13th Annual Child Health Research Days

"Pathways to Equity: Better Health for All Children"

October 3 - 5, 2017


#4 Prenatal and early-life exposure to non-nutritive sweeteners and body composition at 3 years of age in the CHILD cohort

Alyssa J Archibald, University of Manitoba; Atul K Sharma, University of Manitoba; Russell J de Souza, McMaster University; Vernon W Dolinsky, University of Manitoba; Allan B Becker, University of Manitoba; Piushkumar J Mandhane, University of Alberta; Stuart E Turvey, University of British Columbia; Padmaja Subbarao, University of Toronto; Diana L Lefebvre, McMaster University; Malcolm R Sears, McMaster University; Meghan B Azad, University of Manitoba


One in three Canadian children is overweight, putting them at risk for many chronic diseases later in life. Non-nutritive sweeteners (NNS), intended to be a healthier alternative to sugar, may paradoxically induce obesogenic effects, perhaps even in utero. Recent findings in the Canadian Healthy Infant Longitudinal Development (CHILD) cohort show that infants born to mothers who regularly consumed NNS in pregnancy have an increased body mass index (BMI) at 1 year. Our objective was to examine the association between prenatal NNS exposure and body composition at 3 years of age.


Accessing data from 2298 mother-child dyads in the CHILD Study, we examined the association between maternal sugar- and artificially-sweetened beverage intake during pregnancy and child body composition using linear and logistic regression models. Maternal diet including NNS consumption was self-reported by food frequency questionnaire, and child anthropometric measurements were taken at 3 years (height and weight to calculate WHO BMI z-scores, and subscapular skinfold z-scores against WHO standards to assess adiposity). Covariates including maternal BMI, gestational diabetes, breastfeeding, child diet quality and sedentary behaviour were determined by questionnaire.


Children born to mothers reporting daily NNS consumption had higher BMI at 3 years of age (β=0.37, 95% CI 0.19–0.55, n=2298) compared to those with less NNS consumption. This association was attenuated after adjustment for relevant covariates (aβ=0.17, 95% CI -0.05–0.39, n=1797); however, it remained significant among children who were not breastfed for at least 6 months (aβ=0.40, 95% CI 0.02–0.77, n=390). Similar associations were found for adiposity measured by subscapular skinfolds.


Our results suggest that maternal NNS use in pregnancy may increase the risk of early childhood obesity, and that breastfeeding may mitigate this effect. This study furthers our understanding of the development of childhood obesity, and will ultimately help inform dietary recommendations for expectant mothers.