Twenty-one studies were included in the review: four intervention studies (n=157 for three studies; n not stated for fourth); seven treatment studies (n=228); three prospective cohorts (n=27,933); and seven retrospective studies (192,859). The mean participation rate was 81 per cent and the mean attrition rate was 7.5 per cent (where reported). Only two studies described conceptual frameworks.
Five studies (two prospective cohorts; three retrospective) found a moderate effect size (0.40 to 0.46) favouring physical activity before and/or during pregnancy for reducing the risk of onset of GDM. One intervention study (n=96) found a medium–to-large effect on mean maternal weight increase per week, favouring exercise and diet compared to diet only (effect size 0.50).
Three studies (two intervention and one treatment, n=approximately 146) found that exercise did not appear to affect gestational age at delivery compared to control conditions (d=0.13). Results for infant birth weight were inconsistent, but overall showed little difference between the groups (d=-0.04, two intervention and two treatment studies n=146).
Other results were reported in the review.