SUN-268 Vitamin D Supplementation and Falls in Adults: A Systematic Umbrella Review of Meta-Analyses of Randomized Controlled Trials

Abstract M.T. Chakhtoura: None. J.E. Nassar: None. A. Slim: None. B. Slim: None. J. Yammine: None. N. Wahoud: None. M.M. Rahme: None. G. El-Hajj Fuleihan: None. Context: Systematic reviews/meta-analyses (SRs/MAs) of trials investigating the efficacy of Vitamin D (VitD) (± Calcium (Ca)) on fall prevention led to contradictory results and guidelines. Objective: We aim to clarify this controversy with an umbrella review coupled with a rigorous evaluation of the quality of the SRs/MAs and the reasons for discrepancies between them. Methods: We searched 5 databases (2010-2022), references of reviews, contacted vitD experts, and used A MeaSurement Tool to Assess systematic Reviews 2 (AMSTAR-2) for quality assessment. We compared results and investigated reasons for discrepancy using matrices and subgroup analyses (PROSPERO registration: CRD42019129540). Results: We included 20 SRs, with 25 MAs for fallers and 8 MAs for Falls, all of low/critically low quality, except 1 of moderate quality. Numbers of Fallers: 3 SRs (5 MAs) on institutionalised participants had null findings: 2 MAs on VitD±Ca versus (vs) placebo (PBO)/Ca, 1 MA on VitD+Ca (Ca/D) vs control, and 2 MAs pooled VitD vs PBO and Ca/D vs Ca. 6 SRs (8 MAs) targeted community-dwelling participants; 2 MAs on VitD vs PBO had null findings. Conversely, 1/5 MAs on VitD±Ca vs PBO/Ca reported a relative risk (RR) 0.83(0.77;0.89) and 1 MA on Ca/D vs PBO/Ca reported an odds ratio (OR) 0.87(0.80;0.94).10 SRs (12 MAs) pooled data of community-dwelling and institutionalised participants. 3 MAs, 1 MA on VitD vs PBO, 1 MA on Ca/D vs control, and 1 MA pooled Ca/D vs Ca and VitD vs PBO, had null findings. Conversely, 4/8 MAs reported positive findings on VitD ± Ca vs PBO/Ca, OR 0.86(0.79;0.93), RR range 0.86-0.95, and 1 MA on Ca/D vs Ca with an OR 0.73(0.58;0.90). Numbers of Falls: 2 SRs/MAs on institutionalised participants had identical findings; 1 MA pooled VitD vs PBO and Ca/D vs Ca and 1 MA on VitD±Ca vs PBO/Ca, with a RR and an OR of 0.72 (0.55;0.95). Null findings were also reported in 2 SRs/MAs on community-dwelling participants on VitD±Ca vs PBO/Ca. 3 SRs (4 MAs) pooled data from community-dwelling and institutionalised participants: 1 MA was on VitD vs PBO had null findings, whereas two had positive findings, 1 MA on Ca/D vs PBO/Ca with a RR 0.88(0.80;0.87) and 2 MAs on VitD±Ca vs PBO/Ca with an OR 0.79(0.70;0.88), and a RR 0.91(0.85;0.97). Discrepancies in the results were attributed to the heterogeneity in the trials included, in definition of falls/fallers, and inconsistent sample sizes in MAs using the same trials leading to different effect sizes for these trials. Conclusion: VitD±Ca had no effect on the number of fallers. Its potential in preventing the number of falls in institutionalised participants was suggested but remains uncertain given the low quality of evidence. Future MAs should prioritise individual patient data, considering age, residency, and baseline 25-hydroxyvitamin D levels, to better assess the efficacy of VitD in fall prevention. Sunday, June 2, 2024