Race and stress in the incidence of herpes zoster in older adults.

OBJECTIVES

To examine the effect of black race and acute (negative life events) and chronic (lack of social support) psychological stress on the risk of herpes zoster in late life.

DESIGN

A population-based, prospective cohort study.

SETTING

Central North Carolina.

PARTICIPANTS

Duke Established Populations for Epidemiological Studies of the Elderly, a stratified probability sample of community-dwelling persons more than 65 years of age.

MEASUREMENTS: Interviewers administered a comprehensive health survey to the participants in 1986-1987 (P1, n = 4162), 1989-1990 (P2, n = 3336), and 1992-1994 (P3, n = 2568). Incident cases of zoster between P1 and P2 and P2 and P3 served as the dependent variable. Hypothesis-testing variables included race, negative life events, and five measures of social support. Control variables included age, sex, education, cancer, chronic diseases, basic ADLs, instrumental ADLs, depression, self-rated health, hospitalization, and cigarette smoking. Statistical analyses employed chi-square tests and proportional hazards model.

RESULTS

At baseline, the sample had a mean age of 73.6 years and was 55% black, 45% white, and 65% female. There were 65 cases of zoster between P1 and P2 and 102 cases of zoster between P2 and P3. From P1 to P2, 1.4% of blacks and 3.4% of whites developed zoster (P < .001). From P2 to P3, 2.9% of blacks and 7.5% of whites developed zoster (P < .001). After controlling for the above variables, blacks were significantly less likely to develop zoster (adjusted risk ratio = 0.35; 95% confidence interval (CI), 0.24-0.51; P < .001). Negative life events increased the risk of zoster, but the result was borderline for statistical significance (adjusted RR = 1.38, 95% CI 0.96-1.97; P = .078). No measures of social support were significantly associated with zoster.

CONCLUSION

Black race decreased the risk of zoster in late life significantly. Measures of stress were not significantly related to zoster, but study limitations preclude definitive conclusions. Future research should focus on these factors in larger samples and different populations.