Abstract

This study examined the associations between intra-individual variability in, and inter-individual levels of, diurnal cortisol secretion with a marker of low-grade inflammation (i.e., C-Reactive Protein; CRP). Reasoning that greater day-to-day cortisol variability could reflect a dysregulation of the HPA axis, we hypothesized that it would predict higher levels of CRP, above and beyond inter-individual differences in cortisol levels. A 10-year longitudinal study of 130 older adults examined diurnal cortisol secretion on three different days across each of the 6 waves of data collection and levels of CRP during the last 3 waves. Indicators of mean cortisol levels, short-term cortisol variability, and long-term cortisol variability were analyzed. Hierarchical linear modeling showed significant main effects, linking baseline mean cortisol levels, T-ratio = 2.25, p = 0.03, and long-term cortisol variability, T-ratio = 2.63, p = 0.01, with higher CRP values six to ten years after study entry. In addition, a two-way interaction demonstrated that short-term variability in cortisol were associated with higher levels of CRP among individuals who secreted relatively high, T-ratio = 2.68, p = 0.01, but not low, T-ratio = −1.09, p = 0.28, baseline levels of cortisol. Finally, a three-way interaction, T-ratio = 2.24, p = 0.03, suggested that the effect of long-term cortisol variability on CRP became stronger over time among participants who secreted high average levels of cortisol, whereas it became weaker among their counterparts who secreted low average levels of cortisol.