCALE (CES-D)

The Center for Epidemiological Studies-Depression Scale (CES-D) is a 20-item, self-reported instrument used for screening depressive symptomatology (Radloff, 1977). An analysis of the CES-D has usually revealed four factors: somatic symptoms, depressive affect, positive affect, and interpersonal behavior (Berkman et al., 1986; Clark, Aneshensel, Frerichs, & Morgan, 1981; Gatz, Johanasson, Pedersen, Berg, & Reynolds, 1993; Hertzog, Van Alstine, Usala, Hutsch, & Dixon, 1990; Radloff, 1977; Ross & Mirowski, 1984; Sheehan, Fifield, Reisine, & Tennen, 1995). Studies show that the CES-D has a high internal consistency of .85 (Blazer, Landermann, Hays, Simonsick, & Sauders, 1998; Radloff, 1977; Roberts, 1980; Roberts, Vemon,& Rhoades, 1989). Previous studies used principal component factor analysis with normal Varimax rotation to extract four factors using Eigenvalues greater than 1, with each item loading above .40 (Foley et al., 2002; Kim & Mueller, 1978; Radloff, 1977). Although studies have shown that the CES-D is reliable and valid, the utility and psychometric properties of it may be inconsistent when it is administered to an older male minority population.

CES-D AND MINORITY ELDERS

Several studies of older minorities have raised some questions about the CES-D's standard four-factor structure. For instance, in Chapleski, Lamphere, Kaezynski, Lichtenberg, and Dwyer's (1997) study, a four-factor structure was found not to be applicable to American Indian elders (N = 126), while the 12-item CES-D appeared better suited than the 20-item CES-D. Similarly, another study of American Indian elders (N = 120) showed a different factor structure from the original four-factor structure, with no clear distinction between depressive affect and somatic factors (Somervell, Beals, Kinzie, Boehnlein, Leung, & Manson, 1993). Exploratory factor analyses were used for both studies. We must remain cautious, though, due to the small sample sizes of both studies, suggesting that the results may not generalize to all American Indian elders. In addition, the combined depressed and somatic factor was seen with the Hispanic NHANES study of 3,117 Mexican American adults (Guarnaccia, Angel, & Worobey, 1989), with 2,536 Hispanic elders (Miller, Markides, & Blacks, 1997), and with 227 older black men and women (Foley et al., 2002). Further, in Miller et al., only two factors, depression and well-being, were found using both exploratory and confirmatory factor analyses. Studies of minority populations and CES-D have shown that there is no distinction between depressive and somatic symptoms (Baron, Manson, Ackerson, & Brenneman, 1990; Garcia & Marks, 1989; Guarnaccia et al., 1989; Kuo, 1984; Somervell et al., 1993). Large-scale studies such as those by Guarnaccia et al. and Miller et al. that can be generalized to the particular minority group are needed.

CES-D AND GENDER DIFFERENCES

The psychometric properties of the CES-D are different when gender and race are taken into consideration. Callahan and Wolinsky (1994) showed an obscure seven-factor structure among black men, with three new factors for anxiety, introspection, and crying. In the same study, a four-factor structure emerged when CES-D was applied to white and black women; however, two items, "mind" and "failure," did not load onto any factor for black women. In addition, Foley et al. (2002), who studied older blacks, showed a unique fourth factor, "social well-being." Moreover, blacks were significantly more likely than whites to respond affirmatively on two interpersonal questions, "people are unfriendly" and "people dislike me" (Cole, Kawachi, Mailer, & Berkman, 2000). Cole et al. also found that "crying spell" was gender-biased, because fewer male subjects endorsed this item than the other items on CES-D.

The low prevalence of depression observed among men may be due to multiple factors, including sex-role socialization, under-representation of men in mood disorder studies, the effects of aging on men's health, study instrumentation, and culture or minority status. Sex-based differences in behavior appear to be influenced by socially prescribed roles for both men and women. Specifically, the masculine and feminine attributes may influence the way men and women express their depression (Cockerham, 1999). It may be more acceptable for women to be open about depressive symptoms, while for men an expression of similar symptoms could be perceived as a sign of personal weakness. Aspects of the traditional male gender role, as traditionally constructed, discourage both awareness and expression of psychic pain. At the same time, these traditional roles actually contribute to pain by prompting men to follow a path that rewards emotional restriction and denial of psychological pain (Cochran & Rabinowitz, 2001). Men may therefore falsely appear to experience lower observed prevalence of depression than their female counterparts.

Because depression appears to manifest itself differently in men than in women and the risk of depression is high in older minority persons, there is a need to examine depression among older black men. Depression is not simply having the blues or having transitory feelings of sadness. It often involves continual pain, suffering, and dysfunction (Hortwitz & Scheid, 1999). The calculated lifetime risk for a diagnosis of Major Depressive Disorder in the community has varied from 10% to 25% for women and from 5% to 12% for men (American Psychiatric Association, 1994). Within the emerging discourse on the psychology of men, more interest should be devoted to articulating the problem of unrecognized and untreated depression in men.

DEPRESSION IN URBAN BLACK MEN

Studies have shown that blacks are under-diagnosed for affective disorders such as depression, while they are over-diagnosed for schizophrenia (Baker & Bell, 1999; Chung, Mahler, & Kakumat, 1995; Williams, 1995). During the 1970s black men were labeled "an endangered species" (Braithwaite, 2001), a label that thirty-plus years later can still be applied as many black men face serious health, sociopolitical, and psychological concerns. In every age group of blacks up to age 65 years and over, black men experience higher mortality rates than do any other racial or ethnic subgroups. For instance, the mortality rate for heart disease in black men is 1.5 times greater than that of black women and two times greater than that of white men (Courtenay, 2000). About 7% of black men are expected to develop depression during their lifetime, yet this is most likely an underestimate of the true incidence of depression in this group (American Psychiatric Association, 1994). Numerous psychological and sociological hazards exist in the urban black community, such as lower socioeconomic status, lack of perceived control, personal discouragement, discrimination, and thwarted aspirations and expectations of a successful life (Williams & Williams-Morris, 2000).

Located in upper Manhattan (New York City) north of Central Park, Harlem is where black culture developed in the 1920s, and today Harlem is still a black cultural center (McCord & Freeman, 1997). Widespread poverty and inadequate housing are prominent in Harlem (Wilson, 1996). Unemployment rates are high, as are homelessness, crime, and substance abuse among Harlem's inhabitants. Further, in Harlem, witnessing someone seriously injured or murdered is frequently reported (Fullilove et al., 1999). Use of illegal drugs among people in Harlem is two to 20 times higher compared to New York City residents overall.

In general, blacks living in impoverished urban areas suffer greatly from poor physical health and chronic health conditions (Geronimus, 1996). Those who live in public housing are more likely to be depressed (Black, Rabins, German, Roca, McGuire, & Brant, 1998). Communities of those with lower socioeconomic status are unable to provide adequate quality health services.

Given the above, a focus on older black men living in Harlem permits an increase in understanding of the poorer health status of this group. In this present study, we investigate the suitability of the traditional four-factor model of CES-D when applied to samples of older urban black men.

METHOD

PARTICIPANTS

The data for the Harlem Men's Health Cohort Study were collected between February 1995 and June 1996. The eligibility criteria for inclusion in this cohort required (1) residence in the 11 zip code areas that defined central Harlem, (2) being 50 to 74 years of age, (3) being fluent in English, (4) having no reported treatment for cancer in the prior year, and (5) having no reported lifetime diagnosis or treatment for prostate cancer. With the high prevalence of non-telephone households in Harlem (23.9% in the recent Harlem Health Survey), men were enrolled into the study from clinic settings to allow representation of men without telephones (17.5% reported not having telephones). A second sample of community-based men was identified using random-digit dialing (RDD) telephone methods. The New York State Department of Health and the Harlem Hospital Center institutional review board reviewed and approved the study (Ashford et al., 2001).

A clinic-based sample of 404 men was recruited from general medical and urology clinics in central Harlem, thus covering all providers in this geographic area. These clinics included medical and medical subspecialties: Harlem Hospital Center, North General Hospital, Robeson Clinic, and Renaissance Health Care Network, which consists of five community-based ambulatory care centers. To capture a representative volume of service use, clinic recruitment sessions were selected. At each site, trained project interviewers approached all the adult males present at selected clinic sessions and attempted to recruit all the clinic patients. If a subject met the study's criteria, informed consent was obtained and an interview took place in a private room. Contact information was obtained, and interviews were conducted within 1 to 2 days (usually at patients' homes) for the 10% of the patients who preferred another time or location for the interview, usually because the patient had other appointments scheduled. A private firm enrolled 319 men in the community sample using a random-digit dialing (RDD) telephone approach. The firm called numbers within all telephone exchanges in the catchment areas to identify potential respondents and then forwarded the names to the research team. If the respondents met the eligibility criteria and consented to participate, an in-person interview was scheduled and most often completed in the subject's home.

The in-person interview involved a structured, 60 to 90-minute assessment of respondent demographics (age, education, living conditions, income), health care access (insurance coverage, types of care providers, number of physician visits during the prior year), and physical and mental health. Self-reported health measures included the Medical Outcome Study (MOS short form, SF-36 with some revisions). The project interviewers were black research assistants who received extensive training. A field monitor observed and oversaw all the project interviewers conducting interviews. Furthermore, a 10% random sample of respondents was contacted again by the project coordinator and interviewed on a select set of questions to ensure the reliability of the data.

ANALYSES

Descriptive statistics including the sociodemographic variables and self-reported health measures were analyzed for two different sample populations (a RDD community sample and a clinic sample) of the Harlem Men's Health Cohort. Prevalence estimates of depression were calculated separately for the community and the clinic samples. The resulting sum of scores of CES-D indicated the severity of the person's depressive symptoms experienced in the past week, with 0 being the lowest score an individual can obtain and 60 being the highest (Radloff, 1977). A total score of 16 or higher on the CES-D discriminated psychiatric patients from general population samples.

To analyze the psychometric properties of the CES-D, subjects with complete data on all CES-D items were included (community, n = 319; clinic, n = 404). The following analyses were performed on both samples separately: (1) total score of the CES-D, after reverse coding positively framed items; (2) item-by-item statistics; (3) calculation of internal reliability of total scale and item-total correlations using Cronbach's alpha; and (4) a two-stage exploratory factor analysis (EFA) using principal component with no rotation (Stage 1) and an orthogonal (Varimax) rotation (Stage 2) based on scree plot and Eigenvalues of Stage 1. Only factors with an Eigenvalue greater than 1 were considered. To identify the primary factor on which the items loaded, a factor loading of .40 or greater was used (Kim & Mueller, 1978). Any items that loaded onto more than one factor were indicated in the tables. Orthogonal rotation was used to maximize the variance of factor loading by marking high loadings higher and low ones lower for each factor (Tabachnick & Fidel, 2001). Because only a few studies have examined the factor structure of the CES-D among older black men and there are inconsistencies of the factor structure of the CES-D when CES-D is administered to a black population, EFA was used in this study in order not to force a suggested factor pattern, as it is in confirmatory factor analysis (Foley et al., 2002; Kim & Mueller, 1978; Tabachnick & Fidel, 2001). EFA was conducted on the CES-D to determine how many latent variables underlie a set of items on the scale, to provide a way to explain the variation among many variables and to define the substantive content or latent variables that account for the variation in a scale (DeVellis, 1991).

RESULTS

SAMPLES CHARACTERISTICS

Table 1 shows the demographic information on the community and the clinic samples of the Harlem Men's Health Cohort Study. The age distribution and years of living in Harlem were equivalent. Mean ages of the community and clinic samples were 61.4 years old and 60.4 years old, respectively. Mean years of living in Harlem for the community sample was 35.8 years, and the community sample was 36.0 years. The clinic and the community samples differed significantly in education, marital status, household income, employment status, Medicaid status, and type of residence. The community sample was more educated than the clinic sample ([chi square] = 12.01, p = .007). More people were married in the community sample than the clinic sample ([chi square] = 12.27, p = .015). About 82.6% of the subjects in the clinic made less than $15,000 annually, whereas about 65.9% of the subjects in the community made less than $15,000 ([chi square] = 37.61, p = .0001). The clinic sample had significantly more unemployed people ([chi square] = 33.90, p = .0001). More than half of the clinic sample was on Medicaid whereas only a third of the community sample was on Medicaid ([chi square] = 63.48, p = .0001). Although not shown on the table, there were significantly more people living in rented rooms, shelters, and on the streets in the clinic sample than in the community sample ([chi square] = 17.64, p = .001). Furthermore, 6.5% of the clinic sample reported that they were feeling very unsafe in the neighborhood, while about 2.5% of the community sample reported they were feeling very unsafe in the neighborhood ([chi square] = 17.25, p = .001). More subjects in the clinic sample reported having poor general health than those in the community sample ([chi square] = 39.92, p = .0001).

RELIABILITY AND ITEM-LEVEL ANALYSES

The mean CES-D scores for the community and clinic samples were 10.14 and 14.23, respectively, with a range of 0-49 in the community sample and 0-48 in the clinic sample (out of a possible 60). Almost 26% of the community and 41% of the clinic samples scored 16 or higher, which is commonly used as a "cut-off' for potential depression and need for additional screening ([chi square] = 15.65, p = .0001). Cronbach's alpha was high (ct = 0.91 or standardized et = 0.92 for community, and [alpha] = 0.87 or standardized [alpha] = 0.88 for clinic).

In Tables 2 and 3, item-by-item descriptive analyses of CES-D are shown. In the community sample, the responses were skewed toward less depressive symptomatology (Table 2). Approximately 72% of the community sample reported rare or no symptoms at all on these items. In the clinic sample, the responses were also skewed toward less symptomatology (Table 3). Approximately 61% of the clinic sample reported rare or no symptoms at all on these items.

FACTOR STRUCTURE OF CES-D

Due to the inconsistencies of the factor structure of CES-D when administered to a black population, EFA was used to analyze the factor structure of the CES-D on our samples. In Table 4, principal component factor analysis results of two Harlem samples are presented separately along with the original four-factor structure (Radloff, 1977). Textbooks on factor analysis argue that in addition to using the Eigenvalues greater than 1.0 rule, other indicators of factor structure should be examined such as item loadings, scree plots, and percentage of variance explained by each factor (Kim & Mueller, 1978; Tabachnick & Fidel, 2001).

In the community sample, three factors were found, and the scree plot showed that no more than three factors could be extracted. Thirteen of the CES-D items loaded together as the first factor (>.40). Six out of seven depressive items, six out of seven somatic items, and one out of two interpersonal distress items accounted for 39.42% of the variance in the model (Eigenvalue = 7.88). This factor was represented as "D/S" in Table 4 to show the combined depressive-somatic factor, since there was no distinct somatic or depressive factor. The second factor accounted for 6.98% of the variance (Eigenvalue = 1.40), which included one interpersonal item, "unfriendly," one depressive affect, "crying," and one somatic affect, "appetite." This second factor was noted in Table 4 as "I." Positive affect emerged almost as expected for the third factor, and three out of four positive items loaded together ("good," "hopeful," and "dislike"), explaining 5.68% of the variance (Eigenvalue = 1.14).

In the clinic sample, three factors emerged, and the shape of the scree plot supported this finding (Table 4). For the first factor, which included 13 of the CES-D items, six out of seven depressive items and all seven somatic items accounted for 32.75% of the variance in the model (Eigenvalue = 6.55). This factor is represented as "D/S" in Table 4 to indicate the combined depressive-somatic factor. There were no distinct "somatic" or "depressive" factors. The second factor included two interpersonal items, "unfriendly" and "dislike," as well as one depressive affect, "crying," and accounted for 7.86% of the variance (Eigenvalue = 1.57). "I" was used to indicate this second factor in Table 4. Positive affect emerged, as expected, as a third factor, with all four positive items loaded on a single factor, explaining 5.14% of the variance (Eigenvalue = 1.03).

DISCUSSION

Older black men living in Harlem appeared to have higher levels of depressive symptomatology when compared to older blacks in previous studies (Callahan & Wolinsky, 1994; Foley et al., 2001). About 26% of the community sample and 48% of the clinic sample scored 16 or higher on the CES-D compared to Foley et al.'s study, which showed that 14% of the older blacks (men and women) scored 16 or higher. This may be due to the widespread poverty and inadequate housing prominent in Harlem. As noted, high unemployment rates, homelessness, crime, and substance abuse are prevalent in poor neighborhoods such as Harlem (Wilson, 1996). Death rates for those between the ages of 5 and 65 are worse in Harlem than in Bangladesh (McCord & Freeman, 1997). The existence of a disproportionate amount of poverty will continue to have a profound influence upon family structure, health status, and overall well-being among older black men (Miles, 1999; Siegel, 1999).

The results of the factor analyses performed in this study showed that three factors emerged in both community and clinic samples instead of the four factors originally described in Radloff's study (1977) or the seven factors found among older black men living in a North Carolina community (Callahan & Wolinsky 1994). There was no distinction between depressive affect and somatic symptoms in our samples, as had been suggested by previous studies on minority populations (Baron et al., 1990; Foley et al., 2002; Garcia & Marks, 1989; Guarnaccia et al., 1989; Kuo, 1984; Somerville et al., 1993).

In our samples, the "crying" item, which was one of the depressive affect items in the study by Radloff (1977), loaded with the interpersonal distress items in both the community and clinic samples. This may be due to sex-based differences in the perception of socially acceptable behaviors for men and women. Since men and women express their depression differently due to what is socially accepted as typically masculine and feminine attributes (Cockerham, 1999), the "crying" item was not a reliable indicator of depressive symptomatology.

Items like "appetite" and "unfriendly" loaded differently in the two Harlem samples. "Appetite" loaded as an interpersonal factor in the community sample, whereas in the clinic sample "appetite" loaded as a somatic-depressive factor. "Unfriendly" loaded as a somatic-depressive factor in the community sample, while this item loaded as an interpersonal distress factor in the clinic sample. Differences in socioeconomic levels between the community and the clinic samples that we compared may have produced a discrepancy in comprehension or inconsistent response pattern, which led to unequal measures of depressive symptomatology. Although not shown in this study, a majority of the community sample attended church regularly and were an active part of the Harlem community. The black community is a tightly knit community that incorporates church as one of the main social support networks (Braithwaite & Taylor, 1992). Many meals and social interactions are provided in church settings. This may explain why the "appetite" item loaded as an interpersonal factor and the "unfriendly" item loaded as a somatic-depressive factor. Because the clinic sample had lower socioeconomic status and poorer overall health than the community sample, those individuals may be experiencing more hardship than those in the community sample. More than 60% of the clinic sample was on Medicaid, which may indicate that they have a difficult time acquiring food and being accepted in their own community. This may explain why the item "appetite" loaded as a somatic-depressive factor, whereas the item "unfriendly" loaded as an interpersonal distress item for this sample.

The results of this study indicated that reports of depressive symptomatology are measured differently in older urban black men. Our results confirmed Foley et al. (2002), who demonstrated that depressive symptomatology may have a different structure among older blacks. Furthermore, the heterogeneity of the black population was demonstrated by the differences between the community and clinic samples of black men living in Harlem. The men who were in the clinic sample scored higher on the depressive symptomatology measure than did the men who were in the community sample. Because this study was a cross-sectional study, it is impossible to conclude if the poor health of the clinic sample contributed to greater endorsement of depressive symptomatology than was seen in the community sample.

Our samples were collected from the Harlem area. Therefore, the results of this study should be interpreted with caution. These results may not be generalizable to all older urban black men living in the United States.

Many studies that utilized the CES-D have not typically reported sex differences in subscales. A few studies such as Clark et al. (1981) demonstrated that there are sex differences in item correlations and factor loadings. Their conclusion was that the CES-D may measure different phenomena in men and women. Men are more likely to endorse items related to concentration difficulties, work difficulty, or somatic items than items such as crying or feeling depressed. Similar results were seen in Callahan and Wolinsky (1994), who showed different factor loadings for men and women. This may account for the lack of distinction between depressive affect and somatic symptoms that we observed. In our society, it is less acceptable for men to express their depressive symptoms, because they may appear to be showing a sign of weakness. Men are supposed to endure pain and mental anguish (Cochran & Rabinowitz, 2001). Thus, lower rates of depression may be seen in men due to social acceptance of traditional gender roles.

Dispelling the stigma associated with depression is needed among older black men as well as younger black men in their communities. Many still perceive depression as a personal weakness or a character flaw (Rosenbaum, 1996). Furthermore, the racism and discrimination that black men experience during their lifetime may exacerbate depression in this group (Annandale, 2003). The lower health status of black elders is of a particular concern because of their rapid growth in numbers and their tendency to have low socioeconomic status (Ferraro, 1993; Ferraro & Farmer, 1996). Our sample of older black men in Harlem had poorer mental and physical health than a national sample composed of white and black men. Therefore, early intervention programs for the black elderly population will be needed to identify factors that contribute to depression and to dispel stigmas associated with depression in this group.

Finally, it is important for researchers and clinicians to evaluate how depression is measured and defined in older minority population subgroups. The differences found between subgroups themselves may be a product of the instruments used to measure depression. Some researchers feel that CES-D is not an effective instrument to detect depression among older blacks because older blacks are more likely to report somatic symptoms (Collen, 1995). For example, CES-D has a sensitivity of .71 to blacks and .85 to whites (Baker, Parker, Wiley, Velli, & Johnson, 1995). Our study has shown that the measurement suitability of CES-D should be questioned when CES-D is administered to older urban black men. Standard instrumentations used in assessing depression may contribute to masking the "true" prevalence of depression in men. A focus on minority older men and depression is warranted in the mental health field.

The authors would like to thank Dr. Steven Albert, University of Pittsburgh, for guidance, and Dr. Alfred Ashford, of the Harlem Hospital Center and Columbia-Presbyterian Medical Center, for giving us access to this population.

REFERENCES

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Press.

Annandale, E. (2003). A poor man's plight: Uncovering the disparity in men's health. International Journal of Men's Health, 2, 229.

Ashford, A.R., Albert, S.M., Hoke, G., Cushman, L.F., Miller, D.S., & Bassett, M. (2001). Prostate carcinoma knowledge, attitude, and screening behavior among African-American men in central Harlem, New York City. Cancer, 91, 164-172.

Baker, F.M., & Bell, C.C. (1999). Issues in psychiatric treatment of blacks. Psychiatric Services, 50, 362-368.

Baker, F.M., Parker, D.A., Wiley, C., Velli, S.A., & Johnson, J.T. (1995). Depressive symptoms in black medical patients. International Journal of Geriatric Psychiatry, 10, 9-14.

Baron, E.E., Manson, S.M., Ackerson, L.M., & Brenneman, D.L. (1990). Depressive symptomatology in older America Indians with chronic disease. In C. Atkisson & J. Zich (Eds.), Screening for depression in primary care (pp. 217-231). New York: Routledge, Kane and Company.

Berkman, L.F., Berkman, C.S., Kasl, S., Freeman, D.H., Jr, Leo, L., & Ostfeld, A.M. (1986). Depressive symptoms in relation to physical health and function in the elderly. American Journal of Epidemiology, 124, 372-388.

Black, B.S., Rabins, R.V., German, P., Roca, R., McGuire, M., & Brant, L.J. (1998). Use of formal and informal sources of mental health care among older African American public housing residence. Psychological Medicine, 28, 1311-1320.

Blazer, D.G., Landermann, L.R., Hays, J.C., Simonsick, E.M., & Sauders, W.B. (1998). Symptoms of depression among community dwelling elderly black and white older adults. Psychological Medicine, 28, 1311-1320.

Braithwaite, R.L. (2001). The health status of black man. In R.L. Braithwaite & S.E. Taylor (Eds.), Health issues in the black community (pp. 62-80). San Francisco: Jossey-Bass.

Braithwaite, R.L., & Taylor, S.E. (Eds.). (1992). Health issues in the black community. San Francisco: Jossey-Bass.

Callahan, C.M., & Wolinsky, F.D. (1994). The effects of gender and race on the measurement properties of CES-D in older adults. Medical Care, 32, 341-356.

Chapleski, E.E., Lamphere, J.K., Kaezynski, R., Lichtenberg, P.A., & Dwyer, J.W. (1997). Structure of depression measure among American Indian elders: Confirmatory factor analysis of the CES-D scale. Research on Aging, 19, 462-485.

Chung, H., Mahler, J.C., & Kakumat, T. (1995). Racial differences in treatment of psychiatric inpatients. Psychiatric Services, 46, 586-591.

Clark, V.A., Aneshensel, C.S., Frerichs, R.R., & Morgan, T.M. (1981). Analysis of effects of sex and age in response to items of GES-D scale. Psychiatry Research, 5, 171-181.

Cochran, S.V., & Rabinowitz, F.E. (2001). Men and depression: Clinical and empirical perspectives. San Diego: Academic Press.

Cockerham, W. (1999). Mental disorder: Age, gender, and marital status. In C.S. Aneshensel & J.C. Phelan (Eds.), Handbook of sociology of mental health (pp. 154-175). New York: Kluwer Academic Plenum Publishers.

Cole, S.R., Kawachi, I., Mailer, S.J., & Berkman, L.F. (2000). Test of item-response bias in the CES-D scale: Experience from the New Haven EPESE Study. Journal of Clinical Epidemiology, 53, 285-289.

Collen, C.A. (1995). Mortality outlook. An overview of black health. Black research perspectives. Ann Arbor: Institute of Social Research, University of Michigan.

Courtenay, W.H. (2000). Constructions of the masculinity and their influences on men's well being: A theory of gender and health. Social Science Medicine, 50, 1385-1401.

DeVellis, R.F. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage.

Ferraro, K.F. (1993). Are black older adults health-pessimistic? Journal of Health Social Behavior, 34, 201-204.

Ferraro, K.F., & Farmer, M.M. (1996). Double jeopardy to health hypothesis for blacks: Analysis and critique. Journal of Health and Social Behavior, 37, 27-43.

Foley, L.F., Reed, R.S., Mutran, E.J., & DeVellis, R.F. (2002). Measurement adequacy of the CES-D among a sample of older African-Americans. Psychiatry Research, 109, 61-69.

Fullilove, R.E., Fullilove, M.T., Northride, M.E., Ganz, M.L., Bassett, M.T., McLean, D.E., et al. (1999). Risk factors for excess mortality in Harlem. Findings from the Harlem Household Survey. American Journal of Preventive Medicine, 16, 22-28.

Garcia, M., & Marks, G. (1989). Depressive symptomatology among Mexican American adults: An examination with CES-D scale. Psychiatry Research, 27, 137-148.

Gatz, M., Johanasson, B., Pedersen, N., Berg, S., & Reynolds, C. (1993). A cross-national self-report measure of depressive symptomatology. International Psychogeriatrics, 5, 147-156.

Geronimus, A.T. (1996). The health of urban African American men: Excess mortality and causes of death. Aspen Institute's Roundtable on Comprehensive Community Initiatives Project on Race and Community Revitalization, Queenstown, MD.

Guarnaccia, P.J., Angel, R., & Worobey, J.L. (1989). The factor structure of the CES-D in Hispanic Health and Nutrition Examination Survey: The influences of ethnicity, gender, and language. Social Science and Medicine, 29, 85-94.

Hertzog, C., Van Alstine, J., Usala, P.D., Hutsch, D.F., & Dixon, R. (1990). Measurement prosperities of center for epidemiological studies depression scale in older populations: Psychological assessment. A Journal of Consulting and Clinical Psychology, 2, 64-72.

Hortwitz, A.V., & Scheid, T.L. (1999). Approaches to mental health and illness: Conflicting definitions and emphases. In A.V. Horwtiz & T.L. Scheid (Eds.), .4 handbook for the study of mental health: Social context, theories, and systems (pp. 1-11). New York: Cambridge University Express.

Kim, J.O., & Mueller, C.W. (1978). Introduction to factor analysis: What it is and how to do it. In M.S. Lewis-Beck (Ed.), Quantitative application in the social sciences. Newbury Park, CA: Sage.

Kuo, W. (1984). Prevalence of depression among Asian-Americans. Journal of Nervous and Mental Disease, 172, 449-457.

McCord, C., & Freeman, H.P. (1997). Excess mortality in Harlem. In P. Conrad (Ed.), The sociology of health and illness: Critical perspective (5th ed., pp. 35-42). New York: St. Martin Press.

Miles, T.P. (Ed.). (1999). Full-color aging: Facts, goals, and recommendations for America's diverse elders. Washington, DC: Gerontological Society of America.

Miller, T.Q., Markides, K.S., & Blacks, S.A. (1997). The factor structure of the CES-D in two surveys of elderly Mexican Americans. Journal of Gerontology: Social Sciences, 52B, S259-S269.

Radloff, L.S. (1977). The CES-D Scale: A self reported depression scale for research in general population. Applied Psychological Measurement, 1, 385-401.

Roberts, R.E. (1980). Reliability of the CES-D scale in different ethnic contexts. Psychiatry Research, 2, 125-134.

Roberts, R.E., Vernon, S.W., & Rhoades, H.M. (1989). Effects of language and ethnic status on reliability and validity of the Center for Epidemiologic Studies-Depression Scale with psychiatric patients. Journal of Nervous and Mental Disease, 177, 581-592.

Rosenbaum, J.F. (1996). Depression: More and meaner than it seems. The Harvard Maloney Neuroscience Institute Letter, 5, 1-2.

Ross, C.E., & Mirowski, J. (1984). Components of depressed mood in married men and women. American Journal of Epidemiology, 119, 997-1004.

Siegel, J.L. (1999). Demographic introduction to racial Hispanic elder population. In T.P. Miles, (Ed.). Full-color aging: Facts, goals, and recommendations for America's diverse elders. Washington, DC: Gerontological Society of America.

Sheehan, T., Fifield, J., Reisine, S., & Tennen, H. (1995). The measurement structure of the Center for Epidemiologic Studies Depression Scale. Journal of Personality Assessment, 64, 507-521.

Somervell, P.D., Beals, J., Kinzie, J.D., Boehnlein, J., Leung, P., & Manson, S.M. (1993). Criterion validity of the Center for Epidemiologic Studies Depression Scale in a population sample from an American Indian village. Psychiatry Research, 47, 255-266.

Tabachnick, B.G., & Fidel, L.S. (2001). Using multivarate statistics (4th ed.). Boston: Allyn and Bacon.

Williams, D.R. (1995). Black mental health: Persisting questions and paradoxical findings. Black Research Perspective, 1, 1-7.

Williams, D.R., & Williams-Morris, R. (2000). Racism and mental health: The black experience. Ethnicity & Health, 5, 243-268.

Wilson, W. (1996). When work disappears: The world of the new urban poor. New York: Random House.

ASHLEY S. LOVE

University of Texas at San Antonio

San Antonio, TX

ROBERT J. LOVE

University of Texas Health Center at San Antonio

San Antonio, TX

Correspondence concerning this article should be directed to: Dr. Ashley S. Love, 6900 North Loop 1604 West. San Antonio. TX 78249. Electronic mail: [e-mail].

Table 1
Demographics of Harlem Clinic and Community Samples

                              Clinic         Community
                             (N = 404)       (N = 319)

Age, yr (SD)               60.35 (7.02)    61.35 (7.42)
  50-54 years, %           26.2            24.8
  55-59 years              22.5            19.7
  60-64 years              19.8            18.2
  65-69 years              19.3            18.5
  70-74 years              12.1            18.8
Education,
  12th grade education
    or less (no high
    school diploma)        31.8            30.8
  High school graduate     46.3            38.7            [chi square]
  Some college                                             = 12.01 **
    vocational/
    technical/
    business school         7.3            11.7
  College degree or more    9.3            15.3
Marital status,
  Single/never married     18.8            21.5
  Married                   3.0            10.7
  Living together          21.0            27.0            [chi square]
  Divorced/separated       18.1            23.5            = 12.27 *
  Widowed                  16.1            8.5
Insurance status, %
  Medicaid                 61.2            31.3
  Other insurance          14.4            85.6            [chi square]
Employment status, %                                       = 63.48 *
  Full time                10.6            18.8
  Part time                 1.5             2.2
  Irregular                 1.2             2.2            [chi square]
  Retired                  37.6            48.9            = 33.90 *
  Unemployed               49.0            28.2
Household income, %
  Less than $7K to $15K    82.6            65.9
  $15,001 to $25K           8.8            10.9            [chi square]
  $25,001 to $40K           4.7            12.3            = 37.61
  $40,001+                  3.9            10.9
Years in Harlem (SD)       36.00 (17.53)   36.75 (18.71)
  Less than 10 years        9.7            10.4
  10-19 years               8.6             8.5
  20-29 years              11.6            13.3
  30-39 years              20.7            20.4
  40 years and more        49.9            47.5

* p < .05; ** p < .01.

Table 2
Item-Total Correlation and Response Distribution of CES-D Items
among Older Black Men Living in Harlem: Community Sample (N = 319)

                                     How Often Have You Felt This
                                         Way during This Week?

                                                        Item-Total
                                     Abbreviation      Correlation

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 316)                        Blue                  .63
  I felt depressed (n = 317)         Depressed             .71
  I thought my life had been
    a failure (n = 319)              Failure               .66
  I felt fearful (n = 318)           Fearful               .66
  I felt lonely (n = 319)            Lonely                .67
  I had crying spells (n = 318)      Crying                .46
  I felt sad (n = 318)               Sad                   .68
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 318)                        Bothered              .48
  I did not feel like eating;
    my appetite was poor
    (n = 318)                        Appetite              .52
  I had trouble keeping my
    mind on what I was doing
    (n = 316)                        Mind                  .61
  I felt that everything I did
    was an effort (n = 316)          Effort                .53
  My sleep was restless (n = 319)    Sleep                 .59
  It seemed that I talked less
    than usual (n = 318)             Talk                  .45
  I could not get going (n = 316)    Get going             .58
Positive affect
  I felt that I was just as good
    as other people (n = 315)        Good                  .22
  I felt hopeful about the
    future (n = 317)                 Hopeful               .46
  I was happy (n = 319)              Happy                 .47
  I enjoyed life (n = 318)           Enjoy                 .49
Interpersonal distress
  People were unfriendly (n = 319)   Unfriendly            .47
  I felt that people disliked me
    (n = 317)                        Dislike               .47

                                     How Often Have You Felt This
                                        Way during This Week?

                                     Rarely or None   Some or Little
                                      of the Time      of the Time
                                      (<1 day) (%)    (l-2 days) (%)

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 316)                            71.5             15.8
  I felt depressed (n = 317)             65.6             18.3
  I thought my life had been
    a failure (n = 319)                  77.7             11.6
  I felt fearful (n = 318)               75.5             15.1
  I felt lonely (n = 319)                62.7             20.4
  I had crying spells (n = 318)          89.6             7.2
  I felt sad (n = 318)                   68.6             19.8
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 318)                            61.6             22.0
  I did not feel like eating;
    my appetite was poor
    (n = 318)                            73.6             14.5
  I had trouble keeping my
    mind on what I was doing
    (n = 316)                            71.5             13.6
  I felt that everything I did
    was an effort (n = 316)              61.4             17.1
  My sleep was restless (n = 319)        63.9             16.9
  It seemed that I talked less
    than usual (n = 318)                 70.8             13.2
  I could not get going (n = 316)        72.8             14.6
Positive affect
  I felt that I was just as good
    as other people (n = 315)             5.1              3.8
  I felt hopeful about the
    future (n = 317)                      8.8             10.4
  I was happy (n = 319)                   7.8             17.2
  I enjoyed life (n = 318)                6.3              7.5
Interpersonal distress
  People were unfriendly (n = 319)       80.3             11.3
  I felt that people disliked me
    (n = 317)                            84.5             10.4

                                     How Often Have You Felt This
                                        Way during This Week?

                                      Occasionally
                                     or a Moderate     Most or All
                                     Amount of Time    of the Time
                                     (3-4 days) (%)   (5-7 days) (%)

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 316)                              6.0              6.6
  I felt depressed (n = 317)               9.5              6.6
  I thought my life had been
    a failure (n = 319)                    5.6              5.0
  I felt fearful (n = 318)                 5.7              3.8
  I felt lonely (n = 319)                  7.5              9.4
  I had crying spells (n = 318)            1.6              1.6
  I felt sad (n = 318)                     7.2              4.4
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 318)                              8.5              7.9
  I did not feel like eating;
    my appetite was poor
    (n = 318)                              7.2              4.7
  I had trouble keeping my
    mind on what I was doing
    (n = 316)                              8.5              6.3
  I felt that everything I did
    was an effort (n = 316)                9.2             12.3
  My sleep was restless (n = 319)          9.1             10.0
  It seemed that I talked less
    than usual (n = 318)                   8.2              7.9
  I could not get going (n = 316)          5.7              7.0
Positive affect
  I felt that I was just as good
    as other people (n = 315)              2.5             88.6
  I felt hopeful about the
    future (n = 317)                       9.5             71.3
  I was happy (n = 319)                   12.9             62.1
  I enjoyed life (n = 318)                 7.5             78.6
Interpersonal distress
  People were unfriendly (n = 319)         5.6              2.8
  I felt that people disliked me
    (n = 317)                              3.2              1.9

Table 3
Item-Total Correlation and Response Distribution of CES-D Items
among Older Black Men Living in Harlem: Clinical Sample (N = 404)

                                       How Often Have You Felt
                                      This Way during This Week?

                                                        Item-Total
                                     Abbreviation      Correlation

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 401)                        Blue                  .63
  I felt depressed (n = 403)         Depressed             .67
  I thought my life had been
    a failure (n = 403)              Failure               .53
  I felt fearful (n = 403)           Fearful               .55
  I felt lonely (n = 401)            Lonely                .60
  I had crying spells (n = 403)      Crying                .33
  I felt sad (n = 403)               Sad                   .68
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 403)                        Bothered              .52
  I did not feel like eating;
    my appetite was poor (n = 403)   Appetite              .45
  I had trouble keeping my mind
    on what I was doing (n = 403)    Mind                  .50
  I felt that everything I did was
    an effort (n = 403)              Effort                .39
  My sleep was restless (n = 403)    Sleep                 .44
  It seemed that I talked less
    than usual (n = 402)             Talk                  .44
  I could not get going (n = 402)    Get going             .53
Positive affect
  I felt that I was just as good
    as other people (n = 402)        Good                  .28
  I felt hopeful about the
    future (n = 401)                 Hopeful               .33
  I was happy (n = 403)              Happy                 .40
  I enjoyed life (n = 402)           Enjoy                 .42
Interpersonal distress
  People were unfriendly (n = 403)   Unfriendly            .51
  I felt that people disliked me
    (n = 401)                        Dislike               .51

                                       How Often Have You Felt
                                      This Way during This Week?

                                     Rarely or None   Some or Little
                                      of the Time      of the Time
                                      (<1 day) (%)    (1-2 days) (%)

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 401)                             65.1             17.0
  I felt depressed (n = 403)              55.8             25.1
  I thought my life had been
    a failure (n = 403)                   68.5             17.6
  I felt fearful (n = 403)                66.7             19.4
  I felt lonely (n = 401)                 52.4             25.7
  I had crying spells (n = 403)           86.6              8.2
  I felt sad (n = 403)                    59.6             23.6
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 403)                             48.1             61.1
  I did not feel like eating;
    my appetite was poor (n = 403)        60.5             18.4
  I had trouble keeping my mind
    on what I was doing (n = 403)         55.8             23.3
  I felt that everything I did was
    an effort (n = 403)                   40.2             21.3
  My sleep was restless (n = 403)         51.4             24.1
  It seemed that I talked less
    than usual (n = 402)                  58.5             19.2
  I could not get going (n = 402)         61.7             24.9
Positive affect
  I felt that I was just as good
    as other people (n = 402)              7.0              7.5
  I felt hopeful about the
    future (n = 401)                      11.5             13.7
  I was happy (n = 403)                   11.9             19.4
  I enjoyed life (n = 402)                 4.0             12.2
Interpersonal distress
  People were unfriendly (n = 403)        63.8             21.3
  I felt that people disliked me
    (n = 401)                             74.3             16.0

                                       How Often Have You Felt
                                      This Way during This Week?

                                      Occasionally
                                     or a Moderate     Most or All
                                     Amount of Time    of the Time
                                     (3-4 days) (%)   (5-7 days) (%)

Depressed affect
  I felt that I could not shake
    off the blues, even with help
    from my family and friends
    (n = 401)                             12.0             6.0
  I felt depressed (n = 403)              11.7             7.4
  I thought my life had been
    a failure (n = 403)                    9.4             4.5
  I felt fearful (n = 403)                10.4             3.5
  I felt lonely (n = 401)                 12.5             9.5
  I had crying spells (n = 403)            3.0             2.0
  I felt sad (n = 403)                    12.7             4.2
Somatic symptoms
  I was bothered by things that
    don't usually bother me
    (n = 403)                             15.6             8.9
  I did not feel like eating;
    my appetite was poor (n = 403)        15.4             5.7
  I had trouble keeping my mind
    on what I was doing (n = 403)         11.7             5.1
  I felt that everything I did was
    an effort (n = 403)                   17.1            21.3
  My sleep was restless (n = 403)         14.1            10.4
  It seemed that I talked less
    than usual (n = 402)                  14.4            10.0
  I could not get going (n = 402)          8.7             4.7
Positive affect
  I felt that I was just as good
    as other people (n = 402)              9.5            76.5
  I felt hopeful about the
    future (n = 401)                      13.0            61.8
  I was happy (n = 403)                   19.1            49.5
  I enjoyed life (n = 402)                13.9            39.0
Interpersonal distress
  People were unfriendly (n = 403)        10.2             4.7
  I felt that people disliked me
    (n = 401)                              5.7             4.0

Table 4
A Comparison of the Original Factor Structure of the CES-D When
Applied to Older Black Men in Harlem Using Exploratory Factor
Analysis: Community Sample (N = 318) and Clinic Sample (N = 404)

                                   Clinic (c)

                                Rotated
             Radloff
Items        (1977) (a,b)   Loading   Factor    [h.sup.2]

Bothered     S              D/S         .66        .46
Appetite     S              D/S         .63        .40
Blues        D              D/S         .64        .53
Good         P              P           .60        .45
Mind         S              D/S         .62        .42
Depressed    D              D/S         .66        .59
Effort       S              D/S         .53        .30
Hopeful      P              P           .65        .49
Failure      D              D/S         .48        .43
Fearful      D              D/S         .52        .44
Sleep        S              D/S         .56        .37
Happy        P              P           .64        .52
Talk         S              D/S         .53        .32
Lonely       D              D/S         .55        .48
Unfriendly   I              I           .53        .42
Enjoy        P              P           .69        .52
Cry          D              I           .69        .49
Sad          D              D/S         .56        .54
Dislike      I              I           .66        .55
Get going    S              D/S         .51        .47

                                  Community (c)

                                 Rotated

Items                       Factor    Loading   [h.sup.2]

Bothered                    D/S         .56        .39
Appetite                    I           .64        .54
Blues                       D/S         .64        .56
Good                        P           .44        .38
Mind                        D/S         .45        .47
Depressed                   D/S         .58        .66
Effort                      D/S         .60        .43
Hopeful                     P           .68        .49
Failure                     D/S         .60        .57
Fearful                     D/S         .62        .52
Sleep                       D/S         .51        .51
Happy                       P           .59        .46
Talk                        D/S         .61        .48
Lonely                      D/S         .70        .63
Unfriendly                  D/S         .70        .57
Enjoy                       P           .71        .58
Cry                         I           .74        .59
Sad                         D/S         .50        .57
Dislike                     I           .69        .57
Get going                   D/S         .48        .44

Note: Principle Component Factor Analysis was performed for both
clinic and community samples. Three factors were identified for
both groups. In the clinic, they were D/S (Eigenvalue = 6.55,
Variance Accounted = 32.75%), 1 (Eigenvalue = 1.57, Variance

(a) Original factor structure: D = depressed affect; S = somatic
complaints; P = positive; I = interpersonal.

(b) Factor loadings indicated in this text for three white
populations; structure was reported to be similar for blacks.

(c) Factor loadings for this study: D/S = Depressed/Somatic.