| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
Departments of 1 Family, Consumer and Human Development, 2 Psychology, and 3 Mathematics and Statistics, and 4 Center for Epidemiologic Studies, Utah State University, Logan.
5 Institute of Clinical Neurosciences, Göteborg University, Sweden.
6 Department of Mental Hygiene, the Johns Hopkins University, Baltimore, Maryland.
7 Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle.
8 Duke University Medical Center, Durham, North Carolina.
Address correspondence to Maria C. Norton, PhD, Department of Family, Consumer and Human Development, Cache County Study on Memory in Aging, Utah State University, 4440 Old Main Hill, Logan, UT 84322-4440. E-mail: mnorton{at}cc.usu.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
DEPRESSION in old age is a public health concern, with reported prevalence rates of major depression ranging from 1% to 4% in community samples (Regier et al., 1993
; Ritchie et al., 2004). Depression in later life is often underdiagnosed (Ginsberg, 2005) and undertreated (Mulsant & Ganguli, 1999). The adverse consequences of depression late in life include high health care utilization and poor quality of life and productivity, increased risk of suicide, and increased nonsuicide mortality (P
lsson & Skoog, 1997). The suicide rate in the 65-year-old and older age group has been increasing since 1980 (Lebowitz et al., 1997). In addition, studies generally report that older women are 1.5 to 2.0 times as likely as men to suffer from major depression (Gallo & Lebowitz, 1999; Steffens et al., 2000).
Older adults have numerous psychosocial risk factors for depression, including death of loved ones, financial problems, and medical illnesses, that may have differential effects on men and women (Paykel, 1991). Supportive social networks may offset some of their association with depression (George, Blazer, Hughes, & Fowler, 1989). Some studies also suggest that religious involvement may reduce depressive symptoms by offering coping strategies through spiritual, intellectual, and social avenues (Husaini, Blasi, & Miller, 1999; Strawbridge, Shema, Cohen, Roberts, & Kaplan, 1998), but other studies have found no such effect (Courtenay, Poon, Martin, Clayton, & Johnson, 1992; Spendlove, West, & Stanish, 1984). The association between religious involvement and depression has yet to be definitively established, because some studies have used small or nonrepresentative samples of individuals (e.g., medically ill or nursing home patients). Further, not all studies have controlled for potential confounding factors such as education, social supports, physical health, and functional abilities.
Depression rates have also been shown to differ by religious affiliation. In a study by Boey (2003), Catholics and Buddhists reported less depression than Protestants among elderly women in Hong Kong (M = 3.5, 3.9, and 6.9, respectively, on the Geriatric Depression Scale–15). Affiliation also can affect the female-to-male ratio for depression, shown to be 1:1 among Jews, compared with the more common 2:1 among Catholics and Protestants (Levav, Kohn, Golding, & Weissman, 1997). Further, Braam and colleagues (2001) argued that a strong religious climate protects against depression in later life, particularly if it is pervasive throughout cultural life. Thus, the study of the effect of religious involvement on depression in populations with a predominant religious group will expand our understanding of this association.
Here we report an analysis of the relation between religious involvement and current major depression in a large community-based sample of older adults participating in the Cache County (Utah) Study on Memory Health and Aging (Breitner et al., 1999). An unusual characteristic of the Cache County population is that approximately 90% are members of the Church of Jesus Christ of Latter-Day Saints (LDS). LDS church members tend to have high religious involvement, and the church prohibits the use of alcohol or tobacco. Its members in Cache County therefore have low rates of mortality before the age of 85 and have a conditional life expectancy for men at age 65 that is the highest in the United States (Murray, Michaud, & McKenna, 1998), exceeding national norms by almost 10 years (Manton, Stallard, & Tolley, 1991). The prevalence of major depression among those aged 65 and older in this population has been previously reported as 4.3% in women and 3.0% in men (Steffens et al., 2000). This is somewhat higher than the 4.0% in women and 1.8% in men reported in a community study in France (Ritchie et al., 2004), but it is much higher than the 1.4% in women and 0.4% in men in the NIMH Epidemiological Catchment Area Study (Koenig & Blazer, 1992).
The general finding that religious involvement is inversely associated with depression, along with somewhat higher rates of major depression among residents of a community with a predominant religious group (where members generally exhibit high religiosity), begs further study into differences between individuals who demonstrate varying degrees of adherence to their religious beliefs within such a cultural environment.
Our purpose in this study was to examine the effect of religious involvement on major depression risk among older adults in a rural county in northern Utah composed primarily of LDS church members. We hypothesized that more frequent church attendance would decrease depression, with a stronger effect for women and for members of the predominant group. Models tested the effect of religious involvement after adjustment for important confounding variables known to affect depression risk.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Briefly, the study approached all Cache County residents aged 65 and older as of January 1, 1995 (N = 5,677). We enrolled 5,092 participants (90%), 176 of whom resided in nursing homes or assisted living facilities. We excluded 355 individuals who we identified as having prevalent dementia by use of a comprehensive multistage case ascertainment protocol (Breitner et al.), 190 individuals who for other reasons could not provide direct interview responses, and 57 others who endorsed at least one depressive symptom but did not meet the criteria for major depression as established by the American Psychiatric Association's (1994) Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV; Steffens et al., 2000). The remaining 4,490 individuals were eligible to be in the present study, but 22 were missing at least one variable; this yielded a final sample size of 4,468. Trained interviewers visited participants at their place of residence and conducted an in-person interview. We obtained all measures from this interview, except for social network, which we collected from a mail-in questionnaire (80% return rate). We obtained informed consent for all participants, and all procedures received approval by the Institutional Review Boards of Utah State University, Duke University, and the Johns Hopkins University.
Measurement
We ascertained current major depression by using a modified version of the depression section of the Diagnostic Interview Schedule (DIS). The DIS has excellent reliability and validity and helps to establish DSM-IV diagnosis of major depression (Robins, Helzer, Croughan, & Ratcliff, 1981). Interviewers asked participants endorsing a history of two or more consecutive weeks of sadness, anhedonia, or irritability whether they were currently still experiencing these symptoms; if they were, interviewers also asked them about symptoms of appetite or weight change, sleep or concentration difficulties, restlessness, diminished energy level, guilt, or suicidal ideation. Although some depression studies analyze symptom counts as the outcome of interest, we focused on major depression because of its greater clinical significance. Further, the skip-out protocol for those not endorsing sadness, anhedonia, or irritability precluded an examination of subsyndromal or milder forms of depression. We identified instances of current major depression when individuals endorsed five or more symptoms including depressed mood or loss of interest or pleasure. The comparison group comprised those with no current depressive disorder.
We asked participants to identify their religious affiliation, with these available responses: Catholic, Protestant, LDS, Jewish, other, and no religion. Because of the high frequency of LDS affiliation, we dichotomized these responses into LDS (91.9%) versus all others. We coded frequency of attendance at religious services and activities as never, less than once per month, 1 to 2 times per month, weekly, and more than once per week. Religious compliance combined attending church at least weekly, currently being a nonsmoker (no cigarettes, cigars, pipes, chewing tobacco, or snuff) and being a nondrinker (averaging fewer than 2 or more alcoholic beverages per week), three behaviors fundamental to LDS doctrine.
We measured social support from the mail-in questionnaire according to reported frequency of two types of contact—social gatherings with family and friends and participation in social clubs and volunteering. We coded each from 1 (rarely or never) to 6 (usually every day). We combined these two variables into a single measure indicating social contacts at least twice weekly, as this represented the median level of contact and would capture social involvement in excess of a single, possibly routine, activity occurring only weekly.
Lifetime health status was the number of the following vascular health problems endorsed: stroke; transient ischemic attack; hypertension; hyperlipidemia; coronary bypass surgery; heart attack, myocardial infarction, or coronary thrombosis; and diabetes mellitus. The vascular health problems score ranged from 0 to 7 (M = 1.27, SD = 1.17). Functional impairment was the number of activities of daily living (ADLs) for which the participant reported needing assistance, including both personal care and instrumental ADLs. The resulting ADL problems score ranged from 0 to 14 (M = 1.74, SD = 1.60). The number of current medications (an indirect indicator of health) ranged from 0 to 25 (M = 5.01, SD = 3.46) and excluded antidepressants, antipsychotics, sedatives, and hypnotics.
Quality Assurance Measures
We taped a 10% sample of interviews (with consent) and reviewed them to ensure that the interviews were conducted according to specifications, and quality assurance specialists checked responses for consistency. To check the quality assurance on categorization of depression, a study geropsychiatrist (D. Steffens) reviewed each interview in which depression was endorsed.
Statistical Analyses
We first tested for differences between men and women on several health status and demographic factors, using the independent-samples t test. We then explored the marginal associations between depression and health status, religious involvement, and psychosocial variables. We used chi-square tests of independence for categorical factors and independent-samples t test for continuous-type factors.
We then used multiple logistic regression analyses to account for potentially confounding factors with respect to the association between religious involvement and depression. We built models sequentially, first including the effect of gender, given the higher prevalence of depression in women. In the second model we added demographic and health variables, followed by the religious involvement variables in the third model. We tested the interaction between gender and religious involvement in the fourth model. Where the interaction was significant, we calculated within-strata odds ratios (ORs), which were adjusted for all other variables. Finally, in the subset returning the questionnaire, we added the social network measure to evaluate the extent to which the association between church attendance and depression may be accounted for by a possible mediating effect of social network. We tested religious involvement both as the single indicator, church attendance, and as the composite indicator, religious compliance, resulting in two sets of logistic regression models. In all logistic regression models, we used the likelihood-ratio test to compare nested models. We computed reported ORs by using the fitted logistic regression model parameters, with corresponding 95% Wald confidence intervals (CIs). We computed all statistical analyses by using the Statistical Package for the Social Sciences, version 12.0.1.
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
2(1, N = 4,468) = 16.69, p <.001, and were more likely to be married (70% vs 58%), 2(3, N = 4,468) = 53.05, p <.001, have higher education (13.3 vs 12.9 years), t(4,466) = –4.06, p <.001, younger age (75 vs 77 years), t(4,466) = 7.87, p <.001, more frequent church attendance (3.7 vs 3.4 on a scale of 1–5), t(4,466) = –6.98, p <.001, fewer problems with ADLs (1.6 vs 2.3 on a scale of 0–14), t(4,466) = 8.43, p <.001, and fewer lifetime vascular health problems (1.2 vs 1.3 on a scale of 0–7), t(4,466) = 2.04, p =.041. However, they were equally likely to be a LDS member (92%), 2(1, N = 4,468) = 0.122, p =.727, male (57%), 2(1, N = 4,468) = 0.075, p =.784, a current smoker (2.4% vs 2.8%), 2(1, N = 4,468) = 0.63, p =.429), and a current drinker (4.1% vs 3.5%), 2(1, N = 4,459) = 0.72, p =.395), and they had equivalent number of medications (5.0), t(4,466) = –0.53, p =.596).
Individuals whose church attendance was weekly or more often were more likely to be married (70% vs 61%), 2(3, N = 4,399) = 67.29, p <.001); they reported more frequent social network contact (56% vs 49%), 2 (1, N = 3,499) = 13.18, p <.001), higher education (13.5 vs 12.7 years), t(4,397) = –10.08, p <.001), younger age (74.7 vs 75.8 years), t(4,397) = 4.65, p <.001), fewer ADL problems (1.6 vs 2.1), t(4,397) = 8.63, p <.001), and fewer vascular health problems (1.2 vs 1.4), t(4,397) = 5.10, p <.001), but a comparable number of medications (5.0 vs 5.1), t(4,397) = 1.57, p =.118), compared with those whose church attendance was less often than weekly. Church affiliation (LDS vs non-LDS) was significantly associated with religious involvement: 2(4, N = 4,480) = 393.0, p <.001, with 76% of LDS members attending church at least weekly and only 34% of non-LDS members attending church this often.
Gender Differences
Compared with men, women were significantly older (M = 75.5, SD = 7.0 for women; M = 74.4, SD = 6.7 for men; p <.001), reported lower education level (M = 12.8, SD = 2.3 for women; M = 13.9, SD = 3.4 for men; p <.001), equivalent number of vascular health problems (M = 1.3, SD = 1.1 for women; M = 1.3, SD = 1.3 for men; p =.450), more ADL problems (M = 1.9, SD = 1.7 for women; M = 1.5, SD = 1.5 for men; p <.001), and more medications (M = 5.5, SD = 3.5 for women; M = 4.3, SD = 3.3 for men; p <.001). Weekly contact with social network was reported by 60% of women but only 40% of men (p <.001).
Bivariate Associations
In bivariate analyses, being of the male gender, 2(1, N = 4,490) = 10.1, p =.001, having a higher level of education (p =.001), being married (p <.001), attending church more frequently, 2(4, N = 4,480) = 34.6, p <.001, and having twice weekly social contacts (p =.044) were related to lower frequency of major depression. Conversely, having more ADL problems (p =.001), taking more medications, and having more vascular health problems (p <.001 for both) were associated with higher depression rates. Among the nondepressed respondents, 24% reported taking at least one psychotropic medication. Across church attendance levels among men, depression rates were typically at 3%, except for those attending church infrequently (< 1 per month), in which 6% had depression. More dramatic differences in depression rates between church attendance levels were observed in women, with a low of 3% among those attending more often than weekly to a high of 13% among those never attending church. Similarly, compliant versus noncompliant men had comparable rates of depression (3% vs 4%, respectively), whereas these rates were more divergent among women (3.5% vs 9%, respectively; see Table 1).
|
|
The fourth model tested a Gender x Church Attendance interaction, which was significant (p =.043). Within-strata ORs revealed that, for women, attendance less than monthly (OR = 0.21, 95% CI = 0.06–0.70), weekly (OR = 0.37, 95% CI = 0.14–0.97), and more than weekly (OR = 0.25, 95% CI = 0.08–0.76) were all associated with significantly less depression. Sporadic attendance at 1 to 2 times per month was not significant (OR = 0.79, 95% CI = 0.19–3.21). Among men, the trend was in the opposite direction. Men attending church less often than monthly (OR = 4.88, 95% CI = 1.43–16.7), weekly (OR = 2.69, 95% CI = 1.04–7.00), and more than weekly (OR = 3.99, 95% CI = 1.31–12.1) all reported depression significantly more often. Again, sporadic attendance at 1 to 2 times per month was nonsignificant for men (OR = 1.27, 95% CI = 0.31–5.14). The significance levels for each of the remaining variables in the model were essentially unchanged with addition of the interaction term.
Finally, in Model 5, having social contacts with family and friends at least twice weekly was associated with significantly less frequent reports of depression (OR = 0.67, 95% CI = 0.46–0.98) in the subsample of individuals providing these data. We included social contacts in this final model to test for a potential mediating effect on the association between church attendance and depression. We observed within-strata ORs for the effect of each level of church attendance to be of comparable magnitude with the previous model (without control for social contacts). Although the magnitude of effects was unchanged, significance levels were no longer statistically significant; this suggests the possibility of a slight mediating effect.
Religious Compliance
The first two models for religious compliance (testing effects of gender, demographic, and health variables) were equivalent to results found for church attendance (Table 3). The third model tested for the LDS affiliation and religious compliance, again with double risk for LDS respondents and compliance associated with significantly fewer reports of depression (OR = 0.53, 95% CI = 0.38–0.73).
|
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
; Braam et al., 2001; Kennedy, Kelman, Thomas, & Chen, 1996; Meertens, Scheepers, & Tax, 2003; Milstein et al., 2003), and among middle-aged LDS women (Johnson, 2004). We predicted that the effect would be stronger among LDS members, but perhaps because of the relatively small numbers of non-LDS members, we did not find this to be true. In fact, LDS church members reported slightly more depression than non-LDS individuals, and after we adjusted for age, marital status, and health status, this difference became statistically significant at roughly twice as much depression among LDS members as among non-LDS members. Otherwise, LDS and non-LDS members had comparable age and functional and health status.
We found that depression was 1.6 times as frequent in women as in men, a common finding across many studies (Angst et al., 2002; Gallo & Lebowitz, 1999). Women also attended church significantly more often than men (p =.005). The most notable finding, however, was that gender significantly moderated the effect of religious involvement on depression; in fact, effects were in the opposite direction for men and women.
For women, periodic (1–2 times/month), weekly, or even more church attendance was associated with reduced risk for depression, compared with never attending church. This was true even after we made covariate adjustments for demographic and health variables. Many regular church goers enjoy strong interpersonal attachments and companionship with other church members (Duke, 1998) and derive moral purpose and life meaning from their involvement (Bergin, Payne, Jenkins, & Cornwall, 1994), correlates of subjective well-being in late life (Krause, 2003). This may be particularly true for women, given the higher salience that religious involvement holds for women (Stark, 2002). In contrast, women attending church infrequently (less than monthly) may feel "disconnected" from the church community (more so than homebound individuals who receive home delivery of church services) or may be experiencing religious doubt, shown to be correlated with greater psychological distress (Krause, Ingersoll-Dayton, Ellison, & Wulff, 1999).
Religious involvement tended to have the reverse effect for men in this population. Before adjustment for any covariates, depression rates were 3% for men at all levels of church attendance, except for 6% at 1 to 2 times per month. After covariate adjustment, all respondents but those at one attendance level had significantly higher depression rates than those who never attend. This is in contrast to studies of hospitalized men (Koenig et al., 1992) and incarcerated men (Koenig, 1995), which showed that higher religious involvement was associated with less depression, but these are special samples that may not generalize to a larger population. We found no published studies that tested interaction effects between gender and religious involvement on depression.
The higher depression risk among men who attend church may be partially attributable to an organizational power differential by sex. Men tend to hold more organizational power (Beit-Hallahmi, in press; Cnaan & Helzer, 2004), particularly so in more patriarchal religious organizations such as the LDS church where the highest levels of leadership, and hence social recognition, are occupied by (primarily middle-aged) men. Social role loss from retirement can be moderated by formal volunteer service, enhancing psychological well-being (Greenfield, 2004). However, LDS church callings are largely diminished for men by age 70 (except at the highest churchwide organizational level), leaving many older men who attend church to experience a sense of having been displaced.
Another factor that may account for higher depression rates among older men who attend church in this community is the extent to which church attendance serves as a social outlet. Research has shown that men more often rely on their wives for social support whereas women have larger and more varied social networks (Antonucci, 1990). An example of this is that women tend to socialize and confide in friends at church regarding personal issues, which is especially true for LDS women (S. Johnson, personal communication, May 18, 2005). Conversely, for men, much of the experience of church attendance centers on the administrative aspects of implementing church programs and providing instrumental service to church members. Thus, the absence of church attendance functioning as a social and emotional outlet, along with the perception of displacement, may combine to increase depression risk in men.
Some (Husaini et al., 1999) but not all (Braam, Beekman, Deeg, & van Tilburg, 1997) researchers have found that social networks mediate the association between church attendance and depression. However, it has also been shown (among middle-aged LDS women) that the personal experience of religion, rather than mere social religious participation, reduces depression risk (Johnson, 2004). In the present study, the test for a potential mediating effect of social contacts on the association between church attendance and depression was suggestive of a slight mediating effect. However, although significance levels dropped from statistically significant to nonsignificant after we controlled for social contacts, the magnitude of effect of church attendance remained unchanged. Further, because the individuals in the subsample providing social contact data were significantly healthier, younger, more educated, less depressed, and more likely to be married, this test was somewhat limited.
To explore the broad construct of religious compliance and its effect on depression, we defined full religious compliance as weekly church attendance and abstinence from alcohol and tobacco. As the latter behaviors are proscribed in LDS doctrine, we found that in this population being a current smoker (n = 111 or 2.5%) or current drinker of alcohol (n = 179 or 4%) were infrequent behaviors. Results of models for religious compliance were comparable with those of church attendance alone, most likely a reflection of the uniformity in abstinence from these substances and their high correlation with church attendance (p <.001 for both).
Limitations
A limitation of this study is that because depression was operationalized as DSM-IV major depression, we have not tested for the impact of religious involvement on minor or subsyndromal depression, which can also have a significant impact on morbidity. Another limitation is our use of a skip-out interview strategy with the DIS, and though common in epidemiological research, it has the potential for underdiagnosis. As with other population studies, we relied on self-reports of depressive symptoms instead of a clinical evaluation; however, even physician interview is vulnerable to variable recall and willingness to report depressive symptoms, thus risking underreporting of symptoms in some groups (Steffens et al., 2000). Religious individuals may be reluctant to reveal emotional problems, perceiving them to demonstrate weaker faith, but they may also feel more compelled to be truthful, even concerning depression.
Our social network measure was frequency of social contacts and did not inquire as to received support or adequacy of support. Because social networks tend to be only weakly correlated with health outcomes, our test for a mediating effect would have been enhanced if such measures had been available. Additionally, the fact that those without social network data reported more depression represents a limitation in the test for its mediating effect. Further, because there were small numbers in some cells (resulting in wider CIs), the reader should interpret ORs with caution.
Our measurement of religious involvement primarily by use of church attendance was somewhat limiting. Private religiosity, including behavioral, cognitive, and emotional components, may play a greater role in generating coping skills that result in significant protection from depression (Boey, 2003; Husaini et al., 1999; Strawbridge et al., 1998), though the contrary has also been demonstrated (Husaini et al.; Parker et al., 2003). A more in-depth study of various spiritual components of religious practice among members of the predominant church in a religiously homogeneous community would be desirable, including measures of intrinsic and extrinsic religiosity (Allport & Ross, 1967; Koenig, 1995), religious meaning and religious doubt (Krause, 2003; Krause et al., 1999), and religious coping (Pargament, Koenig, Tarakeshwar, & Hahn, 2004).
These findings may not generalize to all populations of older adults but may be applicable to other populations with a predominant religious group, such as the Amish in Lancaster County or the Southern Baptists in the South. Further study of the LDS population and of other religious populations with an organizational power differential by sex is needed. Generalizability of LDS individuals is bolstered by national samples showing that LDS members typically have average socioeconomic status and similar overall happiness, marital happiness, and self-esteem (Heaton, Goodman, & Holman, 1994), and on average "flat" (normal) personality profiles on the Minnesota Multiphasic Personality Inventory (Judd, 1998).
Conclusions
To our knowledge, the present study is the first to address the association between religious involvement and depression among older LDS members, one example of a community with a predominant religious group and a male-dominated clergy. Noted strengths are that it is population based with a 90% participation rate at in-person interview, thus conferring significant protection against response bias (Norton, Breitner, Welsh, & Wyse, 1994). Further, our models controlled for potential confounding factors of physical health, functional abilities, marital status, and education. However, causal relationships cannot be proven from these analyses. As these are cross-sectional data, it may be equally likely that reduced church attendance is a consequence of social withdrawal caused by depression brought on by other factors. However, these cross-sectional data add to the growing body of literature implicating the salience of religion in late-life mental health. The opposite effects we observed between men and women warrant further study to examine potential gender differences in the emphasis on public versus private forms of religious practice and their association with mental health. Longitudinal analyses would help to elucidate directionality, which may even be recursive. To explore this phenomenon in greater detail, in future studies of this population we will include an expanded religiosity measurement with longitudinal depression data.
|
Acknowledgments |
|---|
We thank Truls Østbye and Ronald Munger for their in-depth reviews, and Marie Cornwall and Sherrie Johnson for their helpful input on the manuscript. We also thank the study participants and their families for their generosity in willingness to participate. Other Cache County Study investigators involved in the project include (in alphabetic order) James Anthony, Erin Bigler, James Burke, Michelle Carlson, Robert Green, Kate Hayden, Liz Klein, Carol Leslie, Constantine Lyketsos, Richard A. Miech, Brenda Plassman, Roxane Pfister, Carl Pieper, Martin Steinberg, Leslie Toone, Jeannette J. Townsend, Michael Williams, and Bonita W. Wyse.
|
Footnotes |
|---|
Received for publication February 28, 2005. Accepted for publication September 30, 2005.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
lsson, S., Skoog, I. (1997). The epidemiology of affective disorders in the elderly: A review. International Clinical Psychopharmacology, 12,(Suppl. 7), S3-S13.This article has been cited by other articles:
 |
|
 |
|
  M. C. Norton, A. Singh, I. Skoog, C. Corcoran, J. T. Tschanz, P. P. Zandi, J. C. S. Breitner, K. A. Welsh-Bohmer, D. C. Steffens, and for the Cache County Investigators Church Attendance and New Episodes of Major Depression in a Community Study of Older Adults: The Cache County Study J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2008; 63(3): P129 - P137. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
