. 2017 Oct 5;125(10):106001. doi: 10.1289/EHP1233

Table 8.

Quality of the evidence of the overall effects of bisphenol A on the investigated obesity related outcome measures using the Office of Health Assessment and Translation confidence rating methodology (NTP 2015).

Outcome measure	Body of evidence (animal studies)	Factors for downgrading					Factors for upgrading				Final confidence rating
Outcome measure	Body of evidence (animal studies)	Risk of bias	Unexplained inconsistency	Indirectness	Imprecision	Publication bias	Magnitude^a	Dose response	Residual confounding^a	Consistency species	Final confidence rating
Body weight	Initial high confidence (55 studies)	$↓$ ^b	$↓$ ^c	––	––	––^d	––	––^e	––	––	Low
Fat weight	Initial high confidence (13 studies)	$↓$ ^f	$↓$ ^g	––	––	––^d	––	––^e	––	––	Low
Triglyceride	Initial high confidence (8 studies)	$↓$ ^f	$↓$ ^h	––	––	––^d	––	––^e	––	––	Low
Free fatty acids	Initial high confidence (7 studies)	$↓$ ^f	$↓$ ⁱ	––	––	––^d	––	––^e	––	$↑$ ^j	Low to Moderate
Leptin	Initial high confidence (8 studies)	$↓$ ^b	$↓$ ^g	––	––^k	––^d	––	––^e	––	––	Low

Note: –, no concern, or not present; $↓$ , serious concern; $↑$ , sufficient to upgrade evidence.

^{^a}

The factors “large effect magnitude” and “residual confounding” were not assessed in this study and consequently were not used to upgrade the evidence.

^{^b}

Serious concern because of many “unclear” scores and a change in direction of the association after sensitivity analyses.

^{^c}

Serious concern because of varying point estimates, minimal or no overlap of confidence intervals between studies, and substantial heterogeneity ( $I^{2} > 75 %$ ).

^{^d}

No strongly suspected publication bias observed.

^{^e}

Indications for dose–response effects either within or across studies, but the consistency of these indications was not considered sufficient to upgrade the confidence.

^{^f}

Serious concern because of many “unclear” scores.

^{^g}

Serious concern because of varying point estimates and minimal overlap of confidence intervals between studies.

^{^h}

Serious concern because of varying point estimates and moderate heterogeneity ( $I^{2} > 50 %$ ).

^ⁱ

Serious concern because of varying point estimates.

^{^j}

No subgroup differences were estimated across species and strains.

^{^k}

Body of evidence was already downgraded for unexplained inconsistency and additional downgrading for imprecision was not considered appropriate (NTP 2015).