Sandra India Aldana, Francheska Merced-Nieves, Vishal Midya, Xin Yu, Katherine Svensson, Jennifer W Thompson, Allison Kupsco, Nia McRae, Inessa Lopez, Alejandra Cantoral, Lauren M Petrick, Haibin Guan, Georgia Dolios, Cecilia S Alcalá, Syam S Andra, Jasmin Chovatiya, Sandipkumar Teraiya, Julio Landero, Dinesh Barupal, Andrea L Deierlein, Allan C Just, Mike Z He, Iván Gutiérrez-Avila, Guadalupe Estrada-Gutierrez, Antonia M Calafat, Andrea A Baccarelli, Robert O Wright, Martha María Téllez-Rojo, Damaskini Valvi, Elena Colicino
There is a paucity of research examining the impact of environmental chemical (EC) mixtures on sleep and the role of body mass index (BMI) in this association. We prospectively evaluated 1) the association between EC exposure (bio)marker-mixtures during peri-pregnancy and sleep outcomes after pregnancy, and 2) mediation by BMI. We studied ∼393 mothers from the PROGRESS cohort with 52 EC (bio)markers measured during pregnancy (estimated air pollutants, blood/urinary metal(loid)s, urinary high- and low-molecular-weight phthalate [HMWP, LMWP] and pesticide metabolites) and 1-month postpartum (serum perfluoroalkyl substances), and with sleep outcomes measured ∼8 years after pregnancy. Sleep duration and efficiency were calculated using actigraphy data, and self-reported sleep disturbance and impairment were estimated from validated NIH-Toolbox questionnaires. Covariate-adjusted BWQS regressions were used to assess EC (bio)marker-mixture associations (by chemical group and overall) with sleep outcomes. Sensitivity BKMR and mediation analyses by BMI were also implemented. Mothers were ∼28.2 (SD:5.5) years old at baseline and had an average sleep duration of 6.9 (SD:0.95) hours at the 8-year follow-up visit. Several EC (bio)marker-mixtures during pregnancy were associated with sleep outcomes post-pregnancy. A quartile increase in the air pollutant-mixture was associated with an ∼8-minute lower sleep duration [β=-0.14 hours (95%CrIs: -0.29, 0.00)]. Similarly, a quartile increase in the non-essential metals/metalloids-mixture was associated with lower sleep efficiency [β=-0.95% (95%CrIs: -1.81%, -0.17%)]. A negative marginal association was also found between the HMWP-mixture and sleep efficiency [β=-0.62% (95%CrIs: -1.28%, 0.02%)]. Furthermore, the pesticide-mixture was associated with higher self-reported sleep disturbance T-scores [β=1.16 (95%CrIs: 0.22, 2.07)]. No evidence of non-linear trends was found. No significant mediation by BMI was observed, although BMI was a weak potential mediator in the association between the HMWP mixture and sleep efficiency (indirect effect p=0.22). In summary, several EC exposure (bio)markers during the period of pregnancy were linked to poorer sleep later in life.