Table 17.

Evidence profile table for male reproductive effects of DEHP.

Outcome	Studies (design in parentheses)	Factors that increase confidence	Factors that decrease confidence	Summary of findings	Confidence judgement for outcome	Confidence judgement for overall hazard
Anogenital distance	Medium confidence Bornehag et al., 2015 (C) Jensen et al., 2016 (C) Swan et al., 2015 (C) Low confidence Bustamante-Montes et al., 2013 (C) Suzuki et al., 2012 (C) Swan, 2008 (C)	Among medium confidence studies: Consistency Exposure-response gradient across studies Minimal concerns for bias Overall: finding is consistent with other systematic review/meta-analysis (National Academies of Sciences, Engineering, and Medicine, 2017)	Low precision in Bornehag et al., 2015	Inverse associations between DEHP exposure and anogenital distance reported in 5/6 studies (Jensen et al., 2016, Swan et al., 2015, Bornehag et al., 2015, Swan, 2008, Suzuki et al., 2012), of which 2 were statistically significant (Swan et al., 2015, Swan, 2008). Among 3 medium confidence studies, effect size increased with increasing exposure levels (Fig. 2).	⨁⨁◯ MODERATE	⨁⨁⨁ ROBUST Supported by coherence across outcomes.
Semen parameters	Medium confidence Axelsson et al., 2015a, 2015b (CS) Bloom et al., 2015 (C) Han et al., 2014 (CS) Hauser et al., 2006 (CS) Huang et al., 2013 (CS) Jurewicz et al., 2013 (CS) Liu et al., 2012 (CS) Pan et al., 2015 (CS) Specht et al., 2014 (CS) Thurston et al., 2016 (CS) Wang et al., 2015a, 2015b, 2015c (CS) Wirth et al., 2008 (CS) Low confidence Den Hond et al., 2015 (CS) Herr et al., 2009 (CS)	Minimal concerns for bias Biological plausibility Lower study sensitivity could account for most observed inconsistency (i.e., null results in some studies)	Difficult to assess exposure-response gradient and precision across studies due to differing analyses/presentation of results	Inverse associations between DEHP exposure and sperm concentration in 7/14 studies (Herr et al., 2009, Huang et al., 2011, Wirth et al., 2008, Specht et al., 2014, Bloom et al., 2015, Hauser et al., 2006), of which 1 was statistically significant (Bloom et al., 2015). Four studies reported an inverse association with motility (Axelsson et al., 2015a; Huang et al., 2013; Huang et al., 2011, and Jurewicz et al., 2013), and 2 were statistically significant (Axelsson et al., 2015a; Jurewicz et al., 2013) (Table 4). Biological plausibility-Studies showed increased apoptosis (Huang et al., 2014); (Wang et al., 2016b; You et al., 2015), increased reactive oxygen species generation (Huang et al., 2014), and increased sperm aneuploidy (Jurewicz et al., 2013) with increased exposure.	⨁⨁◯ MODERATE for concentration
Testosterone	Medium confidence Axelsson et al., 2015a, 2015b (CS) Han et al., 2014 (CS) Jurewicz et al., 2013 (CS) Meeker and Ferguson, 2014 (CS) Meeker et al., 2009 (CS) Mendiola et al., 2011 (C) Pan et al., 2015 (CS) Specht et al., 2014 (CS) Wang et al., 2015c (CS) Low confidence Chang et al. (2015) (CS) Den Hond et al., 2015 (CS) Pan et al., 2006 (CS) Park et al., 2010 (CS)	Minimal concerns for bias overall Risk of bias in low confidence studies and lower study sensitivity account for most observed inconsistency Exposure-response gradient within one study	Difficult to assess exposure-response gradient and precision across studies due to differing analyses/presentation of results	Inverse associations between DEHP exposure and testosterone levels in 8/13 studies (Specht et al., 2014, Meeker and Ferguson, 2014, Wang et al., 2015b, Meeker et al., 2009, Axelsson et al., 2015a, Park et al., 2010, Jurewicz et al., 2013, Pan et al., 2015), 2 of which were statistically significant (Specht et al., 2014, Jurewicz et al., 2013), with the former showing an exposure-response trend, other studies did not examine exposure-response gradient (Table 13).	⨁⨁◯ MODERATE
	Time to Pregnancy				⨁◯◯ SLIGHT
	Hypospadias/cryptorchidism and pubertal development				◯◯◯ INDETERMINATE

C: cohort, CC: case-control, CS: cross-sectional.