比较运动控制干预对前交叉韧带重建术后神经肌肉与生物力学结局的影响：系统综述与网状Meta分析
简介
这项系统综述与网状Meta分析纳入32项试验、1437例ACL重建后患者，比较运动控制干预与常规康复（症状控制、活动度、力量、步态或功能训练）对神经肌肉及生物力学结局的影响。与常规康复相比，本体感觉训练改善关节位置觉（SMD 1.02，95%CI 0.54–1.50），为中等CINeMA置信度证据。其他干预对平衡和落地控制可能有益但置信度较低：跳跃落地控制训…
英文摘要
BACKGROUND: Neuromuscular-control and biomechanical deficits may persist after anterior cruciate ligament reconstruction (ACLR). Comparative motor-control effects remain uncertain. OBJECTIVES: To compare motor-control interventions after ACLR. DESIGN: Systematic review and network meta-analysis. METHOD: Seven databases were searched to 12 May 2026. Usual rehabilitation comprised conventional symptom control, motion, strengthening, gait, or functional training. Three neuromuscular-control outcomes underwent network meta-analysis; k ≥ 2 biomechanical comparisons were pooled and k = 1 estimates unpooled. RoB 2 assessed bias; CINeMA and GRADE assessed confidence. RESULTS: Thirty-two trials enrolled 1437 participants. Proprioceptive training improved joint position sense versus usual rehabilitation (SMD 1.02, 95% CI 0.54-1.50; moderate CINeMA confidence); jump-landing control training yielded a favorable dynamic-balance estimate versus usual rehabilitation (1.19, 0.59-1.80; low CINeMA confidence) and was sensitive to exclusion of a late-phase multi-arm athlete trial; feedback-enhanced (0.88, 0.43-1.32) and proprioceptive training (0.59, 0.21-0.97) favored postural control (both low CINeMA confidence). Pooled estimates suggested that neuromuscular training may improve knee valgus and knee flexion kinematics during single-leg landing (1.51, 0.51-2.51; 1.07, 0.49-1.65); pooled gait biomechanical symmetry was inconclusive (-0.41, -0.87-0.05). Across networks, τ² ranged from 0.0275 to 0.1698; global tests detected no statistically clear inconsistency (P = 0.087-0.229), although sparse loop structures limited inference. CONCLUSION: Moderate-confidence evidence supported proprioceptive training for improving joint position sense after ACLR. Other interventions may help balance and landing control, but their benefits are less certain. The evidence does not establish a single best approach across outcomes.