Abstract
STUDY OBJECTIVES: This study aims first, to evaluate the longitudinal sex-specific association between iron status and objective sleep outcomes; second, to explore the cross-sectional relationship between iron status and blood leukocyte DNA methylation of key circadian genes during adolescence. METHODS: Iron status for 371 adolescents from the Mexico City-based ELEMENT cohort was estimated using serum ferritin levels. Sleep duration, sleep midpoint, sleep fragmentation, and movement index during both weekdays and weekends were measured objectively using 7-day wrist actigraphy collected at two timepoints. The Infinium MethylationEPIC BeadChip was used to measure DNA methylation of circadian genes, and data was extracted from 525 CpG sites within 12 key circadian genes. Sex-stratified linear mixed-effects regression with random intercepts were used to assess the relationship between ferritin tertiles and sleep outcomes. The association between ferritin tertiles and DNA methylation was evaluated by linear regression. The Cauchy combination test was used to combine results from individual CpG sites for each gene. A false discovery rate was used to account for multiple comparisons. RESULTS: Females with ferritin levels in the first tertile had 1.35% (95% CI: 0.21, 2.50) higher sleep fragmentation during weekdays compared to the third tertile; males with ferritin levels in the first tertile had later sleep timing during the weekends (β: 0.59 decimal hour, 95% CI: 0.05, 1.13). Among females, lower ferritin levels were associated with DNA methylation differences in CLOCK, PER1, and RORC (FDR = 0.0392). CONCLUSIONS: Differences in ferritin levels may be linked to lower sleep quality and epigenetic modifications in some circadian genes, especially in females.