0011 Later and Irregular Light Exposure in Early Pregnancy Predicts Small-for-Gestational-Age Birth

Abstract Introduction Light is the primary circadian entrainment cue, and insufficient daytime and excessive nighttime light contribute to chronodisruption, which is associated with adverse birth outcomes. Nightshifts during pregnancy are associated with higher rates of low birthweight. However, it is unknown whether naturalistic light exposure patterns during pregnancy are associated with real-world chronodisruption and adverse birth outcomes. We hypothesized that personal light exposure in early pregnancy correlates with lower diurnal melatonin amplitude and predicts higher odds of small-for-gestational-age birth. Methods Participants were 561 women with singleton pregnancies (agemean=29.27±5.20). We assessed light exposure (photopic lux) every minute up to 20-weeks gestation via wrist-worn actigraphy (MotionWatch 8, CamNtech Ltd.). We computed the following light exposure variables: first and last timepoint above 50 lux (FLiT; LLiT) and their standard deviation (SD), and mean timepoint and total time above 100, 500, 1000 lux (MLiT; TAT). We assessed diurnal melatonin via saliva samples provided at 4-hour intervals over 24 hours under real-world conditions and calculated amplitude using Cosinor analysis. We tested associations between light exposure and small-for-gestational-age birth with logistic regression models covarying for solar photoperiod (Model 1), age, parity, nightshift work, pregestational hypertension and diabetes, and tobacco, alcohol, or drug use (Model 2), and social disadvantage (Model 3) and associations between light exposure and diurnal melatonin amplitude with linear regression covarying for solar photoperiod. Results Later MLiT 100, MLiT 500, and FLiT 50 lux, greater FLiT and LLiT 50 lux SD, and lower TAT 100 lux were significantly associated with higher odds of small-for-gestational-age birth in Model 1 (p’s<.05). Later mLIT 100 and fLIT 50 lux and higher fLIT 50 lux SD remained significant predictors (p’s<.05) in Model 2. Light exposure did not significantly predict small-for-gestational-age birth in Model 3 (p’s>.05). Later mLIT 100 and fLIT 50 lux, greater fLIT 50 lux SD, and lower TAT 100 and 500 lux were significantly associated with lower diurnal melatonin amplitude (p’s<.05). Conclusion Later and irregular light exposure was associated with real-world chronodisruption and small-for-gestational-age birth. Personal light exposure is a potential modifiable target to improve circadian and obstetric health. Support (if any) March of Dimes, NIH R01 HD037831, NIH K23 MH137376, NIH K23 HD095075-03