Cardio can eat into lifting gains — how much depends on the dose and the modality

Most training advice tells people to do both cardio and strength work, yet coaches have long suspected the two adaptations compete — the so-called interference effect. What has been less clear is which specific features of endurance training actually cause the problem: is it the modality, how often it is done, or how long each session lasts? This meta-analysis set out to identify which components of endurance training are detrimental to resistance training outcomes, giving the interference question a quantitative answer rather than a rule of thumb.

The authors pooled 21 studies yielding a total of 422 effect sizes. Included studies had to compare strength training alone against strength plus endurance training, or compare combinations of concurrent training, and report at least one measure of strength, power or hypertrophy with the data needed to calculate effect sizes. Beyond the main comparisons, the analysis examined moderator variables — endurance modality (running versus cycling), training frequency and session duration — and ran correlational analyses between those variables and the size of the training effects.

For hypertrophy, the mean effect size was 1.23 for strength training alone, 0.85 for concurrent training and 0.27 for endurance training alone. For strength, the pattern was similar: 1.76, 1.44 and 0.78 respectively. Power showed significant differences between all three groups, at 0.91 for strength training, 0.55 for concurrent and 0.11 for endurance alone. Among moderators, lifting concurrently with running — but not cycling — produced significant decrements in both hypertrophy and strength. Interference correlated with endurance frequency (-0.26 to -0.35) and duration (-0.29 to -0.75), and decreased body fat correlated with percentage of maximal heart rate (r = -0.60).

This analysis helped move the interference effect from anecdote to quantified phenomenon, showing the cost of concurrent training is real but graded — concurrent lifters still gained substantially more than endurance-only trainees on every outcome. Its most practical contribution is the dose-response signal: more frequent and longer endurance sessions predicted greater interference. The modality finding — running implicated, cycling not — remains an open question, since other analyses of concurrent training have probed modality with differing results, and the correlational nature of the moderator data means these are patterns across studies rather than direct experimental tests.

If you lift and do cardio, this is a reason to think about the total endurance dose, not just whether you do cardio at all. In these pooled data, the interference with muscle and strength gains scaled with how often and how long the endurance sessions ran, and power was the quality most affected. It is also worth noting what the data do not say: concurrent training still produced solid gains in size and strength — the effect is a partial blunting relative to lifting alone, not a cancellation.