Relationship between proton pump inhibitors and the risk of fractures in the observational studies: a systematic review and Meta-analysis

Abstract

Objective To systematically evaluate the relationship between proton pump inhibitors (PPI) and the risk of fractures. Methods The related databases were electronically searched for the cohort studies and the case-control studies about the relationship between proton pump inhibitors and the risk of fractures from inception to July 12, 2016. Meta-analysis were conducted using RevMan 5.3 softwear. The results were presented as odds ratio (OR) or relative risk (RR) and 95% confidence interval (CI). The evidence quality of the enrolled literatures were evaluated by Grades of Recommendation, Assessment, Development, and Evaluation. Results A total of 20 reports which were related to 21 studies involving 9 cohort studies, 12 case-control studies and 2 076 154 subjects were enrolled in the study. There were 262 298 cases who used PPI and 1 813 856 cases who did not use PPI in the 2 076 154 subjects. The pooled results of Meta-analysis of cohort studies showed that compared with the subjects who had not used PPI, the risk of fracture increased 85% in the subjects who used PPI (OR=1.85, 95%CI: 1.55-2.21, P<0.000 01); the risk of hip fracture increased 58% (OR=1.58, 95%CI: 1.07-2.33, P<0.000 01); the risk of fracture in the subjects who were ≥50 years old increased 58% (OR=1.58, 95%CI: 1.47-1.70, P<0.000 01). The pooled results of Meta-analysis after combination of case control studies showed that compared with the subjects who had not used PPI, the risk of fracture increased 39% in the subjects who used PPI (OR=1.39, 95%CI: 1.26-1.52, P<0.000 01); the risk of hip fracture increased 45% (OR=1.45, 95%CI: 1.23-1.69, P<0.000 01); the risk of fracture in the subjects who were ≥50 years old increased 43% (OR=1.43, 95%CI: 1.14-1.79, P=0.002). There were no explicit dose-effect relation and time-effect relation between PPI and hip fracture. The risk of fracture in female and male patients who used PPI increased 110% and 88% as compared with those in female and male patients who never used PPI (OR=2.10, 95%CI: 1.50-2.92, P=0.000 1; OR=1.88, 95%CI: 1.14-3.11, P=0.01), respectively. The result of quality evaluation by GRADE showed that the risk factors (including fracture, hip fracture, female, male, and the patient’s age 50-years or older) which increased the degree of fracture after PPI use were all very low quality evidences. Conclusion PPI use modestly could increase the degree of risk of all kinds of fracture and hip fracture, especially in females. Key words: Proton pump inhibitors; Fractures, bone; Meta-analysis