Concurrent training: a meta-analysis examining interference of aerobic and resistance exercises.
Level 1 - systematic review of randomized trials
Meta-analysis of 21 comparative studies evaluating concurrent vs single-modality exercise training
PubMed 22002517 · doi:10.1519/JSC.0b013e31823a3e2d
What was done
A meta-analysis of 21 studies (422 effect sizes) was performed to evaluate which endurance training components (modality, duration, frequency) interfere with resistance training adaptations. Studies met criteria if they compared strength training alone to strength plus endurance training (concurrent) or combinations of concurrent training, reported at least one measure of strength, power, or hypertrophy, and provided adequate data to calculate effect sizes (ES).
What was found
Mean ES for muscle hypertrophy was 1.23 for strength training, 0.27 for endurance training, and 0.85 for concurrent training, with strength and concurrent training significantly greater than endurance. Mean ES for strength was 1.76 for strength, 0.78 for endurance, and 1.44 for concurrent training (strength and concurrent significantly greater than endurance). Mean ES for power was 0.91 for strength, 0.11 for endurance, and 0.55 for concurrent training, with significant differences among all three groups. Moderator analysis showed concurrent running, but not cycling, led to significant decrements in hypertrophy and strength. Significant negative correlations were identified between endurance frequency (r = -0.26 to -0.35) and duration (r = -0.29 to -0.75) for hypertrophy, strength, and power. Decreased body fat ES correlated significantly with percentage of maximal heart rate (r = -0.60, p < 0.05).
Why it matters
This review shows that interference from aerobic exercise on resistance adaptations is specific: power development is most vulnerable, and running alongside high frequency or long duration endurance work impairs strength and muscle growth more than cycling.
Limits
The abstract does not specify participant characteristics (e.g., sex, age, trained status) or total participant count. Heterogeneity metrics, confidence intervals, and risk of bias across the 21 included studies are not detailed in the abstract.