Modeling day-to-day changes in diurnal cortisol rhythms using 3-level lagged multilevel models reveals reciprocal relations among sleep, cortisol, affect and wellbeing

Utilizing a 3-day diary methodology including objective (actigraph) measures of sleep and repeated sampling of salivary cortisol across the day, we examine how between-person differences in sleep hours and timing relate to individual differences diurnal cortisol rhythms, and also how within-person day-to-day variations in hours of sleep and sleep timing relate to day-to-day variations in diurnal cortisol. Older adolescents (N = 119; 77% female) were wore an actigraph for 3 days of sampling and provided 6 salivary samples each day (wakeup, wakeup + 30 min, 2, 8, and 12 hours post-awakening and bedtime). The mean age of the sample was 17.93 years and individuals were from a variety of ethnic/racial backgrounds. We utilized a 3-level (cortisol samples nested within days, nested within individuals) multilevel growth-curve analysis, with same-day and prior-day lagged predictors at day level. Between-person analyses revealed that across individuals, greater hours of sleep related to significantly steeper declines in cortisol across the day. A day-to-day analysis revealed significantly higher waking cortisol levels and steeper cortisol declines after nights adolescents slept more. Further, on days adolescents had higher wakeup cortisol values and a steeper diurnal decline in their cortisol, they experienced greater hours of sleep that night and tended to wake up earlier the following day. Together, our findings suggest a complex, reciprocal relation between sleep and diurnal cortisol. Additional brief examples of other day-lagged 3-level multilevel models, showing reciprocal relations among social and affective experience, diurnal cortisol rhythms and perceived physical wellbeing may also be provided.