The effects of lengthened-partial range of motion resistance training of the limbs on arm and thigh muscle cross-sectional area

A multi-site cluster trial

Abstract

It has recently been hypothesised that resistance training (RT) performed using a partial range of motion (ROM) at long muscle lengths (i.e., "lengthened partials") might optimise gains in muscle size. Improvements in muscle size however are typically small, even smaller in trained people due to the linear-logarithmic adaptation to RT over time, and thus between intervention differences in effects are likely to be very small. As such, in contrast to most studies in the field which aim to detect differences between interventions, we sought to conduct a highly powered pre-registered test of the statistical equivalence of two RT interventions in previously trained participants; namely full ROM (fROM) and “lengthened partial” ROM (lpROM). A randomised controlled cluster trial across 15 sites was employed. Our primary outcome was hypertrophy operationalised as muscle cross sectional area (CSA) estimated from circumference and skinfold measurements of the upper arm and thigh. Secondary outcomes were strength operationalised as estimated one-repetition maximum estimated from submaximal repetitions to momentary failure for chest press, leg press, and pulldown machines. At the participant level, participants were randomly assigned to either the lpROM (n = 163) or fROM (n = 134) RT intervention condition. Participants underwent pre-intervention testing and then participated in a 12-week intervention with post-intervention testing following this. Our primary estimand of interest was the condition by time interaction effect from our pre-registered analysis reflecting the standardised between condition difference in change in hypertrophy over time. The estimate for this effect for the arm estimated muscle CSA was -0.032 [95%CI: -0.123, 0.058] and for the thigh estimated muscle CSA was 0 [95%CI: -0.094, 0.094]. The p-values for equivalence were p=0.071 for the arm muscle, and p=0.019 for the thigh muscle. As such, considering our inference criteria with alpha set at 0.01 and adjusted to 0.005 for multiple outcomes, we were unable to reject the null hypothesis that the condition:time interaction effect was outside of the SESOI [-0.1, 0.1]. Main effects for time were small for estimated muscle CSA and also strength, with both being in line with prior predictions from theoretical linear-log growth models. Additional exploratory analysis suggest that both the main effects of time, and any interaction effects for condition by time, are likely small. These findings are in line with other recent evidence regarding the comparison of fROM and lpROM specifically and suggest that between condition effects are likely small and practically equivalent. More broadly, this study highlights that the effects of RT in trained persons should be expected to be small and that current studies in the field of RT are woefully underpowered to be able to detect their effects, let alone test between intervention comparisons.