Use of linked administrative and laboratory data to confirm that serum 25(OH)D levels in pregnant women can be predicted from satellite estimates of ultraviolet radiation

Abstract

Background: Serum 25 hydroxyvitamin D [25(OH)D] levels of pregnant women have been linked to various health outcomes in their offspring. Satellite-derived ultraviolet radiation (UVR) data have been used as a proxy for 25(OH)D levels, as individual-level cohort studies are time-consuming, costly and only feasible for common outcomes. Methods: Data on 25(OH)D levels from a public laboratory database were linked to data from the Western Australian Midwives' Notification System and daily erythemal UVR dose from NASA satellites. Regression analysis was used to identify the time period prior to venesection where daily UVR dose best predicted 25(OH)D levels. A predictive model was used to validate the use of daily UVR dose as a proxy for personal sun exposure during pregnancy. Results: Data from 19 173 pregnancies in women aged 18-43 years in Western Australia were included. The daily UVR dose averaged over the 90 days before venesection was the strongest UVR predictor of 25(OH)D level (a 5% increase per 1000 J m-2; equal to 3.3 nmol L-1 at the median of 66 nmol L-1). Ethnicity was the strongest predictor of 25(OH)D levels (21% lower in non-Caucasian vs Caucasian: equal to 7.2 nmol L-1 difference). Other significant predictors were gestation, age, year, parity, socio-economic status, remoteness, medical conditions and season. Conclusion: NASA-derived erythemal UVR dose in the 90 days prior to venesection is a significant predictor of 25(OH)D levels in pregnant women. Linked administrative data can be used to investigate associations between UVR during pregnancy and health outcomes in offspring.