Table 4:

Updates to assessment of probable associations between exposures and metabolic outcomes

	Outcome	Strength of human evidence (2015)	Probability of causation (2015), %	Updates to literature (since 2015)
Prenatal DDE	Childhood obesity	Moderate	40–69%	Not reassessed
Prenatal PFAS	Childhood obesity	Not assessed	Not assessed	Multiple cohorts report positive findings consistent with Barker hypothesis14 and possible mechanism of impaired glucose tolerance; less consistent associations than with birthweight
Prenatal BPA	Childhood obesity	Very low to low	20–69%	Increases in body fat measures (more consistent results than BMI); highly variable approaches to exposure assessment complicate interpretation; pattern of sexual dimorphism not consistent
Prenatal and peripubertal phthalates	Childhood obesity	Not assessed	Not assessed	Pattern of association across studies with increases in BMI and fat mass measures; one longitudinal study showed associations with peripubertal exposure
Pregnancy PFAS	Impaired glucose tolerance	Not assessed	Not assessed	Multiple studies with consistent associations; others with gestational diabetes
Prenatal phthalates	Impaired glucose tolerance	Not assessed	Not assessed	Multiple studies with consistent associations; others with gestational diabetes
Adult DEHP	Adult obesity	Low	40–69%	Positive findings strengthen existing evidence
Adult PFAS	Adult obesity	Not assessed	Not assessed	No significant association at highest levels of (modelled) exposure; associations with lower levels of exposure in multiple cohorts with mechanistic insight and effect modification by diet
Adult DEHP	Adult diabetes	Low	40–69%	One study in adults modestly supports existing evidence of association
Prenatal DDE	Adult diabetes	Low	20–39%	Not reassessed
Pregnancy PFAS	Adult diabetes	Not assessed	Not assessed	Two longitudinal studies of low exposures show associations with indices of insulin resistance; inverse association in higher range of exposure noted in one study
Adult BPA and BPS	Adult diabetes	Not assessed	Not assessed	Case-control, small-scale intervention, and longitudinal studies all consistent with associations found in laboratory studies

Adapted from the data first reported in Trasande et al (2015)¹¹ and updated in Trasande et al (2016).¹² See appendix for full list of studies mentioned here that have updated the literature (appendix pp 24–29). DDE=dichlorodiphenyltrichloroethane. PFAS=perfluoroalkyl substances. BPA=bisphenol A. DEHP=di-2-ethylhexyl phthalate. BPS=bisphenol S.