Table 2.

The effect of OC-use on CAR evaluated in generalized least square analyses in alternative models with increasing complexity.

Covariates for adjustment	Effect (nmol/L*minutes)	95% CI	p-value
A. No adjustment	−228	[−354; −103]	<0.001
B. Age, work day status, BMI	−203	[−343; −63]	0.006
C. Age, BMI, Cohen’s PSS, 5HTTLPR genotype, TMD, smoking, work day status, season, sleep quality*	−238	[−418; −57]	0.013

Model B was chosen as our main model. Work day status describes whether cortisol saliva samples were collected on a work/study or rest day. 5-HTTLPR genotype status is defined as LA/LA versus not-LA/LA. Sleep quality was assessed with global score of the Pittsburgh Sleep Quality Index (PSQI). Season, summer versus winter. Cohen’s PSS, Cohen’s Perceived stress scale; TMD, total mood disturbance. * Model C only includes 63 participants due to missing PSQI data (n=12) and missing TMD data (n=2).