Give it a rest: a systematic review with Bayesian meta-analysis on the effect of inter-set rest interval duration on muscle hypertrophy

Abstract

We systematically searched the literature for studies with a randomized design that compared different inter-set rest interval durations for estimates of pre-/post-study changes in lean/muscle mass in healthy adults while controlling all other training variables. Bayesian meta-analyses on non-controlled effect sizes using hierarchical models of all 19 measurements (thigh: 10; arm: 6; whole body: 3) from 9 studies meeting inclusion criteria analyses showed substantial overlap of standardized mean differences across the different inter-set rest periods [binary: short: 0.48 (95%CrI: 0.19-0.81), longer: 0.56 (95%CrI: 0.24-0.86); Four categories: short: 0.47 (95%CrI: 0.19-0.80), intermediate: 0.65 (95%CrI: 0.18-1.1), long: 0.55 (95%CrI: 0.15-0.90), very long: 0.50 (95%CrI: 0.14-0.89)], with substantial heterogeneity in results. Univariate and multivariate pairwise meta-analyses of controlled binary (short vs. longer) effect sizes showed similar results for the arm and thigh with central estimates tending to favor longer rest periods [arm: 0.13 (95%CrI: -0.27 to 0.51); thigh: 0.17 (95%CrI: -0.13 to 0.43)]. In contrast, central estimates closer to zero but marginally favoring shorter rest periods were estimated for the whole body [whole body: -0.08 (95%CrI: -0.45 to 0.29)]. Subanalysis of set end-point data indicated that training to failure or stopping short of failure did not meaningfully influence the interaction between rest interval duration and muscle hypertrophy. In conclusion, results suggest a small hypertrophic benefit to employing inter-set rest interval durations >60 s, perhaps mediated by reductions in volume load. However, our analysis did not detect appreciable differences in hypertrophy when resting >90 s between sets, consistent with evidence that detrimental effects on volume load tend to plateau beyond this time-frame. Systematic Review Registration: OSF, https://doi.org/10.17605/OSF.IO/YWEVC.

Keywords: muscle cross-sectional area; muscle development; muscle growth; muscle thickness; recovery interval; rest period.

Figure 1

PRISMA flow chart of the search process.

Figure 2

Meta-analysis of non-controlled effect sizes separated by binary categorization of short (≤60 s) vs. long (>60 s) inter-set rest periods. Plots illustrate shrunken posterior distribution of effect sizes following application of meta-analytic model. Circle: median, error bars represent 75 and 95% credible intervals. Small, medium, and large effect size thresholds are presented according to previous research in strength and conditioning (22).

Figure 3

Meta-analysis of non-controlled effect sizes separated by short (≤60 s), intermediate (61 s–119 s), long (120–179 s), and very long (≥180 s) categorization of inter-set rest period. Plots illustrate shrunken posterior distribution of effect sizes following application of meta-analytic model. Circle: median, error bars represent 75 and 95% credible intervals. Small, medium, and large effect size thresholds are presented according to previous research in strength and conditioning (22).

Figure 4

Meta-analysis of controlled effect sizes of muscular hypertrophy of the upper arm with direct comparisons of binary categorization of inter-set rest period. Plots illustrate shrunken posterior distribution of effect sizes following application of meta-analytic model. Circle: median, error bars represent 75 and 95% credible intervals. Small, medium, and large effect size thresholds are presented according to previous research in strength and conditioning (25). Probability of effect size greater than 0 favoring longer rest period = 0.74; probability of effect size greater than small favoring longer rest period = 0.45; probability of effect size greater than medium favoring longer rest period = 0.18; probability of effect size greater than large favoring longer rest period = 0.03.

Figure 6

Meta-analysis of controlled effect sizes of muscular hypertrophy of the whole body with direct comparisons of binary categorization of inter-set rest period. Plots illustrate shrunken posterior distribution of effect sizes following application of meta-analytic model. Circle: median, error bars represent 75 and 95% credible intervals. Small, medium, and large effect size thresholds are presented according to previous research in strength and conditioning (25). Probability of effect size greater than 0 favoring short rest period = 0.69; probability of effect size greater than small favoring short rest period = 0.36; probability of effect size greater than medium favoring short rest period = 0.12; probability of effect size greater than large favoring short rest period = 0.01.

Figure 5

Meta-analysis of controlled effect sizes of muscular hypertrophy of the thigh with direct comparisons of binary categorization of inter-set rest period. Plots illustrate shrunken posterior distribution of effect sizes following application of meta-analytic model. Circle: median, error bars represent 75 and 95% credible intervals. Small, medium, and large effect size thresholds are presented according to previous research in strength and conditioning (25). Probability of effect size greater than 0 favoring longer rest period = 0.88; probability of effect size greater than small favoring longer rest period = 0.54; probability of effect size greater than medium favoring longer rest period = 0.15; probability of effect size greater than large favoring longer rest period = 0.01.