Physical activity (PA) has been shown to influence salivary cortisol concentrations in small studies conducted among athletes. We assessed the association of activity status and patterns with salivary cortisol in the general population.
Cross-sectional study including 1948 adults (54.9% women, 45-86 years). PA and sedentary behaviour (SB) were measured for 14 days by accelerometry. Low PA and high SB status were defined, respectively, as the lowest and highest tertile of each behaviour. 'Inactive', 'Weekend warrior', and 'Regularly active' patterns were also defined. Four salivary cortisol samples were collected over a single day and the following parameters were calculated: area under the curve to ground (AUCg), awakening response (CAR) and diurnal slope.
After multivariable adjustment, low SB remained associated to steeper slopes relative to high SB (- 1.54 ± 0.03 vs. - 1.44 ± 0.04 nmol/l per hour). Non-significant trends were found for high PA relative to low PA with steeper slopes (- 1.54 ± 0.03 vs. - 1.45 ± 0.04) and lower AUCg (208.7 ± 2.0 vs. 215.9 ± 2.9 nmol.h/l). Relative to 'Inactives', 'Regularly actives' had lower AUCg (205.4 ± 2.4 vs. 215.5 ± 2.9) and 'Weekend warriors' had steeper slopes (- 1.61 ± 0.05 vs. - 1.44 ± 0.04). No associations were found for CAR.
Low SB and high PA are related to lower cortisol secretion as measured by different parameters of salivary cortisol, but the effects were only modest.