Main Findings
We found that women born small or premature may have future risk of GDM. However, the evidence certainty was low, and the presented findings may be overestimated, as we observed some evidence of publication bias. These findings were approximately consistent across the subgroups, including different populations, exposures, and studies of varied methodological quality; these findings were robust in sensitivity analyses.
Our finding that the mother’s size at birth or premature birth may affect GDM risk was consistent with that of a previous narrative review20. The strength of this association was similar to that observed in women with a family history of diabetes mellitus, an established risk factor for GDM42-44. However, the importance of the risk factor in clinical decision depends on the absolute risk difference. Our findings suggested that careful review of the mother’s birth status may indicate her risk of GDM and guide pregnancy management in moderate to high prevalence settings. The mother’s preterm birth status and size at birth are not currently considered risk factors for GDM in any of the major guidelines or risk models 43-46. Our findings may help further refine these guidelines and models or to develop new ones.
The certainty of evidence for the association between premature birth or SGA status and GDM was low due to the high risk of publication bias, as shown by funnel-plot analysis. The contour-enhanced funnel plot suggests that studies with non-significant findings may not have been published. Although we did not identify any ongoing or unpublished studies, this did not eliminate the risk of publication bias, as observational studies are less likely to be registered than clinical trials 47. Thus, the reported estimates may be overestimates. The studies included in this review tended not to adjust for confounders, such as smoking, obesity, socioeconomic status, and family history of diabetes. Future studies should adjust for these factors.
The main result of this review was subject to substantial between-study heterogeneity, as shown by the I 2 statistic27. This heterogeneity may be due to the different types of exposure (LBW, SGA, or preterm birth) considered in this study. However, as all exposure types were associated with increased GDM risk, the high I 2 statistic may be due to the large number of participants and narrow CIs of the primary studies 48. Given these findings, we did not assign a low rating to the inconsistency domain of the GRADE criteria29.
The underlying mechanism of the association between preterm birth or SGA status and subsequent GDM may be gestational malnutrition due to maternal malnutrition or placental insufficiency49. Findings from animal studies have suggested that malnutrition in utero is associated with reduced β-cell counts, pancreas weight, and pancreatic insulin content50-52. According to the Barker foetal origin hypothesis, these foetal programming events may affect the future risk of disease 17. A review of epidemiological studies has suggested that LBW and preterm birth are associated with the risk of type 2 diabetes in adulthood; a similar mechanism is possible for GDM 16.