Effects of aerobic, resistance, and combined exercise on postmenopausal women with osteopenia or osteoporosis: a systematic review and network meta-analysis

Objective This study aims to evaluate the clinical effects, characteristics, and comparative advantages of aerobic exercise, resistance exercise, and their combination in postmenopausal women with osteopenia or osteoporosis. Methods A systematic review was conducted following PRISMA guidelines and registered in PROSPERO (CRD42023404292). We searched PubMed, Embase, Web of Science, Cochrane Library, CNKI, Wanfang, VIP, and SinoMed from inception to 31 August 2026 for parallel or crossover randomized controlled trials enrolling naturally postmenopausal women with osteopenia (T-score −2.5 to −1.0) or osteoporosis (T-score≤ − 2.5) according to WHO criteria, comparing aerobic, resistance, or combined aerobic and resistance exercise with blank control, routine treatment, or routine treatment plus anti-osteoporotic drugs, and reporting at least one outcome of interest (lumbar spine, femoral neck, or total hip BMD; VAS score; or serum calcium, phosphorus, ALP, or P1NP). A random-effects network meta-analysis was performed using Stata 18.0, with mean differences or standardized mean differences (95% CI) as effect measures, I 2 to quantify heterogeneity, SUCRA for treatment ranking, and node-splitting to assess consistency. Certainty of evidence was evaluated using GRADE. Results A total of 30 studies involving 2,149 participants were included. For lumbar spine BMD, all three exercise modalities significantly improved BMD compared with control, with aerobic exercise ranking highest (MD = 0.042 g/cm 2 , 95% CI: 0.016–0.069, SUCRA = 0.854), followed by resistance (MD = 0.034 g/cm 2 , 95% CI: 0.019–0.049, SUCRA = 0.692) and combined exercise (MD = 0.024 g/cm 2 , 95% CI: 0.001–0.046, SUCRA = 0.447). At the femoral neck, aerobic (MD = 0.042 g/cm 2 , 95% CI: 0.004–0.080, SUCRA = 0.722) and resistance (MD = 0.039 g/cm 2 , 95% CI: 0.016–0.062, SUCRA = 0.678) exercise were both effective, whereas combined exercise did not reach statistical significance (MD = 0.033 g/cm 2 , 95% CI: −0.013 to 0.078, p = 0.157). At the total hip, resistance exercise was the only modality that significantly increased BMD (MD = 0.062 g/cm 2 , 95% CI: 0.029–0.096, SUCRA = 0.959). None of the three modalities significantly improved VAS scores, serum calcium, serum phosphorus, ALP, or P1NP levels. Subgroup analyses revealed that heterogeneity was substantially reduced in high-frequency (>3 sessions/week) subgroups for lumbar spine BMD ( I 2 = 8.2%), in blank-control and osteopenia subgroups for femoral neck BMD (both I 2 = 0.0%), and in moderate-to-high intensity subgroups for total hip BMD ( I 2 ≤ 21.4%). The routine treatment plus drug subgroup consistently exhibited the highest heterogeneity across all three BMD sites ( I 2 = 82.0–94.8%). Conclusion Exercise significantly improves BMD at the lumbar spine, femoral neck, and total hip in postmenopausal women with osteopenia or osteoporosis, with the optimal modality being site-specific. Aerobic exercise is preferred for the lumbar spine, resistance exercise for the total hip, and either modality is suitable for the femoral neck. Exercise frequency is a key determinant of lumbar spine BMD consistency, while intensity is critical for total hip BMD gains. None of the exercise modalities significantly improved pain or biochemical markers, suggesting that the current evidence does not support their routine use as primary measures for evaluating short-to-medium term exercise efficacy. A site-directed approach to exercise prescription—selecting the modality that ranks highest for the patient’s primary site of BMD decline—may be considered for clinical practice.