I T Liu1,2,3,4, F W Liang5, S T Wang3,6, C M Chang3,7,8, T H Lu9, C H Wu10,11. 1. Department of Family Medicine, E-DA Hospital, Kaohsiung, Taiwan. 2. Department of Geriatric Medicine, E-DA Hospital, Kaohsiung, Taiwan. 3. Institute of Geriatrics, College of Medicine, National Cheng Kung University, Tainan, Taiwan. 4. School of Medicine, College of Medicine, I-Shou University, Kaohsiung, Taiwan. 5. Department of Public Health, College of Health Sciences, Kaohsiung Medical University, Kaohsiung, Taiwan. 6. Deputy Superintendent, Ditmanson Medical Foundation, Chiayi Christian Hospital, Chiayi, Taiwan. 7. Department of Geriatric Medicine, National Cheng Kung University Hospital, Tainan, Taiwan. 8. Department of Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan. 9. Department of Public Health, College of Medicine, National Cheng Kung University, Tainan, Taiwan. 10. Institute of Geriatrics, College of Medicine, National Cheng Kung University, Tainan, Taiwan. paulo@mail.ncku.edu.tw. 11. Department of Family Medicine, College of Medicine, National Cheng Kung University Hospital, National Cheng Kung University, 138 Sheng-Li Road, Tainan, 70428, Taiwan. paulo@mail.ncku.edu.tw.
Abstract
Fall is the major risk factor of fracture that has not been included in FRAX®. Whether different age may determine the effect of falls on FRAX® is still uncertain. This epidemiological cohort study reveals that history of fall is a significant predictor of incident fracture independent of FRAX probability, especially in subjects < 75 years old. INTRODUCTION: The Fracture Risk Assessment Tool (FRAX) calculates 10-year fracture risk using 11 clinical risk factors and bone mineral density (BMD); however, it does not include fall history in its risk assessment. Here, we investigated whether fall history is an independent risk factor on fracture prediction after adjustment of FRAX scores in two age subgroups (40-75 and ≥ 75 years). METHODS: Beginning in 2009 to 2010, 1975 people (914 men) from Taiwan were followed for 6.8 ± 1.1 years by matching them with their records in the 2008-2016 National Health Insurance databank. We validated FRAX predictive accuracy with or without fall history by Cox proportional hazards regression. RESULTS: After adjusting for FRAX risk, a history of falling was still a significant predictor of major osteoporotic fractures (MOFs) (using BMD, hazard ratio [HR], 1.47; p = 0.03; without using BMD, HR, 1.54; p = 0.01). A history of recurrent falls was also a significant predictor of both incident MOFs and hip fractures. However, when the subjects were stratified based on age group, a history of falling and recurrent falls were strong predictors of MOFs and hip fractures in the younger but not the older subgroup. CONCLUSION: A fall history can predict incident fracture independently of FRAX probability, particularly in subjects younger than 75 years old.
Fall is the major risk factor of fracture that has not been included in FRAX®. Whether different age may determine the effect of falls on FRAX® is still uncertain. This epidemiological cohort study reveals that history of fall is a significant predictor of incident fracture independent of FRAX probability, especially in subjects < 75 years old. INTRODUCTION: The Fracture Risk Assessment Tool (FRAX) calculates 10-year fracture risk using 11 clinical risk factors and bone mineral density (BMD); however, it does not include fall history in its risk assessment. Here, we investigated whether fall history is an independent risk factor on fracture prediction after adjustment of FRAX scores in two age subgroups (40-75 and ≥ 75 years). METHODS: Beginning in 2009 to 2010, 1975 people (914 men) from Taiwan were followed for 6.8 ± 1.1 years by matching them with their records in the 2008-2016 National Health Insurance databank. We validated FRAX predictive accuracy with or without fall history by Cox proportional hazards regression. RESULTS: After adjusting for FRAX risk, a history of falling was still a significant predictor of major osteoporotic fractures (MOFs) (using BMD, hazard ratio [HR], 1.47; p = 0.03; without using BMD, HR, 1.54; p = 0.01). A history of recurrent falls was also a significant predictor of both incident MOFs and hip fractures. However, when the subjects were stratified based on age group, a history of falling and recurrent falls were strong predictors of MOFs and hip fractures in the younger but not the older subgroup. CONCLUSION: A fall history can predict incident fracture independently of FRAX probability, particularly in subjects younger than 75 years old.