Effect of exercise on slowing breastfeeding‐induced bone loss: A meta‐analysis and trial sequential analysis

AIM This study systematically reviewed evidence regarding the effectiveness of exercise in slowing breastfeeding-induced bone loss. METHODS The evidence-based approach of a systematic review (PROSPERO registration No. CRD42019111623) was adopted. The inclusion criteria were randomized controlled trials or observational studies. Study samples were breastfeeding women, the intervention was any form of exercise, and bone mineral density (BMD) of the total body, lumbar spine and hip/femur neck before 6 months and at 1 year were the outcome measures. Meta-analyses were performed using a random effect model, and calculations of mean differences of BMD change and 95% confidence intervals (CIs) were carried out. Studies were further evaluated through trial sequential meta-analysis (TSA), and the 'Grading of Recommendations Assessment, Development and Evaluation' methodology was used to assess the certainty of evidence (CoE). RESULTS A total of 1049 studies were screened, and 4 met the inclusion criteria. Weight-bearing aerobic exercise and resistance training before the 6-month evaluation slowed breastfeeding-induced bone loss in the lumbar spine (1.62% BMD change [95% CI = 0.53-2.71]; I2 = 8%). The TSA Z-curve revealed crossing of the TSA boundary and line of information size, indicating sufficient sampling and significance. The CoE of exercise benefit in the lumbar spine at 6 months was low, whereas the CoE for other areas ranged from low to very low. CONCLUSION This first systematic review and meta-analysis provided some evidence of the advantages of exercise for slowing breastfeeding-induced bone loss. However, additional randomized controlled trials are warranted to generate more conclusive evidence.