Background: It has been reported that morning cortisol levels increase with age, although there is some controversy in the literature.
Aim: The aim of this study was to examine associations of cortisol levels with advancing age in an elderly population and investigate possible interactions with metabolic and hormonal parameters.
Subjects and methods: From 372 subjects initially evaluated, we studied 251 ambulatory subjects aged 51-90 yr, median 71 yr (169 women), all permanent residents of a small town in southern Greece. Anthropometric parameters, glucose, insulin, cortisol, and biochemical parameters were recorded.
Results: Fasting cortisol levels (08:00-09:00 h) varied between 150.9- 854 nmol/l (mean 362.4 nmol/l). A significant association was found between age and cortisol levels (Spearman's rho =0.170, p=0.01). There was a positive correlation between cortisol levels and creatinine (Spearman's rho =0.144, p=0.023), homocysteine (Spearman's rho =0.283, p<0.001) and a negative correlation with body mass index (Spearman's rho =-0.128, p=0.047). Multivariate analysis showed that when creatinine was taken into account, the association of cortisol with age and with homocysteine was no longer significant. When, however, diabetic subjects were included in the analysis, the adjusted for creatinine association of cortisol with age was significant (β=0.168, p<0.05).
Conclusions: It is concluded that, in elderly ambulatory subjects, the reported associations between cortisol levels, age, and homocysteine may be affected by coexisting co-morbidities or possibly by a decline in renal function. In subsequent studies it is important that fasting glycemia is taken into account.