Risk of fracture following gastric surgery for benign and malignant conditions: A study level pooled analysis of population-based cohort studies.

Background: Metabolic changes may occur following gastric surgery, which has been reported to contribute to bone loss, osteoporosis and even bone fracture. However, the evidence regarding the relationship between gastric surgery for benign and malignant conditions and risk of fracture is controversial. This study was conducted with the aim to evaluate whether gastric surgery is associated with a high risk of fracture.

Methods: Major electronic databases were searched from inception through October 2021 for population-based cohort studies investigating the associations between gastric surgery (including bariatric gastric surgeries and surgeries for gastric benign and malignant gastric tumors) and risk of fracture compared with controls. Pooled relative risks (RRs) with 95% confidence intervals (CIs) were derived using the random-effects Mantel-Haenszel model. Multiple subgroup analyses and sensitivity analyses were carried out to test sources of heterogeneity stratified by various study characteristics and the robustness of the results.

Results: A total of 14 studies comprising 693134 individuals were identified for analysis. The RR for the risk of fracture in people undergoing gastric surgery was 1.45 [95% confidence interval (CI) 1.23 - 1.72; I 95.8%; P < 0.001] compared with that in control populations, among which the fracture sites of upper limb, spine, lower limb, pelvis and hip showed consistent significant results (all P < 0.05), whereas nonsignificant associations was noted for other fracture sites. Significant associations were also observed for patients having total or subtotal gastrectomy (RR 2.22, 95% CI 1.66 to 3.00), gastric bypass (RR 1.48, 95% CI 1.26 to 1.74), and a similar trend was observed for preserved passage procedures (including sleeve gastrectomy, gastric banding, vertical banded gastroplasty and other procedures that preserved the passage through the duodenum and proximal small bowel, in contrast to gastric bypass), though the difference did not reach statistically significant (RR 1.10, 95% CI 0.95 to 1.26). An evident increased risk in the age range from 40-59 years was observed (40-49 years: RR 1.36, 95% CI 1.19-1.55; 50-59 years: RR 2.48, 95% CI 1.58-3.90).

Conclusion: From this large pooled analysis of population-based cohort studies, evidence supports that fracture risk is increased in gastric surgery survivors compared with the control population. Early prevention and effective intervention strategies of bone fracture should be taken from clinicians and health policy makers.

Clinical Trial Registration: PROSPERO (https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=291394), identifier CRD42021291394.