Duration, timing, and consistency of sleep in relation to inflammatory cytokines in Mexican adolescents

Abstract

Objective: To evaluate whether sleep duration, timing, and variability were associated with inflammatory cytokines in a cohort of Mexico City adolescents.

Methods: The analytic sample comprised >500 adolescents who were part of an ongoing longitudinal study in Mexico City. At two time points during mid-to-late puberty (average age 14, n = 391) and late-to-post puberty (average age 16, n = 345), adolescents completed a follow-up visit that included 7-day wrist actigraphy and clinical assessment of plasma inflammatory cytokines (high-sensitivity C-reactive protein, Interleukin 1β, Interleukin 6, and Tumor Necrosis Factor ɑ). Sleep characteristics included weekday and weekend sleep duration and midpoint (median of bed and wake time), as well as sleep variability (SD of sleep duration across 7 days) and social jetlag (midpoint difference from weekdays to weekends). At each time point, multivariable linear regression models were run with log inflammatory levels as the outcome and categories of sleep characteristics as predictors, while adjusting for potential confounders (specific to each model). Analyses were run unstratified and sex-stratified.

Results: In the mid-to-late pubertal visit, weekday sleep duration was inversely associated with natural log hs-CRP after adjustment (Q4 vs Q1: β = -0.41, 95% Confidence Interval (CI) -0.81 to -0.01) and later sleep midpoint was positively associated with log hs-CRP (Q4 vs Q1: β = 0.55, 95% CI 0.13 to 0.97). Sleep duration variability was associated with higher IL-1β among boys, while in girls social jetlag was associated with higher IL-1β and weekend sleep duration was inversely associated with IL-6. At the late-to-post pubertal visit, there were few associations except for a positive association between weekday sleep duration and hs-CRP among boys (β = 0.60, 95% CI 0.04 to 1.16) and a non-linear positive association between social jetlag and hs-CRP among girls (β = 0.80, 95% CI 0.22 to 1.37 comparing 2 to 3 h of social jetlag vs <1 h).

Conclusion: Later timing, shorter duration, and inconsistency of sleep were related to higher levels of inflammatory biomarkers, but associations were more evident at the mid-to-late pubertal visit than the late-to-post pubertal visit.