The differential influence of distinct clusters of psychiatric symptoms, as assessed by the general health questionnaire, on cause of death in older persons living in a rural community of Japan

OBJECTIVES: To examine the influence of distinct clusters of psychiatric symptoms on subsequent cause-specific mortality in older persons living in the community. DESIGN: A prospective cohort study over 7.5 years. SETTING: Otsuki-town, Kochi prefecture, Japan. PARTICIPANTS: Nine hundred eighty community-dwelling persons, aged 65 to 84 in 1991. MEASUREMENTS: Psychiatric symptoms at baseline were measured using the 30-item version of the General Health Questionnaire (GHQ-30). A factor analysis was performed on the responses of all 980 subjects. The relationships between subscale scores on the identified factors and causes of deaths occurring over 7.5 years in 817 respondents with no physical disability at baseline were assessed using a proportional hazards model adjusted for age, gender, chronic conditions under treatment, regular physical activity, and availability of close or casual neighbors. RESULTS: The factor analysis identified three clusters of psychiatric symptoms: depression, apathy/anergia, and anxiety. In the proportional hazards model, which included three GHQ subscales depicting these factors simultaneously, the depression subscale was associated with increased mortality from cerebrovascular disease (multivariate adjusted hazard ratio per unit increase in the standard score on the depression subscale = 2.04, 95% confidence interval (CI) = 1.17-3.55), and the apathy/anergia subscale was associated with increased mortality from noncancer and noncardiovascular causes (multivariate adjusted hazard ratio per unit increase in the standard score on the apathy/anergia subscale = 1.71, 95% CI = 1.25-2.34). The anxiety subscale was not associated with any cause of death. CONCLUSION: Depressive symptoms and symptoms indicating apathy/anergia have differential influences on subsequent causes of death in older persons living in the community. Identification of specific psychiatric symptom clusters may contribute to the prevention of deaths from specific causes in older populations.