OBJECTIVE: We investigated the association between reaction time (RT) and mortality in middle-aged and older atomic bomb survivors and their unexposed controls over a period of 30 years. METHODS: During 1970-72, 4912 participants of the Adult Health Study cohort in Hiroshima, Japan, underwent biologic tests including RT. Mortality was followed to the end of 2003. RESULTS: In a multivariate-adjusted model, the hazard ratio (HR) for 1-standard deviation increments of RT was 1.08 (95% confidence interval [CI]=1.03-1.13) for men, 1.22 (95% CI=1.16-1.28) for women, and 1.13 (95% CI=1.09-1.16) for all. When the analysis was performed by sex, age, and follow-up period, a consistent increase of mortality with increments of RT was observed. The HR for mortality for the highest RT quintile was higher than that of the lowest quintile in all sex-age groups. A significant positive association between mortality risk and RT was observed even after 20 years of follow-up (p=.03 in men, p<.001 in women). RT and radiation dose were risk factors for mortality independent of conventional risk factors such as smoking, high blood pressure, and diabetes mellitus. Interaction between RT and radiation dose had no significant effect on mortality in men. Although increased radiation dose reduced the HR for mortality per RT increment in women, RT and radiation dose were still significant predictors of mortality. CONCLUSIONS: RT is a consistently strong predictor of mortality. Although mortality risk increased with radiation dose, radiation did not accelerate the relationship between RT and mortality.
OBJECTIVE: We investigated the association between reaction time (RT) and mortality in middle-aged and older atomic bomb survivors and their unexposed controls over a period of 30 years. METHODS: During 1970-72, 4912 participants of the Adult Health Study cohort in Hiroshima, Japan, underwent biologic tests including RT. Mortality was followed to the end of 2003. RESULTS: In a multivariate-adjusted model, the hazard ratio (HR) for 1-standard deviation increments of RT was 1.08 (95% confidence interval [CI]=1.03-1.13) for men, 1.22 (95% CI=1.16-1.28) for women, and 1.13 (95% CI=1.09-1.16) for all. When the analysis was performed by sex, age, and follow-up period, a consistent increase of mortality with increments of RT was observed. The HR for mortality for the highest RT quintile was higher than that of the lowest quintile in all sex-age groups. A significant positive association between mortality risk and RT was observed even after 20 years of follow-up (p=.03 in men, p<.001 in women). RT and radiation dose were risk factors for mortality independent of conventional risk factors such as smoking, high blood pressure, and diabetes mellitus. Interaction between RT and radiation dose had no significant effect on mortality in men. Although increased radiation dose reduced the HR for mortality per RT increment in women, RT and radiation dose were still significant predictors of mortality. CONCLUSIONS: RT is a consistently strong predictor of mortality. Although mortality risk increased with radiation dose, radiation did not accelerate the relationship between RT and mortality.