Discriminatory Accuracy of Fracture Risk Assessment Tool in Asian Populations: A Systematic Review and Meta-Analysis

Abstract

Background: This review explores the discriminative ability of fracture risk assessment tool (FRAX) in major osteoporotic fracture (MOF) and hip fracture (HF) risk prediction and the densitometric diagnosis of osteoporosis in Asian populations.

Methods: We systematically searched the EMBASE, Cochrane, and PubMed databases from the earliest indexing date to January 2024. Studies were included if FRAX was used to identify future osteoporotic fractures or a densitometric diagnosis of osteoporosis in an Asian population and reported the area under the curve (AUC) values. Meta-analyses were conducted after quality assessment for AUC with 95% confidence intervals across the following categories: standard FRAX without/with bone mineral density (BMD), adjusted FRAX, and BMD alone for fracture prediction, as well as standard FRAX for densitometric diagnosis of osteoporosis.

Results: A total of 42 studies were included. The AUC values for predicting fracture risk using FRAX-MOF with BMD (0.73 [0.70-0.77]) was highest compared to FRAX-MOF without BMD (0.72 [0.66-0.77]), and adjusted FRAX-MOF (0.71 [0.65-0.77]). The AUC values for predicting fracture risk using FRAX-HF with BMD (0.77 [0.71-0.83]) was highest compared to FRAX-HF without BMD (0.72 [0.65-0.80]), and adjusted FRAX-HF (0.75 [0.63-0.86]). The AUC values for BMD alone (0.68 [0.62-0.73]) was lowest for fracture prediction. The AUC values for identifying a densitometric diagnosis of osteoporosis was 0.77 [0.70-0.84] and 0.76 [0.67-0.86] using FRAX-MOF and FRAX-HF, respectively.

Conclusions: FRAX with BMD tends to perform more reliably in predicting HF compared to MOF in Asia. However, its accuracy in predicting fracture risk in Asian populations can be improved through region-specific, long-term epidemiological data.