Search

Background: Few studies have explored the variance in individual symptoms by race in older adults., Methods: Data were analysed from the Duke site of the Established Populations for Epidemiologic Studies of the Elderly (EPESE), a community sample of persons 65 years-of-age and older, 54% of whom were African-Americans. Of the 3401 subjects with adequate data on depressive symptomatology, confirmatory factor analysis and LISREL were first used to confirm the presence of the factor structure previously reported for the CES-D. Next, bivariate analysis was performed to determine the prevalence of individual symptoms by race. Finally, LISREL analysis was performed to control for potential confounding variables., Results: When bivariate comparisons of specific symptoms by race were explored, African-Americans were more likely to report less hope about the future, poor appetite, difficulty concentrating, requiring more effort for usual activities, less talking, feeling people were unfriendly, feeling disliked by others and being more 'bothered' than usual. When LISREL analyses were applied to these data (controlling for education, income, cognitive impairment, chronic health problems and disability and other factors) racial differences in somatic complaints and life satisfaction disappeared, yet differences in interpersonal relations persisted., Conclusions: This study confirms earlier findings of minimal overall differences in symptom frequency between African-American and non-African-American community-dwelling older adults in controlled studies.

Personality assessed with the Minnesota Multiphasic Personality Inventory (MMPI) in college was used to predict exercise behavior measured at midlife in 3,630 men and 796 women enrolled in the University of North Carolina Alumni Heart Study. Logistic regression models were fitted for each of the MMPI clinical scales to test the predictive effect of personality, gender, and their interaction on adult exercise behavior. Lower depression, social introversion, and psychopathic deviance scores were associated with increased probability of exercising in midlife for both men and women. Furthermore, better psychological health (indexed by lower hypochondriases and psychasthenia) in college was generally predictive of increased exercise for men, whereas higher scores on these same factors predicted midlife exercise for women. There were two other patterns of gender interactions: (a) for men, lower scores on hysteria and schizophrenia scales were associated with increased probability of exercising at midlife, whereas these factors were unrelated to exercise for women and (b) for women, lower ego strength and higher college scores on paranoia and mania were associated with exercise behavior at midlife. These data suggest that early adulthood personality predictors of exercise behavior at midlife are both gender-neutral and gender-specific; that is, where no gender differences exist, healthier personality traits predict exercise at midlife, and when gender differences do occur, healthier college patterns of personality predict exercise behavior for men and sedentary behavior for women.