Evidence for disruption of diurnal salivary cortisol rhythm in childhood obesity: relationships with anthropometry, puberty and physical activity

Table 4 Compositional behavior model for the associations between HPA axis measures and the proportion of the day spent in screen time, non-screen sedentary behaviours, MVPA, and sleep duration

Cortisol variables	Model P	Model R²	γ_Screentime	P	γ_{non-screen/SB}	P	γ_MVPA	P	γ_sleep	P
Cortisol_08:00	0.002	0.350	−0.004	0.895	−0.030	0.686	−0.107	0.018	0.142	0.183
Cortisol_16:00	0.286	0.136	0.016	0.618	0.147	0.045	−0.009	0.377	−0.153	0.129
Cortisol_23:00	0.004	0.332	0.004	0.941	0.224	0.080	0.086	0.246	−0.315	0.080
AUC_{08:00–16:00}	0.005	0.324	0.002	0.936	0.030	0.653	−0.081	0.038	0.048	0.603
AUC_{16:00–23:00}	0.035	0.252	0.024	0.385	0.169	0.009	0.019	0.594	−0.212	0.018
AUC_{08:00–23:00}	0.009	0.309	0.010	0.711	0.064	0.292	−0.059	0.091	−0.015	0.860
\|DCS\|	0.001	0.382	−0.003	0.936	−0.069	0.445	−0.150	0.007	0.222	0.085

All models were adjusted for pubertal stage and BMI z-score using multiple linear regression. Cortisol variables were logarithmically transformed. Regression coefficients correspond to change in the log-ratio of the given behaviour relative to other behaviours. MVPA moderate to vigorous physical activity, SB sedentary behaviours, |DCS| absolute values of diurnal cortisol slope

ISSN: 1471-2431