Abstract
Supervision during resistance training (RT) may enhance strength gains by optimizing trainee effort. We investigated supervision’s role on effort during RT in a unique setting with private strength clinics, where members train either unsupervised (“Core” membership) or supervised by a qualified exercise scientist (“Assisted” membership). Using both retrospective analysis of member training records and a prospective experimental study, we examined supervision’s impact on exercise performance, measured as time under load (TUL) where a longer TUL at a given load is indicative of greater effort, rating of perceived effort (RPE), and rating of perceived discomfort (RPD). Bayesian methods were applied, using empirically informed prior distributions from retrospective data to model the experimental study. The previous observational sample included ~1000 members training sessions from each membership type, while the experimental study involved 45 Core members performing both supervised and unsupervised sessions in randomized order, using their current training loads to momentary failure. Our findings suggest that, in real-world settings (in situ), exercise performance differed little between supervised and unsupervised training with trainees likely using suboptimal effort under both conditions. However, in our experimental study, supervision improved TUL (Core = 125.12 [95%QI: 113.70, 131.90] sec; Assisted = 147.35 [95%QI: 134.29, 154.81] sec; contrast = -22.10 [95%QI: -26.60, -17.61] sec). In percentage points RPE was slightly higher with supervision in both previous observational real-world (Core = 53% [95%QI: 51%, 55%]; Assisted = 59% [95%QI: 57%, 61%]; contrast = -6% [95%QI: -8%, -4%]) and experimental settings (Core = 81% [95%QI: 75%, 86%]; Assisted = 87% [95%QI: 83%, 91%]; contrast = -6% [95%QI: -10%, -4%]), suggesting trainees push closer to failure under supervision. This was further supported by higher RPD during the experimental study (Core = 6.3 [95%QI: 5.1, 7.3]; Assisted = 7.5 [95%QI: 6.5, 8.3]; contrast = -1.2 [95%QI: -1.6, -0.9]). Overall, these results reinforce research on the benefits of supervision in RT, indicating that, during a single session under experimental conditions, trainees likely train with greater effort when supervised.