- Email: [email protected]
Depression in Older Adults With Schizophrenia Spectrum Disorders: Prevalence and Associated Factors
SPECIAL ARTICLE
Depression in Older Adults With Schizophrenia Spectrum Disorders: Prevalence and Associated Factors Shilpa Diwan, M.D., M.P.H., Carl I. Cohen, M.D., Azziza O. Bankole, M.D., Ipsit Vahia, M.D., Michelle Kehn, M.A., Paul M. Ramirez, Ph.D.
Rationale: Although depression is common in older adults with schizophrenia, it has not been well studied. The authors examine those factors that are related to depression in a multiracial urban sample of older persons with schizophrenia. Methods: The schizophrenia group consisted of 198 persons aged 55 or older who lived in the community and developed schizophrenia before age 45. Persons with substantial cognitive impairment were excluded from the study. A community comparison group (N⫽113) was recruited using randomly selected census tract data. The authors adapted George’s Social Antecedent Model of Depression, which consists of six categories comprising 16 independent variables, and used a dichotomous dependent variable based on a Center for Epidemiologic Studies Depression Scale cutoff score of ⱖ16. Results: The schizophrenia group had significantly more persons with clinical depression than the community comparison group (32% versus 11%, respectively; 2 ⫽ 28.23, df⫽1, p⫽0.001). Bivariate analysis revealed that eight of the 16 variables were significantly related to clinical depression in the schizophrenia group. In logistic regression, six variables retained significance: physical illness (odds ratio [OR]⫽1.60, 95% confidence interval [CI], 1.17–2.18), quality of life (OR⫽0.84, 95% CI, 0.76 – 0.93), presence of positive symptoms (OR⫽1.12, 95% CI, 1.02–1.21), proportion of confidants (OR⫽0.03, 95% CI, 0.01– 0.39), copes by using medications (OR⫽2.12, 95% CI, 1.08 – 4.13), and copes with conflicts by keeping calm (OR⫽1.34, 95% CI, 1.03–1.74). Conclusion: Consistent with earlier studies of schizophrenia in older persons, the authors found physical health, positive symptoms, and several nonclinical variables to be associated with depression. Potential points for intervention include strengthening social supports, improving physical well-being, more aggressive treatment of positive symptoms, and increasing the recognition and treatment of depression. (Am J Geriatr Psychiatry 2007; 15:991–998) Key Words: schizophrenia, depression, urban, multiracial
Received January 21, 2007; revised September 7, 2007; accepted September 10, 2007. From SUNY Downstate Medical Center (SD, CIC, AOB, IV) and Long Island University (MK, PMR), Brooklyn, New York. Send correspondence and reprint requests to Shilpa Diwan, M.D., M.P.H., Resident in Psychiatry-PGY IV, SUNY Downstate Medical Center, Box 1203, 450 Clarkson Avenue, Brooklyn, NY 11203; e-mail: [email protected] © 2007 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 15:12, December 2007
991
Depression in Older Adults With Schizophrenia Spectrum Disorders
D
epression is common among older persons with schizophrenia, with rates ranging from 44% to 75%.1– 4 Although depression is thought to have a potentially deleterious effect on the outcome of schizophrenia,5 there have been only four published studies that have looked specifically at depression among older persons with schizophrenia.1– 4 Among the studies that compared depressed and nondepressed older persons with schizophrenia,1,3,4 a consistent finding was the relationship between positive symptoms and depression. In these studies, depression was also found to be significantly associated with other variables such as quality of well-being, instrumental activities of daily living, various social network measures, physical limitations, lower income, and several factors on the SF-36, such as social functioning, role limitations, and general health.3,4 Lower age was found to be significant in one study but not in the other two studies.1,3 Earlier studies of depression among older persons with schizophrenia have had several limitations. In the studies from San Diego, California,1,3 the absence of a multivariate theoretical model made the findings more difficult to generalize to other sites and precluded an assessment of the overlapping effects of the various predictor variables on depression. Moreover, these studies did not include several variables that have been found to be important in studies of depression in older adults in general (e.g., life stressors, financial strain, social support, and coping styles).6 –12 Although the New York City study4 utilized a theoretical model and examined more variables, the number of variables was limited by the sample size (N⫽117) and by the use of instruments for measuring psychotic symptoms and depression that had not been used previously with schizophrenia populations. Moreover, this study lacked an adequate community comparison group with which to contrast the prevalence of depression. To address some of the limitations of earlier investigations, we have used an adaptation of George’s Social Antecedent Model of Depression.13 This is a conceptual model of the social precursors of psychiatric disorders based on the consensus in the social science, epidemiologic, and social psychiatry literature. The model consists of six categories of social and clinical factors that predict depression: demographic variables; early life events such as education level and childhood traumas; later events and
992
achievements such as income and marital status; social integration; vulnerability and protective factors such as chronic stressors and social supports; and provoking agents and coping efforts such as major life events and coping styles. Using a large multiracial urban sample of persons with schizophrenia who were age 55 and older and a community comparison group, we address the following questions: 1) What is the prevalence of clinical depression among older persons with schizophrenia and how does it compare with their age peers in the general population? 2) What factors are associated with depression among older persons with schizophrenia?
METHODS We recruited subjects aged 55 and older who lived in the community and developed schizophrenia before the age of 45 using stratified sampling in which we attempted to interview approximately one-half of the subjects from outpatient clinics and day programs and the other one-half from supported community residences, including sites with varying degrees of on-site supervision. Inclusion was based on a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, chart diagnosis of schizophrenia or schizoaffective disorder that was supplemented by a lifetime illness review adapted from Jeste et al.14 Thus, respondents were required to have one or more of the following: 1) symptoms such as social withdrawal, loss of interest in school or work, deterioration in hygiene and grooming, unusual behaviors, or outbursts of anger before age 45; 2) evidence of hallucinations, delusions, or disorganized thinking or behavior before age 45; 3) hospitalization for schizophrenia before age 45; and/or 4) treatment for schizophrenia before age 45. Subjects were excluded if they had medical conditions or trauma, seizures, dementia, head injury with unconsciousness of ⱖ30 minutes, mental retardation, substance abuse, or evidence of bipolar illness that could better account for the subject’s psychotic symptoms before age 45. We excluded persons with cognitive impairment too severe for completion of the questionnaire (i.e., defined as scores of ⬍5 on the Mental Status Questionnaire).15 A community comparison group was recruited using Wessex Census STF3 files of Kings County
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al. (Brooklyn, NY). Persons age 55 and older were extracted. A computer program was written to randomly select without replacement Census Bureau groupings of blocks (“block-groups”) as the primary sampling unit. Without replacement sampling was used to avoid unnecessary duplication of effort and waste of resources. An effort was made to interview all persons in a selected block-group by knocking on doors. To enhance response rates, subjects from the selected block-group were also recruited at senior centers and churches and through personal references. Persons were excluded if they reported a history of outpatient treatment or hospitalization for schizophrenia or schizoaffective disorder. Subjects were offered $75 for completing the 2.5hour interview. The rejection rate was 7% in the schizophrenia group and 48% of those contacted in the community group. The schizophrenia sample consisted of 198 persons with schizophrenia; 39% were living independently in the community, and 61% lived in supported community residences. From the original community sample (N⫽206), we selected 113 persons who more closely matched the schizophrenia sample for age, sex, race, and income.
TABLE 1.
Forty-nine percent of the schizophrenia group were women versus 49% of the community comparison group, and 35% of the schizophrenia group was African American versus 36% of the community comparison group (2 ⫽ 0.07, df ⫽1, p⫽0.70). The mean ages of the schizophrenia and community comparison groups were 61.5 years and 63.0 years, respectively (t⫽2.37, df⫽ 306, p⫽0.02). Thus, despite a small absolute difference in age between the two groups, this difference attained statistical significance. Median income for both groups fell within the same category ($7,000 –$12,999). Instruments We adapted George’s model that consisted of six categories comprising 16 independent variables. These variables were selected based on their theoretical importance as well as their having been identified in the literature as impacting on depression. The 16 independent variables (Table 1) were derived from the following scales: a “Physical Illness” score that represented the sum of seven health care items and 13 illness categories derived from the Multilevel
Bivariate Analysis of Variables for Depressed (CES-D Score, >16) and Nondepressed Older Persons With Schizophrenia Using George’s Social Antecedent Model
Variable
Nondepressed (N ⴝ 134)
Depressed (N ⴝ 64)
61.8 (5.5) 46 34
60.9 (5.7) 55 38
3.2 (1.97)
Demographic factor Age, years Female Black Event and achievement Household income Social integration Proportion of contacts considered intimates Vulnerability factor Physical illness Instrumental activities of daily living Dementia rating scale Positive symptoms Negative symptoms Quality of Life Index Provoking agent Acute stressors Coping strategy Copes by using medication Copes with conflicts by keeping calm Use of spiritualists or their products Takes antidepressants
2/t-Value
Df
p
0.97 1.23 0.29
193 1 1
0.33 0.27 0.59
3.1 (1.1)
0.27
191
0.79
0.3 (0.3)
0.2 (0.2)
2.45
177
0.02a
1.0 (1.1) 22.3 (3.8) 127.2 (13.9) 11.5 (5.3) 11.6 (6.0) 23.3 (4.6)
1.9 (1.8) 21.9 (4.1) 128.7 (11.2) 14.8 (7.4) 12.7 (6.0) 18.4 (5.1)
⫺3.35 0.68 ⫺0.85 ⫺3.25 ⫺1.15 6.64
84 196 150 96 196 193
0.001a 0.50 0.40 0.002a 0.25 0.001a
6.7 (5.8)
12.9 (8.2)
⫺5.47
94
0.001a
1.2 (0.7) 2.2 (1.6) 34 27
1.4 (0.6) 2.8 (1.7) 63 39
⫺2.37 ⫺2.46 14.78 3.02
193 196 1 1
0.02a 0.02a 0.001a 0.08
Notes: Unless stated otherwise, data are mean percentage (SD). For t-tests, the unequal variance t-test was used when findings of the Levene test were significant. CES-D: Center for Epidemiologic Studies Depression Scale. a Author will explain.
Am J Geriatr Psychiatry 15:12, December 2007
993
Depression in Older Adults With Schizophrenia Spectrum Disorders Assessment Inventory; Physical Health Status (lower scores indicate worse health)16; instrumental activities of daily living (lower scores indicate more impairment)16; Acute Stressors Scale17 that examined 11 life events that may have occurred in the past month; the Dementia Rating Scale,18 which assesses five areas of cognitive functioning (namely, attention, initiation and perseveration, construction, conceptualization, and memory [higher scores indicate better functioning]); the Network Analysis Profile,19 from which we used the confidant variable (i.e., shares intimate thoughts); the Positive and Negative Syndrome Scale (PANSS), which consists of seven items assessing positive symptoms, seven items assessing negative symptoms, and 16 items assessing general psychopathology; the Quality of Life Index, which consists of 64 Likert-type items divided into two sections (assessing satisfaction or importance in various life domains); the Pearlin Coping Scale,20 a 23item scale used to assess coping strategies to various stressors that was subdivided into eight subscales using principal components analysis with varimax rotation; and the Conflict Tactics Scale,21 which was subdivided into seven subscales using principal components analysis with varimax rotation. From each of the latter two measures, we selected the subscale that had the strongest association with the dependent variable: namely, Use of Medication and/or Professional Assistance subscale from the Pearlin Coping Scale and the Keeping Calm subscale from the Conflict Tactics Scale. The latter consisted of three items: discussed situation calmly, got information, and brought someone to help. “Spiritualism” was rated positive if the respondent saw a spiritualist or purchased items from a religious shop. The dependent variable was “syndromal depression,” which was based on the traditional cutoff score of ⱖ16 on the Center for Epidemiologic Studies Depression Scale (CES-D).22 We also made comparisons with the community sample using “subsyndromal depression,” which was derived from our earlier work and based on a CES-D score of 8 to 15.22 The internal reliability (Cronbach ␣) scores for the scales were as follows: Quality of Life Index, 0.97; CES-D, 0.88; PANSS positive scale, 0.83; PANSS negative scale, 0.78; Acute Stressors Scale, 0.88; instrumental activities of daily living, 0.74; Dementia Rating Scale, 0.89; Pearlin Coping Scale, 0.75; and
994
Conflict Tactics Scale, 0.82. All scales attained a recommended ␣ of ⱖ0.60.23 Interviewers were trained using audiotapes and videotapes, and they were generally matched to respondents from similar ethnic backgrounds. Interviewers were periodically monitored using audiotapes. The intraclass correlations ranged from 0.79 to 0.99 on the various scales. Raters had a mean of 99% and 94% agreement with expert ratings (CIC) on the Physical Illness score and the Network Analysis Profile contact scores, respectively. Data Analysis Initially, we conducted bivariate analyses to examine the relationship between independent and dependent variables; t-tests and 2 analyses were used for continuous and categorical independent variables, respectively. For instruments that generated potentially more than one variable for inclusion in the model, we conducted preliminary bivariate analyses and selected variables from that measure that had substantial associations (p ⬍0.10) with the dependent variable (data from preliminary analyses are not shown). In instances in which several variables attained this cutoff, we selected the variable with the lowest p value. Using this method, we selected the two coping variables described above and a social network variable (proportion of contacts considered intimates). Logistic regression analysis was used to examine those variables associated with depression. There was no evidence of collinearity among the independent variables.
RESULTS As seen in Table 2, 32%, 29%, and 39% of persons with schizophrenia had syndromal depression, subsyndromal depression, and no depression, respectively. In the community comparison group, 11%, 19%, and 71% of persons had syndromal depression, subsyndromal depression, and no depression, respectively. The difference between the groups was significant (2 ⫽31.36, df⫽2, p ⬍0.001). There was no statistical difference in syndromally depressed persons who reported taking medications for depression in the schizophrenia group (39%) and the community comparison group (42%) (2 ⫽0.00, df⫽1, p⫽1.0).
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al.
TABLE 2.
Persons in the Community and Schizophrenia Groups Who Had Various Levels of Depression
Group
Syndromal Depression, CES-D Score, >16
Subsyndromal Depression, CES-D Score, 8 –15
No Depression, CES-D Score, <8
11% 32%
19% 29%
71% 39%
Community comparison (N ⫽ 113) Schizophrenia (N ⫽ 198) Note: 2 ⫽ 31.36, df ⫽ 2, p ⬍0.001.
In the schizophrenia group, 44% had psychotic symptoms at the time of interview, which was defined as scores of ⬎2 on one or more on the PANSS positive scale items associated with psychosis (i.e., hallucinations, delusions, or disorganized thinking). Among persons with psychotic symptoms, 43% had syndromal depression; among persons without psychotic symptoms, 25% had syndromal depression. Bivariate analysis revealed that eight of the 16 variables in the model were significantly related to syndromal depression in the schizophrenia group (Table 1). Depression was associated with more physical disorders, more acute stressors, more positive symptoms, lower proportion of confidants in their social network, greater use of medications/ health professionals as a coping strategy for stress, greater use of keeping calm when having conflicts with others, lower quality of life score, and greater use of spiritualists or their products. When the model was tested in logistic regression analysis, six variables retained significance: more physical disorders, presence of positive symptoms, smaller proportion of confidants, copes by using medications/health professionals, copes by keeping calm, and a lower quality of life score (Table 3). The overall model was statistically significant (2 ⫽ 84.99, df⫽16, p ⫽0.001).
TABLE 3.
Logistic Regression Analysis of Variables Predicting Depression (CES-D Score, >16) in Older Persons With Schizophrenia
Variable Demographic factor Age Female Black Event and achievement Household income Social integration Proportion of intimate contacts Vulnerability factor No. of physical illnesses Instrumental activities of daily living Dementia Rating Scale Positive symptoms Negative symptoms Quality of Life Index Provoking agent Acute stressors Coping strategy Copes by using medication Copes with conflicts by keeping calm Use of spiritualists or their products Use of antidepressants
Odds Ratio
95% CI
1.01 1.20 0.90
0.94–1.09 0.51–2.78 0.35–2.28
0.73 0.68 0.81
1.18
0.93–1.48
0.17
0.03
0.01–0.39
0.01a
1.60 1.05
1.17–2.18 0.93–1.18
0.003a 0.44
1.02 1.12 0.99 0.84
0.99–1.06 1.02–1.21 0.91–1.08 0.76–0.93
0.23 0.04a 0.81 0.001a
1.04
0.96–1.11
0.34
2.12 1.34
1.08–4.13 1.03–1.74
0.03a 0.03a
2.31
0.94–5.67
0.07
0.88
0.35–2.18
0.78
p
Notes: Overall model: ⫽ 84.99, df ⫽ 16, p ⫽ 0.001. Odds ratios of ⬎1 are associated with depression; odds ratios of ⬍1 are associated with absence of depression. Wald 2 tests with df ⫽ 1 were used to obtain p values. CI: confidence interval; CES-D: Center for Epidemiologic Studies Depression Scale. a Author will explain. 2
DISCUSSION The prevalence of syndromal depression was threefold greater in the schizophrenia group than in the community comparison group, and the prevalence of any level of depression in the schizophrenia group was twice that in the community comparison group. Moreover, the proportion of persons with schizophrenia who had subsyndromal depression and syndromal depression (61%) was similar to the levels found among older persons with schizophrenia in San Diego, California (68%),1,3 although our group
Am J Geriatr Psychiatry 15:12, December 2007
had more severe depression than their group (32% versus 13%, respectively). Thus, our study confirms earlier findings that older people with schizophrenia have substantially higher rates of depression than their age peers in the general population. Our findings were remarkably consistent with those of an der Heiden et al;24 in their longitudinal investigation of a younger schizophrenia cohort, about one-third of persons had depressive symptoms at any point in time, and the prevalence of depression
995
Depression in Older Adults With Schizophrenia Spectrum Disorders rose to one-half during exacerbation or recurrence of psychoses. We found that 32% of our sample had syndromal depression but that 43% of persons with psychosis had syndromal depression. On the basis of work with younger populations, several investigators have proposed that depressive features may be a “core” feature of schizophrenia.24 –26 Our findings for older persons would suggest that this may be premature. At a minimum, there should be a consensus on how depression is defined (i.e., subsyndromal or syndromal) and what level of prevalence would make it a “core” feature (e.g., in our sample, 39% had no clinically relevant depression). These concerns notwithstanding, our data strongly support the argument that depression remains a common associated feature of schizophrenia throughout the life span. With respect to our overall theoretical model, we found that eight variables were significant in bivariate analysis and that six retained significance in logistic regression analysis. Four of the six categories had one significant variable in bivariate analysis, and three categories (vulnerability, social integration, and coping) had one significant variable in logistic regression analysis. Although the overall model was statistically significant, the fact that 10 variables did not retain independent significance in the final analysis suggests that trimming of the model or use of alternative variables may be indicated. With respect to specific variables in the vulnerability category of the model, we identified a number of items that were associated with higher rates of depression, suggesting potential points for clinical intervention. Most notably, our data support earlier findings that depression is associated with positive symptoms.1,27 Although significant, the odds ratio for this positive symptoms was low. However, the consistency of the relationship across several studies suggests that it is a meaningful finding that may have clinical implications. Because of the cross-sectional nature of this study, it is unclear whether depression exacerbates positive symptoms, whether positive symptoms result in more depression, or whether there is a mutually reinforcing effect. Another important finding is the relationship among physical illness, disability, and depression in older persons with schizophrenia. Although there is some variability across studies (i.e., in some studies, depression was related to physical or activity limitations and not the number of physical disorders,
996
whereas in this study the obverse was obtained), there is a strong indication that like other older persons, physical illness and/or disability is linked to depression. Here again, causal direction cannot be determined, although other researchers have suggested that the relationship between physical illness and depression is probably bidirectional.28,29 In recent years, there has been an increased recognition that older persons with schizophrenia are not receiving adequate medical treatment despite having access to care.30 The interplay of physical illness and depression should be examined with respect to its impact on healthcare in older persons with schizophrenia. Our findings highlight several other important variables that may play a role in affecting depression that have not been considered previously. For example, within the coping category of the model, we found that certain coping strategies were more commonly used by depressed persons with schizophrenia, although it is not clear whether these strategies serve to diminish or exacerbate the symptoms. For example, keeping calm to deal with interpersonal conflicts may merely be a preferred coping strategy among depressed persons with schizophrenia or it may serve to further worsen depressive symptoms. Similarly, within the social integration category of the model, we found that depressed persons had a smaller proportion of network confidants, and although paucity of network support has been thought to exacerbate depression in older persons in general,31 we could not accurately determine causal direction. However, it is evident that future studies need to assess the impact of various social and coping strategies as well as the potential impact of alternative treatments such as the use of spiritualists or their products, the latter being significantly associated with depression in our sample in bivariate analysis and marginally significant in logistic regression analysis. Our findings are consistent with those of Jin et al.3 that depression is associated with lower perceived quality of life. We demonstrated that this effect was independent of other variables. Thus, in addition to its associations with physical impairment and psychotic symptoms, depression is associated with diminished perception of life quality. A thorny issue is whether quality of life is really a measure of depression. In another study of this population, we found that the Quality of Life Index accounted for 28% of the explained variance in the CES-D.32 Thus, al-
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al. though there is overlap with depressive symptoms, quality of life is not identical with depression. Although causality could not be determined in our study, several prospective studies with younger persons with schizophrenia found that depressive symptoms led to the worsening of quality of life indicators.33,34 For example, Fitzgerald et al.34 found a strong relationship between depression scale scores and quality of life in both cross-sectional and longitudinal analyses. Future studies should try to replicate such studies in older schizophrenia populations. Finally, our study generally confirmed earlier reports that depression in schizophrenia was not associated with any demographic variables, negative symptoms, or cognitive functioning. We also found no associations with acute stressors or household income. These findings differ from those of studies of depression in older persons in general, in which these variables have often been found to be important. Thus, our data suggest that depression in schizophrenia, although having several similar associations with variables linked to depression in older persons in general, may be less impacted by factors that affect depressed persons without schizophrenia. Although this study has several strengths (e.g., large multiracial sample of persons with schizophrenia who were age 55 and older, a theoretical model that concurrently examined 16 independent variables, and a comparison group), there are several limitations. For example, the study was conducted in one geographic area, the study used cross-sectional data, and schizophrenia respondents were not randomly selected but represented a convenience sample derived from a variety of residential and clinical sites serving this population. Furthermore, it was difficult to reliably distinguish persons with schizophrenia from those with schizoaffective disorder, and this may have contributed to the higher rates of depression among the schizophrenia sample. Finally, by using 16 variables in logistic regression analysis, there was an increased likelihood of a type
I error: that is, one of the variables might be found significant by chance. Using a more conservative p value of ⬍0.01 would have reduced the possibility of a type I error from 56% to 15%. However, this reduction would have eliminated four of the six significant variables (Table 3) and created the possibility of a type II error (i.e., power for a medium effect would have been reduced from 90% to 75%; the latter falling below the conventional standard of 80%). Because this was an exploratory study meant to lay the groundwork for longitudinal research and clinical interventions based on a theoretical model, we believed that retaining a p value of ⬍0.05 (thus avoiding the increased likelihood of a type II error) was appropriate in this instance. Our findings have important implications for clinicians and policy makers. Depending on causal direction, some variables associated with depression in schizophrenia may be alleviated by more vigorous treatment of depression. Only two-fifths of syndromally depressed persons with schizophrenia in our sample were taking antidepressants. It is not clear if any of them were receiving nonpharmacologic therapy specifically targeted toward their depression. Moreover, on the basis of our findings, medical interventions to improve physical well-being or psychosocial strategies to enhance social supports and coping strategies may also have the potential to alleviate depressive symptoms. Indeed, the latter two factors, especially intimacy and copes using medications, had the most powerful odds ratios among the significant variables. Finally, given the relationship between positive symptoms and depression, it may be possible to diminish depression by more aggressive treatment of the positive symptoms. The authors thank Dr. Robert Yaffee, Joseph Eimicke, Dr. Jeanne Teresi, Dr. Carol Magai, and Community Research Applications for their assistance. This study was supported by a grant from the National Institute of General Medical Sciences (SO6GM54650).
References 1. Zisook S, McAdams LA, Kuck J, et al: Depressive symptoms in schizophrenia. Am J Psychiatry 1999; 156:1736–1743 2. Zisook S, Nyer M, Kasckow J, et al: Depressive symptom patterns in patients with chronic schizophrenia and subsyndromal depression. Schizophr Res 2006; 86:226–233 3. Jin H, Zisook S, Palmer BW, et al: Association of depressive
Am J Geriatr Psychiatry 15:12, December 2007
symptoms with worse functioning in schizophrenia: a study in older outpatients. J Clin Psychiatry 2001; 62:797–803 4. Cohen CI, Talavera N, Hartung R: Depression among aging persons with schizophrenia who live in the community. Psychiatr Serv 1996; 47:601–607 5. Siris SG: Depression in schizophrenia: perspective in the era of
997
Depression in Older Adults With Schizophrenia Spectrum Disorders “atypical” antipsychotic agents. Am J Psychiatry 2000; 157:1379– 1389 6. Tennant C: Life events, stress and depression: a review of recent findings. Aust N Z J Psychiatry 2002; 36:173–182 7. O’Sullivan C: The psychosocial determinants of depression: a lifespan perspective. J Nerv Ment Dis 2004; 192:585–594 8. Krause N: Chronic financial strain, social support, and depressive symptoms among older adults. Psychol Aging 1987; 2:185–192 9. Krause N, Jay G, Liang J: Financial strain and psychological wellbeing among the American and Japanese elderly. Psychol Aging 1991; 6:170–181 10. Jongenelis K, Pot AM, Eisses AM, Beekman AT, Kluiter H, Ribbe MW: Prevalence and risk indicators of depression in elderly nursing home patients: the AGED study. J Affect Disord 2004; 83:135– 142 11. Moos RH, Schutte KK, Brennan PL, Moos BS: The interplay between life stressors and depressive symptoms among older adults. J Gerontol B Psychol Sci Soc Sci 2005; 60:P199–P206 12. Glass TA, Kasl SV, Berkman LF: Stressful life events and depressive symptoms among the elderly. Evidence from a prospective community study. J Aging Health 1997; 9:70–89 13. George LK: Social factors and depression in late life, in Diagnosis and Treatment of Depression in Late Life. Edited by Schneider LS, Reynolds CF, Lebowitz BD. Washington, DC, American Psychiatric Press, 1994 14. Jeste DV, Symonds LL, Harris MJ, et al: Nondementia nonpraecox dementia praecox? Late-onset schizophrenia. Am J Geriatr Psychiatry 1997; 5:302–317 15. Wilson L, Brass W: Brief assessment of the mental state in geriatric domiciliary practice: the usefulness of the mental status questionnaire. Age Ageing 1973; 2:92–101 16. Lawton M, Moss M, Fulcomer M, et al: A research and serviceoriented multilevel assessment instrument. J Gerontology 1982; 37:91–99 17. Chatters LM: Health disability and its consequences for subjective stress, in Aging in Black America. Edited by Jackson JS, Chatters LM, Taylor RJ. Newbury Park, CA, Sage Publications, 1993 18. Coblentz JM, Mattis S, Zingesser LH, et al: Presenile dementia. Clinical aspects and evaluation of cerebrospinal fluid dynamics. Arch Neurol 1973; 29:299–308 19. Sokolovsky J, Cohen CI: Toward a resolution of methodological dilemmas in network mapping. Schizophr Bull 1981; 7:109–116
998
20. Pearlin LI, Mullan JT, Semple SJ, Skaff MM: Caregiving and the stress process: an overview of concepts and their measures. Gerontologist 1990; 30:583–594 21. Strauss M: Measuring intrafamily conflict and violence: The conflict tactics (CY) scales. J Marriage Fam 1979; 41:75–88 22. Thompson LW, Futterman A, Gallagher D: Assessment of late-life depression. Psychopharmacol Bull 1988; 24:577–586 23. Nunally JC: Psychometric Theory. New York, McGraw-Hill, 1967 24. an der Heiden W, Konnecke R, Maurer K, Ropeter D, Hafner H: Depression in the long-term course of schizophrenia. Eur Arch Psychiatry Clin Neurosci 2005; 255:174–184 25. Koreen AR, Siris SG, Chakos M, et al: Depression in first-episode schizophrenia. Am J Psychiatry 1993; 150:1643–1648 26. Planansky K, Johnston R: Depressive syndrome in schizophrenia. Acta Psychiatr Scand 1978; 57:207–218 27. Lancon C, Auquier P, Reine G, Bernard D, Addington D: Relationships between depression and psychotic symptoms of schizophrenia during an acute episode and stable period. Schizophr Res 2001; 47:135–140 28. Koenig HG, George LK: Depression and physical disability outcomes in depressed medically ill hospitalized older adults. Am J Geriatr Psychiatry 1998; 6:230–247 29. Von Korff M, Ormel J, Katon W, Lin EH: Disability and depression among high utilizers of health care. A longitudinal analysis. Arch Gen Psychiatry 1992; 49:91–100 30. Vahia I, Bankole A, Diwan S, Cohen CI: Improving adequacy of medical care for older persons with schizophrenia, in Annual Meeting of the American Association of Geriatric Psychiatry. San Juan, PR, 2006 31. Reyes P, Ramirez P, Cohen CI: Depression, psychosis and functioning in older adults with schizophrenia, in 58th Institute for Psychiatric Services. New York, NY, American Psychiatric Association, 2006, p 177 32. Bankole A, Vahia I, Diwan S, et al: Enhancing Quality of Life in the Elderly With Schizophrenia. New York, NY, Institute for Psychiatric Services, 2006 33. Sands JR, Harrow M: Depression during the longitudinal course of schizophrenia. Schizophr Bull 1999; 25:157–171 34. Fitzgerald PB, de Castella AR, Filia K, et al: A longitudinal study of patient- and observer-rated quality of life in schizophrenia. Psychiatry Res 2003;119:55–62
Am J Geriatr Psychiatry 15:12, December 2007
Depression in Older Adults With Schizophrenia Spectrum Disorders: Prevalence and Associated Factors Shilpa Diwan, M.D., M.P.H., Carl I. Cohen, M.D., Azziza O. Bankole, M.D., Ipsit Vahia, M.D., Michelle Kehn, M.A., Paul M. Ramirez, Ph.D.
Rationale: Although depression is common in older adults with schizophrenia, it has not been well studied. The authors examine those factors that are related to depression in a multiracial urban sample of older persons with schizophrenia. Methods: The schizophrenia group consisted of 198 persons aged 55 or older who lived in the community and developed schizophrenia before age 45. Persons with substantial cognitive impairment were excluded from the study. A community comparison group (N⫽113) was recruited using randomly selected census tract data. The authors adapted George’s Social Antecedent Model of Depression, which consists of six categories comprising 16 independent variables, and used a dichotomous dependent variable based on a Center for Epidemiologic Studies Depression Scale cutoff score of ⱖ16. Results: The schizophrenia group had significantly more persons with clinical depression than the community comparison group (32% versus 11%, respectively; 2 ⫽ 28.23, df⫽1, p⫽0.001). Bivariate analysis revealed that eight of the 16 variables were significantly related to clinical depression in the schizophrenia group. In logistic regression, six variables retained significance: physical illness (odds ratio [OR]⫽1.60, 95% confidence interval [CI], 1.17–2.18), quality of life (OR⫽0.84, 95% CI, 0.76 – 0.93), presence of positive symptoms (OR⫽1.12, 95% CI, 1.02–1.21), proportion of confidants (OR⫽0.03, 95% CI, 0.01– 0.39), copes by using medications (OR⫽2.12, 95% CI, 1.08 – 4.13), and copes with conflicts by keeping calm (OR⫽1.34, 95% CI, 1.03–1.74). Conclusion: Consistent with earlier studies of schizophrenia in older persons, the authors found physical health, positive symptoms, and several nonclinical variables to be associated with depression. Potential points for intervention include strengthening social supports, improving physical well-being, more aggressive treatment of positive symptoms, and increasing the recognition and treatment of depression. (Am J Geriatr Psychiatry 2007; 15:991–998) Key Words: schizophrenia, depression, urban, multiracial
Received January 21, 2007; revised September 7, 2007; accepted September 10, 2007. From SUNY Downstate Medical Center (SD, CIC, AOB, IV) and Long Island University (MK, PMR), Brooklyn, New York. Send correspondence and reprint requests to Shilpa Diwan, M.D., M.P.H., Resident in Psychiatry-PGY IV, SUNY Downstate Medical Center, Box 1203, 450 Clarkson Avenue, Brooklyn, NY 11203; e-mail: [email protected] © 2007 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 15:12, December 2007
991
Depression in Older Adults With Schizophrenia Spectrum Disorders
D
epression is common among older persons with schizophrenia, with rates ranging from 44% to 75%.1– 4 Although depression is thought to have a potentially deleterious effect on the outcome of schizophrenia,5 there have been only four published studies that have looked specifically at depression among older persons with schizophrenia.1– 4 Among the studies that compared depressed and nondepressed older persons with schizophrenia,1,3,4 a consistent finding was the relationship between positive symptoms and depression. In these studies, depression was also found to be significantly associated with other variables such as quality of well-being, instrumental activities of daily living, various social network measures, physical limitations, lower income, and several factors on the SF-36, such as social functioning, role limitations, and general health.3,4 Lower age was found to be significant in one study but not in the other two studies.1,3 Earlier studies of depression among older persons with schizophrenia have had several limitations. In the studies from San Diego, California,1,3 the absence of a multivariate theoretical model made the findings more difficult to generalize to other sites and precluded an assessment of the overlapping effects of the various predictor variables on depression. Moreover, these studies did not include several variables that have been found to be important in studies of depression in older adults in general (e.g., life stressors, financial strain, social support, and coping styles).6 –12 Although the New York City study4 utilized a theoretical model and examined more variables, the number of variables was limited by the sample size (N⫽117) and by the use of instruments for measuring psychotic symptoms and depression that had not been used previously with schizophrenia populations. Moreover, this study lacked an adequate community comparison group with which to contrast the prevalence of depression. To address some of the limitations of earlier investigations, we have used an adaptation of George’s Social Antecedent Model of Depression.13 This is a conceptual model of the social precursors of psychiatric disorders based on the consensus in the social science, epidemiologic, and social psychiatry literature. The model consists of six categories of social and clinical factors that predict depression: demographic variables; early life events such as education level and childhood traumas; later events and
992
achievements such as income and marital status; social integration; vulnerability and protective factors such as chronic stressors and social supports; and provoking agents and coping efforts such as major life events and coping styles. Using a large multiracial urban sample of persons with schizophrenia who were age 55 and older and a community comparison group, we address the following questions: 1) What is the prevalence of clinical depression among older persons with schizophrenia and how does it compare with their age peers in the general population? 2) What factors are associated with depression among older persons with schizophrenia?
METHODS We recruited subjects aged 55 and older who lived in the community and developed schizophrenia before the age of 45 using stratified sampling in which we attempted to interview approximately one-half of the subjects from outpatient clinics and day programs and the other one-half from supported community residences, including sites with varying degrees of on-site supervision. Inclusion was based on a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, chart diagnosis of schizophrenia or schizoaffective disorder that was supplemented by a lifetime illness review adapted from Jeste et al.14 Thus, respondents were required to have one or more of the following: 1) symptoms such as social withdrawal, loss of interest in school or work, deterioration in hygiene and grooming, unusual behaviors, or outbursts of anger before age 45; 2) evidence of hallucinations, delusions, or disorganized thinking or behavior before age 45; 3) hospitalization for schizophrenia before age 45; and/or 4) treatment for schizophrenia before age 45. Subjects were excluded if they had medical conditions or trauma, seizures, dementia, head injury with unconsciousness of ⱖ30 minutes, mental retardation, substance abuse, or evidence of bipolar illness that could better account for the subject’s psychotic symptoms before age 45. We excluded persons with cognitive impairment too severe for completion of the questionnaire (i.e., defined as scores of ⬍5 on the Mental Status Questionnaire).15 A community comparison group was recruited using Wessex Census STF3 files of Kings County
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al. (Brooklyn, NY). Persons age 55 and older were extracted. A computer program was written to randomly select without replacement Census Bureau groupings of blocks (“block-groups”) as the primary sampling unit. Without replacement sampling was used to avoid unnecessary duplication of effort and waste of resources. An effort was made to interview all persons in a selected block-group by knocking on doors. To enhance response rates, subjects from the selected block-group were also recruited at senior centers and churches and through personal references. Persons were excluded if they reported a history of outpatient treatment or hospitalization for schizophrenia or schizoaffective disorder. Subjects were offered $75 for completing the 2.5hour interview. The rejection rate was 7% in the schizophrenia group and 48% of those contacted in the community group. The schizophrenia sample consisted of 198 persons with schizophrenia; 39% were living independently in the community, and 61% lived in supported community residences. From the original community sample (N⫽206), we selected 113 persons who more closely matched the schizophrenia sample for age, sex, race, and income.
TABLE 1.
Forty-nine percent of the schizophrenia group were women versus 49% of the community comparison group, and 35% of the schizophrenia group was African American versus 36% of the community comparison group (2 ⫽ 0.07, df ⫽1, p⫽0.70). The mean ages of the schizophrenia and community comparison groups were 61.5 years and 63.0 years, respectively (t⫽2.37, df⫽ 306, p⫽0.02). Thus, despite a small absolute difference in age between the two groups, this difference attained statistical significance. Median income for both groups fell within the same category ($7,000 –$12,999). Instruments We adapted George’s model that consisted of six categories comprising 16 independent variables. These variables were selected based on their theoretical importance as well as their having been identified in the literature as impacting on depression. The 16 independent variables (Table 1) were derived from the following scales: a “Physical Illness” score that represented the sum of seven health care items and 13 illness categories derived from the Multilevel
Bivariate Analysis of Variables for Depressed (CES-D Score, >16) and Nondepressed Older Persons With Schizophrenia Using George’s Social Antecedent Model
Variable
Nondepressed (N ⴝ 134)
Depressed (N ⴝ 64)
61.8 (5.5) 46 34
60.9 (5.7) 55 38
3.2 (1.97)
Demographic factor Age, years Female Black Event and achievement Household income Social integration Proportion of contacts considered intimates Vulnerability factor Physical illness Instrumental activities of daily living Dementia rating scale Positive symptoms Negative symptoms Quality of Life Index Provoking agent Acute stressors Coping strategy Copes by using medication Copes with conflicts by keeping calm Use of spiritualists or their products Takes antidepressants
2/t-Value
Df
p
0.97 1.23 0.29
193 1 1
0.33 0.27 0.59
3.1 (1.1)
0.27
191
0.79
0.3 (0.3)
0.2 (0.2)
2.45
177
0.02a
1.0 (1.1) 22.3 (3.8) 127.2 (13.9) 11.5 (5.3) 11.6 (6.0) 23.3 (4.6)
1.9 (1.8) 21.9 (4.1) 128.7 (11.2) 14.8 (7.4) 12.7 (6.0) 18.4 (5.1)
⫺3.35 0.68 ⫺0.85 ⫺3.25 ⫺1.15 6.64
84 196 150 96 196 193
0.001a 0.50 0.40 0.002a 0.25 0.001a
6.7 (5.8)
12.9 (8.2)
⫺5.47
94
0.001a
1.2 (0.7) 2.2 (1.6) 34 27
1.4 (0.6) 2.8 (1.7) 63 39
⫺2.37 ⫺2.46 14.78 3.02
193 196 1 1
0.02a 0.02a 0.001a 0.08
Notes: Unless stated otherwise, data are mean percentage (SD). For t-tests, the unequal variance t-test was used when findings of the Levene test were significant. CES-D: Center for Epidemiologic Studies Depression Scale. a Author will explain.
Am J Geriatr Psychiatry 15:12, December 2007
993
Depression in Older Adults With Schizophrenia Spectrum Disorders Assessment Inventory; Physical Health Status (lower scores indicate worse health)16; instrumental activities of daily living (lower scores indicate more impairment)16; Acute Stressors Scale17 that examined 11 life events that may have occurred in the past month; the Dementia Rating Scale,18 which assesses five areas of cognitive functioning (namely, attention, initiation and perseveration, construction, conceptualization, and memory [higher scores indicate better functioning]); the Network Analysis Profile,19 from which we used the confidant variable (i.e., shares intimate thoughts); the Positive and Negative Syndrome Scale (PANSS), which consists of seven items assessing positive symptoms, seven items assessing negative symptoms, and 16 items assessing general psychopathology; the Quality of Life Index, which consists of 64 Likert-type items divided into two sections (assessing satisfaction or importance in various life domains); the Pearlin Coping Scale,20 a 23item scale used to assess coping strategies to various stressors that was subdivided into eight subscales using principal components analysis with varimax rotation; and the Conflict Tactics Scale,21 which was subdivided into seven subscales using principal components analysis with varimax rotation. From each of the latter two measures, we selected the subscale that had the strongest association with the dependent variable: namely, Use of Medication and/or Professional Assistance subscale from the Pearlin Coping Scale and the Keeping Calm subscale from the Conflict Tactics Scale. The latter consisted of three items: discussed situation calmly, got information, and brought someone to help. “Spiritualism” was rated positive if the respondent saw a spiritualist or purchased items from a religious shop. The dependent variable was “syndromal depression,” which was based on the traditional cutoff score of ⱖ16 on the Center for Epidemiologic Studies Depression Scale (CES-D).22 We also made comparisons with the community sample using “subsyndromal depression,” which was derived from our earlier work and based on a CES-D score of 8 to 15.22 The internal reliability (Cronbach ␣) scores for the scales were as follows: Quality of Life Index, 0.97; CES-D, 0.88; PANSS positive scale, 0.83; PANSS negative scale, 0.78; Acute Stressors Scale, 0.88; instrumental activities of daily living, 0.74; Dementia Rating Scale, 0.89; Pearlin Coping Scale, 0.75; and
994
Conflict Tactics Scale, 0.82. All scales attained a recommended ␣ of ⱖ0.60.23 Interviewers were trained using audiotapes and videotapes, and they were generally matched to respondents from similar ethnic backgrounds. Interviewers were periodically monitored using audiotapes. The intraclass correlations ranged from 0.79 to 0.99 on the various scales. Raters had a mean of 99% and 94% agreement with expert ratings (CIC) on the Physical Illness score and the Network Analysis Profile contact scores, respectively. Data Analysis Initially, we conducted bivariate analyses to examine the relationship between independent and dependent variables; t-tests and 2 analyses were used for continuous and categorical independent variables, respectively. For instruments that generated potentially more than one variable for inclusion in the model, we conducted preliminary bivariate analyses and selected variables from that measure that had substantial associations (p ⬍0.10) with the dependent variable (data from preliminary analyses are not shown). In instances in which several variables attained this cutoff, we selected the variable with the lowest p value. Using this method, we selected the two coping variables described above and a social network variable (proportion of contacts considered intimates). Logistic regression analysis was used to examine those variables associated with depression. There was no evidence of collinearity among the independent variables.
RESULTS As seen in Table 2, 32%, 29%, and 39% of persons with schizophrenia had syndromal depression, subsyndromal depression, and no depression, respectively. In the community comparison group, 11%, 19%, and 71% of persons had syndromal depression, subsyndromal depression, and no depression, respectively. The difference between the groups was significant (2 ⫽31.36, df⫽2, p ⬍0.001). There was no statistical difference in syndromally depressed persons who reported taking medications for depression in the schizophrenia group (39%) and the community comparison group (42%) (2 ⫽0.00, df⫽1, p⫽1.0).
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al.
TABLE 2.
Persons in the Community and Schizophrenia Groups Who Had Various Levels of Depression
Group
Syndromal Depression, CES-D Score, >16
Subsyndromal Depression, CES-D Score, 8 –15
No Depression, CES-D Score, <8
11% 32%
19% 29%
71% 39%
Community comparison (N ⫽ 113) Schizophrenia (N ⫽ 198) Note: 2 ⫽ 31.36, df ⫽ 2, p ⬍0.001.
In the schizophrenia group, 44% had psychotic symptoms at the time of interview, which was defined as scores of ⬎2 on one or more on the PANSS positive scale items associated with psychosis (i.e., hallucinations, delusions, or disorganized thinking). Among persons with psychotic symptoms, 43% had syndromal depression; among persons without psychotic symptoms, 25% had syndromal depression. Bivariate analysis revealed that eight of the 16 variables in the model were significantly related to syndromal depression in the schizophrenia group (Table 1). Depression was associated with more physical disorders, more acute stressors, more positive symptoms, lower proportion of confidants in their social network, greater use of medications/ health professionals as a coping strategy for stress, greater use of keeping calm when having conflicts with others, lower quality of life score, and greater use of spiritualists or their products. When the model was tested in logistic regression analysis, six variables retained significance: more physical disorders, presence of positive symptoms, smaller proportion of confidants, copes by using medications/health professionals, copes by keeping calm, and a lower quality of life score (Table 3). The overall model was statistically significant (2 ⫽ 84.99, df⫽16, p ⫽0.001).
TABLE 3.
Logistic Regression Analysis of Variables Predicting Depression (CES-D Score, >16) in Older Persons With Schizophrenia
Variable Demographic factor Age Female Black Event and achievement Household income Social integration Proportion of intimate contacts Vulnerability factor No. of physical illnesses Instrumental activities of daily living Dementia Rating Scale Positive symptoms Negative symptoms Quality of Life Index Provoking agent Acute stressors Coping strategy Copes by using medication Copes with conflicts by keeping calm Use of spiritualists or their products Use of antidepressants
Odds Ratio
95% CI
1.01 1.20 0.90
0.94–1.09 0.51–2.78 0.35–2.28
0.73 0.68 0.81
1.18
0.93–1.48
0.17
0.03
0.01–0.39
0.01a
1.60 1.05
1.17–2.18 0.93–1.18
0.003a 0.44
1.02 1.12 0.99 0.84
0.99–1.06 1.02–1.21 0.91–1.08 0.76–0.93
0.23 0.04a 0.81 0.001a
1.04
0.96–1.11
0.34
2.12 1.34
1.08–4.13 1.03–1.74
0.03a 0.03a
2.31
0.94–5.67
0.07
0.88
0.35–2.18
0.78
p
Notes: Overall model: ⫽ 84.99, df ⫽ 16, p ⫽ 0.001. Odds ratios of ⬎1 are associated with depression; odds ratios of ⬍1 are associated with absence of depression. Wald 2 tests with df ⫽ 1 were used to obtain p values. CI: confidence interval; CES-D: Center for Epidemiologic Studies Depression Scale. a Author will explain. 2
DISCUSSION The prevalence of syndromal depression was threefold greater in the schizophrenia group than in the community comparison group, and the prevalence of any level of depression in the schizophrenia group was twice that in the community comparison group. Moreover, the proportion of persons with schizophrenia who had subsyndromal depression and syndromal depression (61%) was similar to the levels found among older persons with schizophrenia in San Diego, California (68%),1,3 although our group
Am J Geriatr Psychiatry 15:12, December 2007
had more severe depression than their group (32% versus 13%, respectively). Thus, our study confirms earlier findings that older people with schizophrenia have substantially higher rates of depression than their age peers in the general population. Our findings were remarkably consistent with those of an der Heiden et al;24 in their longitudinal investigation of a younger schizophrenia cohort, about one-third of persons had depressive symptoms at any point in time, and the prevalence of depression
995
Depression in Older Adults With Schizophrenia Spectrum Disorders rose to one-half during exacerbation or recurrence of psychoses. We found that 32% of our sample had syndromal depression but that 43% of persons with psychosis had syndromal depression. On the basis of work with younger populations, several investigators have proposed that depressive features may be a “core” feature of schizophrenia.24 –26 Our findings for older persons would suggest that this may be premature. At a minimum, there should be a consensus on how depression is defined (i.e., subsyndromal or syndromal) and what level of prevalence would make it a “core” feature (e.g., in our sample, 39% had no clinically relevant depression). These concerns notwithstanding, our data strongly support the argument that depression remains a common associated feature of schizophrenia throughout the life span. With respect to our overall theoretical model, we found that eight variables were significant in bivariate analysis and that six retained significance in logistic regression analysis. Four of the six categories had one significant variable in bivariate analysis, and three categories (vulnerability, social integration, and coping) had one significant variable in logistic regression analysis. Although the overall model was statistically significant, the fact that 10 variables did not retain independent significance in the final analysis suggests that trimming of the model or use of alternative variables may be indicated. With respect to specific variables in the vulnerability category of the model, we identified a number of items that were associated with higher rates of depression, suggesting potential points for clinical intervention. Most notably, our data support earlier findings that depression is associated with positive symptoms.1,27 Although significant, the odds ratio for this positive symptoms was low. However, the consistency of the relationship across several studies suggests that it is a meaningful finding that may have clinical implications. Because of the cross-sectional nature of this study, it is unclear whether depression exacerbates positive symptoms, whether positive symptoms result in more depression, or whether there is a mutually reinforcing effect. Another important finding is the relationship among physical illness, disability, and depression in older persons with schizophrenia. Although there is some variability across studies (i.e., in some studies, depression was related to physical or activity limitations and not the number of physical disorders,
996
whereas in this study the obverse was obtained), there is a strong indication that like other older persons, physical illness and/or disability is linked to depression. Here again, causal direction cannot be determined, although other researchers have suggested that the relationship between physical illness and depression is probably bidirectional.28,29 In recent years, there has been an increased recognition that older persons with schizophrenia are not receiving adequate medical treatment despite having access to care.30 The interplay of physical illness and depression should be examined with respect to its impact on healthcare in older persons with schizophrenia. Our findings highlight several other important variables that may play a role in affecting depression that have not been considered previously. For example, within the coping category of the model, we found that certain coping strategies were more commonly used by depressed persons with schizophrenia, although it is not clear whether these strategies serve to diminish or exacerbate the symptoms. For example, keeping calm to deal with interpersonal conflicts may merely be a preferred coping strategy among depressed persons with schizophrenia or it may serve to further worsen depressive symptoms. Similarly, within the social integration category of the model, we found that depressed persons had a smaller proportion of network confidants, and although paucity of network support has been thought to exacerbate depression in older persons in general,31 we could not accurately determine causal direction. However, it is evident that future studies need to assess the impact of various social and coping strategies as well as the potential impact of alternative treatments such as the use of spiritualists or their products, the latter being significantly associated with depression in our sample in bivariate analysis and marginally significant in logistic regression analysis. Our findings are consistent with those of Jin et al.3 that depression is associated with lower perceived quality of life. We demonstrated that this effect was independent of other variables. Thus, in addition to its associations with physical impairment and psychotic symptoms, depression is associated with diminished perception of life quality. A thorny issue is whether quality of life is really a measure of depression. In another study of this population, we found that the Quality of Life Index accounted for 28% of the explained variance in the CES-D.32 Thus, al-
Am J Geriatr Psychiatry 15:12, December 2007
Diwan et al. though there is overlap with depressive symptoms, quality of life is not identical with depression. Although causality could not be determined in our study, several prospective studies with younger persons with schizophrenia found that depressive symptoms led to the worsening of quality of life indicators.33,34 For example, Fitzgerald et al.34 found a strong relationship between depression scale scores and quality of life in both cross-sectional and longitudinal analyses. Future studies should try to replicate such studies in older schizophrenia populations. Finally, our study generally confirmed earlier reports that depression in schizophrenia was not associated with any demographic variables, negative symptoms, or cognitive functioning. We also found no associations with acute stressors or household income. These findings differ from those of studies of depression in older persons in general, in which these variables have often been found to be important. Thus, our data suggest that depression in schizophrenia, although having several similar associations with variables linked to depression in older persons in general, may be less impacted by factors that affect depressed persons without schizophrenia. Although this study has several strengths (e.g., large multiracial sample of persons with schizophrenia who were age 55 and older, a theoretical model that concurrently examined 16 independent variables, and a comparison group), there are several limitations. For example, the study was conducted in one geographic area, the study used cross-sectional data, and schizophrenia respondents were not randomly selected but represented a convenience sample derived from a variety of residential and clinical sites serving this population. Furthermore, it was difficult to reliably distinguish persons with schizophrenia from those with schizoaffective disorder, and this may have contributed to the higher rates of depression among the schizophrenia sample. Finally, by using 16 variables in logistic regression analysis, there was an increased likelihood of a type
I error: that is, one of the variables might be found significant by chance. Using a more conservative p value of ⬍0.01 would have reduced the possibility of a type I error from 56% to 15%. However, this reduction would have eliminated four of the six significant variables (Table 3) and created the possibility of a type II error (i.e., power for a medium effect would have been reduced from 90% to 75%; the latter falling below the conventional standard of 80%). Because this was an exploratory study meant to lay the groundwork for longitudinal research and clinical interventions based on a theoretical model, we believed that retaining a p value of ⬍0.05 (thus avoiding the increased likelihood of a type II error) was appropriate in this instance. Our findings have important implications for clinicians and policy makers. Depending on causal direction, some variables associated with depression in schizophrenia may be alleviated by more vigorous treatment of depression. Only two-fifths of syndromally depressed persons with schizophrenia in our sample were taking antidepressants. It is not clear if any of them were receiving nonpharmacologic therapy specifically targeted toward their depression. Moreover, on the basis of our findings, medical interventions to improve physical well-being or psychosocial strategies to enhance social supports and coping strategies may also have the potential to alleviate depressive symptoms. Indeed, the latter two factors, especially intimacy and copes using medications, had the most powerful odds ratios among the significant variables. Finally, given the relationship between positive symptoms and depression, it may be possible to diminish depression by more aggressive treatment of the positive symptoms. The authors thank Dr. Robert Yaffee, Joseph Eimicke, Dr. Jeanne Teresi, Dr. Carol Magai, and Community Research Applications for their assistance. This study was supported by a grant from the National Institute of General Medical Sciences (SO6GM54650).
References 1. Zisook S, McAdams LA, Kuck J, et al: Depressive symptoms in schizophrenia. Am J Psychiatry 1999; 156:1736–1743 2. Zisook S, Nyer M, Kasckow J, et al: Depressive symptom patterns in patients with chronic schizophrenia and subsyndromal depression. Schizophr Res 2006; 86:226–233 3. Jin H, Zisook S, Palmer BW, et al: Association of depressive
Am J Geriatr Psychiatry 15:12, December 2007
symptoms with worse functioning in schizophrenia: a study in older outpatients. J Clin Psychiatry 2001; 62:797–803 4. Cohen CI, Talavera N, Hartung R: Depression among aging persons with schizophrenia who live in the community. Psychiatr Serv 1996; 47:601–607 5. Siris SG: Depression in schizophrenia: perspective in the era of
997
Depression in Older Adults With Schizophrenia Spectrum Disorders “atypical” antipsychotic agents. Am J Psychiatry 2000; 157:1379– 1389 6. Tennant C: Life events, stress and depression: a review of recent findings. Aust N Z J Psychiatry 2002; 36:173–182 7. O’Sullivan C: The psychosocial determinants of depression: a lifespan perspective. J Nerv Ment Dis 2004; 192:585–594 8. Krause N: Chronic financial strain, social support, and depressive symptoms among older adults. Psychol Aging 1987; 2:185–192 9. Krause N, Jay G, Liang J: Financial strain and psychological wellbeing among the American and Japanese elderly. Psychol Aging 1991; 6:170–181 10. Jongenelis K, Pot AM, Eisses AM, Beekman AT, Kluiter H, Ribbe MW: Prevalence and risk indicators of depression in elderly nursing home patients: the AGED study. J Affect Disord 2004; 83:135– 142 11. Moos RH, Schutte KK, Brennan PL, Moos BS: The interplay between life stressors and depressive symptoms among older adults. J Gerontol B Psychol Sci Soc Sci 2005; 60:P199–P206 12. Glass TA, Kasl SV, Berkman LF: Stressful life events and depressive symptoms among the elderly. Evidence from a prospective community study. J Aging Health 1997; 9:70–89 13. George LK: Social factors and depression in late life, in Diagnosis and Treatment of Depression in Late Life. Edited by Schneider LS, Reynolds CF, Lebowitz BD. Washington, DC, American Psychiatric Press, 1994 14. Jeste DV, Symonds LL, Harris MJ, et al: Nondementia nonpraecox dementia praecox? Late-onset schizophrenia. Am J Geriatr Psychiatry 1997; 5:302–317 15. Wilson L, Brass W: Brief assessment of the mental state in geriatric domiciliary practice: the usefulness of the mental status questionnaire. Age Ageing 1973; 2:92–101 16. Lawton M, Moss M, Fulcomer M, et al: A research and serviceoriented multilevel assessment instrument. J Gerontology 1982; 37:91–99 17. Chatters LM: Health disability and its consequences for subjective stress, in Aging in Black America. Edited by Jackson JS, Chatters LM, Taylor RJ. Newbury Park, CA, Sage Publications, 1993 18. Coblentz JM, Mattis S, Zingesser LH, et al: Presenile dementia. Clinical aspects and evaluation of cerebrospinal fluid dynamics. Arch Neurol 1973; 29:299–308 19. Sokolovsky J, Cohen CI: Toward a resolution of methodological dilemmas in network mapping. Schizophr Bull 1981; 7:109–116
998
20. Pearlin LI, Mullan JT, Semple SJ, Skaff MM: Caregiving and the stress process: an overview of concepts and their measures. Gerontologist 1990; 30:583–594 21. Strauss M: Measuring intrafamily conflict and violence: The conflict tactics (CY) scales. J Marriage Fam 1979; 41:75–88 22. Thompson LW, Futterman A, Gallagher D: Assessment of late-life depression. Psychopharmacol Bull 1988; 24:577–586 23. Nunally JC: Psychometric Theory. New York, McGraw-Hill, 1967 24. an der Heiden W, Konnecke R, Maurer K, Ropeter D, Hafner H: Depression in the long-term course of schizophrenia. Eur Arch Psychiatry Clin Neurosci 2005; 255:174–184 25. Koreen AR, Siris SG, Chakos M, et al: Depression in first-episode schizophrenia. Am J Psychiatry 1993; 150:1643–1648 26. Planansky K, Johnston R: Depressive syndrome in schizophrenia. Acta Psychiatr Scand 1978; 57:207–218 27. Lancon C, Auquier P, Reine G, Bernard D, Addington D: Relationships between depression and psychotic symptoms of schizophrenia during an acute episode and stable period. Schizophr Res 2001; 47:135–140 28. Koenig HG, George LK: Depression and physical disability outcomes in depressed medically ill hospitalized older adults. Am J Geriatr Psychiatry 1998; 6:230–247 29. Von Korff M, Ormel J, Katon W, Lin EH: Disability and depression among high utilizers of health care. A longitudinal analysis. Arch Gen Psychiatry 1992; 49:91–100 30. Vahia I, Bankole A, Diwan S, Cohen CI: Improving adequacy of medical care for older persons with schizophrenia, in Annual Meeting of the American Association of Geriatric Psychiatry. San Juan, PR, 2006 31. Reyes P, Ramirez P, Cohen CI: Depression, psychosis and functioning in older adults with schizophrenia, in 58th Institute for Psychiatric Services. New York, NY, American Psychiatric Association, 2006, p 177 32. Bankole A, Vahia I, Diwan S, et al: Enhancing Quality of Life in the Elderly With Schizophrenia. New York, NY, Institute for Psychiatric Services, 2006 33. Sands JR, Harrow M: Depression during the longitudinal course of schizophrenia. Schizophr Bull 1999; 25:157–171 34. Fitzgerald PB, de Castella AR, Filia K, et al: A longitudinal study of patient- and observer-rated quality of life in schizophrenia. Psychiatry Res 2003;119:55–62
Am J Geriatr Psychiatry 15:12, December 2007