“…Analyses were performed throughout to explore the influence of methodological decisions, such as using unadjusted data and Asian-specific BMI categories. The conversion of categorical BMI was necessary due to limited reporting of directly comparable obesity categories: 17 studies combined data for obesity classes I-III (19,34,35,38,(42)(43)(44)(48)(49)(50)(51)(52)(53)(54)(55)(56)(57), three reported obesity classes I-III separately (58)(59)(60), four combined obesity classes I and II (20,41,61,62), six combined classes II and III (39,40,(63)(64)(65)(66), seven had further inconsistent noncomparable categories such as combining overweight and obese (36,37,(67)(68)(69)(70)(71), and two studies did not define their BMI categories (21,72). The possible groups to combine for categorical analyses would have been further reduced when applying additional analysis criteria such as the gestational age stratification, definition of the reference BMI group, etc.…”