Serum cotinine and wrist-worn ambient light exposure patterns in U.S. adults: A cross-sectional analysis of NHANES 2011-2014.

OBJECTIVES

To examine whether serum cotinine, a biomarker of tobacco smoke exposure, is associated with objectively measured wrist-worn ambient light in U.S. adults, and whether any association is accounted for by objective sleep and physical activity.

METHODS

We analyzed NHANES 2011-2014 data from 6937 adults aged ≥20 years with valid serum cotinine and wrist-worn light data. Seven light outcomes were modeled with survey-weighted linear regression using log₂(cotinine+1) across four adjustment levels. Exploratory models added actigraphy sleep timing and duration and wrist-derived activity. Benjamini-Hochberg false discovery rate (FDR) correction was applied across the 7 outcomes within each model.

RESULTS

In the primary model (N = 6350), higher cotinine was associated with higher nighttime light (β = +0.024, p_FDR = 0.014) and lower evening light (β = -0.031, p_FDR = 0.042). Adding objective sleep and activity attenuated the nighttime association to the null (β = +0.002, P = .70), whereas evening light persisted (β = -0.023, P = .011). A cotinine-by-sex interaction for nighttime light was identified (p_interaction = 0.010), significant in females but not males.

CONCLUSIONS

Higher serum cotinine was associated with higher nighttime and lower evening wrist-worn light. The nighttime association was largely accounted for by objective sleep and activity. The cross-sectional design precludes causal inference. Longitudinal mediation studies are needed.