Effects of Taekwondo on Bone Mineral Density: A Systematic Review and Meta-analysis of Short-term Interventions and Long-term Participation

Context: Taekwondo provides impact and multidirectional loading, but its effects on bone mineral density (BMD) across life stages remain uncertain. Objective: To evaluate femoral, spinal, and whole-body BMD after short-term taekwondo interventions and long-term participation. Data Sources: PubMed, Web of Science, KISS, RISS, DBpia, CNKI, and Wanfang Data were searched from inception through July 22, 2026. Study Selection: Controlled studies of healthy participants reporting dual-energy x-ray absorptiometry-derived BMD after 8 to 16 weeks of training or ≥2 years of participation. Study Design: Systematic review and meta-analysis. Level of Evidence: Level 2. Data Extraction: Two reviewers independently screened studies, extracted data, and assessed risk of bias. Random-effects models generated mean differences (MDs) with 95% CIs; life-stage-specific analyses were primary. Results: A total of 18 studies involving 820 participants were included: 8 short-term randomized controlled trials and 10 cross-sectional studies of long-term participation. Long-term participants had higher femoral-region BMD among children and adolescents (MD, 0.15 g/cm 2 ; 95% CI, 0.08-0.21) and young adults (MD, 0.17 g/cm 2 ; 95% CI, 0.10-0.23). Lumbar spine BMD was also higher in both life-stage groups (MD, 0.08 and 0.09 g/cm 2 , respectively). For short-term interventions, the most consistent finding was higher femoral neck BMD in children and adolescents (MD, 0.14 g/cm 2 ; 95% CI, 0.09-0.19), which remained stable in change-score analyses. Whole-body findings were less consistent, and several adult results came from single studies. All short-term trials were judged to have a high overall risk of bias; among the long-term studies, 9 had a serious overall risk and 1 had a critical risk of bias. Conclusion: Taekwondo may support site-specific skeletal health, particularly femoral neck BMD during growth. Evidence in middle-aged and older adults remains limited, long-term associations are vulnerable to residual confounding, and an effective training dose cannot yet be defined.