This study examined relationships between chronic stress and insulin/glucose in two groups of nondiabetics, M age = 69.4: spouse caregivers (CGs) of persons with Alzheimer's disease (n = 73) and age- and gender-matched spouses of nondemented controls (COs) (n = 69). Fasting insulin/glucose and psychological variables were assessed twice (Time 1, Time 2) over a 15-18 month period. CGs had significantly higher insulin levels at Times 1 and 2 than did COs even when obesity, exercise, gender, age, alcoholic drinks, hormone replacement therapy (HRT), lipids, and hypertension (HTN) were considered in the analyses. CGs generally reported significantly more psychological distress (higher burden, depression, hassles, and lower uplifts) than did COs at each time. Differences in psychological distress at Time 1 between CGs and COs did not mediate the insulin difference in the groups at Time 1, but differences in distress at Time 2 between CGs and COs did mediate their difference in insulin at Time 2. Although caregiver status was not associated with glucose at Time 1 or Time 2, psychological distress was positively associated with glucose at Time 2. Moreover, psychological distress at Time 1 was associated with higher glucose at Time 2 after controlling for glucose at Time 1. These data suggest that relationships between psychological and physiological distress exist both cross-sectionally and over time. These results may be important because higher insulin and glucose levels are associated with increased coronary risk and coronary heart disease.