Factors influencing cortisol level and slope among community dwelling older adults in Minnesota

Cortisol has frequently been used as a stress marker and has been shown to be elevated in response to laboratory stressors, severe real-life stressors, and daily hassles. Furthermore, variation in cortisol rhythms has been observed in some disease states and may be related to health outcomes. The majority of cortisol and stress research has been conducted on young adults. This study examines factors associated with cortisol level and slope in healthy older adults. Forty-eight older adults from the Twin Cities, MN (age = 76.4 +/- 5.8) were interviewed regarding health, stress, affect, and social networks. Participants collected saliva every two hours over a three-day period while keeping a record of their emotions and activities. Cortisol was assayed by radioimmunoassay. Data were analyzed using a random mixed effects model and linear regression. In univariate models cortisol was associated with age (p < 0.0001), time of day (p < 0.001), stress level (p = 0.01), positive affect measured in interview (p = 0.005), positive mood state (p < 0.0001), negative mood state (p = 0.09), and morningness (p = 0.0006). In multivariate models, affect was no longer significantly associated with cortisol. Age (p < 0.001), morningness (p = 0.014), physical activity (p = 0.017), and hours slept (p < 0.001) predicted cortisol slope. These results suggest that while cortisol reactivity to current mood and daily stressors may be diminished in older adults, perception of average stress on a trait level is predictive of cortisol level. As such, salivary cortisol might not be a good marker of acute stress in older adults, while still valuable to measure stress over a longer period of time.