Associations between organophosphate esters metabolites and sleep disorder and trouble sleeping in adults: a machine-learning approach

Kang, Xiao; Li, Jingxian; Luo, Jia; Zhang, Dongfeng

Associations between organophosphate esters metabolites and sleep disorder and trouble sleeping in adults: a machine-learning approach

2022

Kang, Xiao | Li, Jingxian | Luo, Jia | Zhang, Dongfeng

Organophosphate esters (OPEs) are used widely as flame retardants and plasticizers. However, the associations between OPEs metabolites and sleep outcomes (sleep disorder and trouble sleeping) remain unknown. Data utilized in this cross-sectional study was from the National Health and Nutrition Examination Survey 2013–2014, including 1393 adults aged ≥ 20 years. We conducted weighted logistic regression and Bayesian kernel machine regression (BKMR) models to analyze the associations between OPEs metabolites and sleep outcomes. We included data from 2011 to 2012 cycle in our sensitivity analysis to explore the association further. Logistic regression model presented a significant positive association between diphenyl phosphate (DPHP) and sleep disorder in all participants (odds ratio (95% confidence interval) for the second quartile was 2.46 (1.85, 3.28)). We observed positive associations between OPEs metabolites and sleep outcomes in males. Among females, no significant association was observed in the logistic model. BKMR presented that dibutyl phosphate (DBUP) was the relatively important exposure. There was a positive association between OPEs metabolites mixture and trouble sleeping. Univariable exposure–response functions demonstrated U-shaped associations between DBUP and sleep outcomes, while bis(2-chloroethyl) phosphate (BCEP) was associated with sleep disorder negatively in females. No substantial changes appeared in the results after including the data from 2011 to 2012 cycle. This current study indicated that OPEs metabolites might be associated with sleep disorder and trouble sleeping, and the associations seemed to be sex-dependent.

Afficher plus [+]