Neuromechanical associations in vertical jump adaptation after plyometric training: a multivariate meta-analysis

Plyometric training (PT) is widely used to enhance vertical-jump performance, but the associations between muscle-tendon mechanical adaptation, lower-limb sEMG amplitude, and" performance across jump tasks with different neuromechanical demands remain incompletely understood. In particular, whether stiffness adaptations are uniformly beneficial across Stretch-Shortening Cycle(SSC)-dependent and predominantly concentric jump tasks remains unclear.

This study employed a multivariate meta-analysis to quantify associations between changes in vertical stiffness (K_vert), Achilles tendon stiffness (K_tendon), knee-extensor and plantar-flexor sEMG amplitude, and vertical-jump adaptation, while examining whether these associations differed according to jump type and intervention duration.

In strict accordance with PRISMA guidelines, seven major databases were searched for randomized controlled trials from inception to December 2025. Multivariate random-effects models were constructed using the metaSEM package in R, supplemented by Pearson correlation analysis. The GRADE framework was utilized to assess the certainty of evidence.

Twenty-nine randomized controlled trials were included. Plyometric training significantly improved vertical jump outcomes (all p < 0.05; standardized mean differences(SMD) > 0.40), vertical stiffness (SMD = 0.37; 95% CI, 0.07-0.67), and Achilles tendon stiffness (SMD = 0.43). Descriptive K_vert-VJ correlations were positive for CMJ (r = 0.51) and negative for DJ (r = - 0.72). Formal moderator analysis supported differences in the K_vert-VJ association across jump types (LRT = 11.33, df = 2, p = 0.003) but not across intervention-duration categories (LRT = 2.05, df = 2, p = 0.358). Formal moderation by jump type or intervention duration was not observed for K_tendon or the sEMG outcomes, although these analyses were based on relatively few studies.

Plyometric-training adaptations showed formal evidence of task-specific K_vert-VJ associations, whereas duration-specific moderation was not supported. Associations involving K_tendon and sEMG amplitude were less consistent and should be regarded as exploratory.