Lifting
Adding load pulled ahead late, as fixed-load sets drifted far from failure
In untrained young adults, both legs grew; the progressing leg only gained significantly more between sessions 18 and 24, once the fixed load had gotten easy.
Thirty-one untrained young adults trained one leg with load increases and the other with a constant load, using knee extensions for 3 sets of 10 to 12 reps, twice a week for 12 weeks. Both legs grew. Muscle cross-sectional area rose 3.36 cm² on the progressing side and 2.67 cm² on the fixed side, but muscle size did not differ significantly between legs at any single measurement point.
The separation came late. Between sessions 18 and 24, the progressing leg gained 1.22 cm² more than the fixed leg (95% CI 0.59 to 1.84). Effort tells part of the story: reps in reserve in the fixed leg climbed from 1.50 at the first session to 4.08 at session 24, while the progressing leg stayed near failure at 0.47 and 0.20. The biggest limitation is the population: untrained young adults on a single exercise, so this says little about experienced lifters.
In practice. Supports raising the load as sets get easier, at least in untrained young adults training one exercise over 12 weeks.
Trust level: Moderate. Randomized within-person design over 12 weeks reduces between-person noise, but 31 untrained participants, one exercise and no significant difference at individual time points limit confidence.
In untrained women, adding load grew arm muscle more than keeping it fixed
Over eight weeks of arm curls, raising the load as reps allowed produced greater muscle thickness gains than a fixed load, and both beat no training.
Forty-one untrained young women did elbow curls one arm at a time for 8 weeks, three sessions a week, 3 sets of 8 to 12 reps. Arms were randomly assigned to raise the load once the top of the rep range was reached, to keep the load fixed, or to a no-exercise control. Both training approaches thickened the elbow flexors more than control, and progressing the load produced larger gains than a fixed load. The abstract does not report how large the difference was.
The authors also measured how much the muscle swelled during a single session before and after the programme, as a stand-in for how much muscle was used. Before training the swelling was similar across arms. Afterward it was larger in control and fixed-load arms than in progressing arms. They read this as progressive overload keeping muscle use higher. That mechanism is the weakest part: it rests on a proxy measure, in untrained young women only.
In practice. Adds to evidence that raising the load over time drives more muscle growth than a fixed load, at least in untrained young women over 8 weeks.
Trust level: Moderate. Randomized, with a control condition and a clear direction of effect, but the abstract gives no effect sizes, the groups are untrained women over only 8 weeks, and the mechanism is inferred from a proxy.
Time-restricted feeding with lifting showed no fat-free mass edge in five small trials
Across five randomized trials, healthy adults who lifted and used time-restricted feeding did not end up with more fat-free mass than those eating conventionally.
This meta-analysis pooled five randomized trials with 120 healthy adults in total, all doing strength training, and compared time-restricted feeding (eating inside a set daily window) with conventional eating. Fat-free mass stayed similar between groups by the end. Only two of the five trials found a significant difference in body weight and fat mass.
The meta-analysis’s pooled estimate was -0.48 kg, with a 95% CI from -2.52 to 1.56 kg. That interval includes zero, but it is wide: it is compatible with a between-group difference of about 2.5 kg in one direction to about 1.6 kg in the other, so five small trials cannot rule out a real difference in either direction. Intervention length and risk-of-bias results are not reported in the abstract. The main limitation is simply how little data there is.
In practice. Suggests that time-restricted feeding offers no clear fat-free mass advantage over conventional eating in healthy adults doing strength training, though the evidence is too thin to exclude small differences.
Trust level: Low. Only five randomized trials and 120 participants, a very wide confidence interval, and no duration or risk-of-bias results given in the abstract.
Beyond the gym
Cold-water immersion was linked to less soreness, with no standout water temperature
A pooled analysis of 52 studies linked a single cold-water immersion session to less soreness and lower muscle damage markers, but could not show that colder water worked better.
This meta-analysis pooled 52 studies of a single cold-water immersion session after exercise. Immersion was associated with lower muscle soreness (DOMS) and lower biochemical markers of muscle damage, creatine kinase and myoglobin. The effects, expressed as standardized mean differences (effect size in standard-deviation units), were rated moderate to large: -0.81 for soreness, -0.64 for creatine kinase and -0.93 for myoglobin.
Temperature is where the story weakens. Soreness fell numerically less at about 5 °C (-0.2) than at about 10 °C (-0.8) or 15 °C (-0.9), but that gap was not statistically significant (p = 0.79), and the authors label the comparison exploratory. The main caveat is consistency: heterogeneity was substantial and prediction intervals were wide, so the effect may not reproduce in a given setting. The abstract does not say who was studied or what exercise preceded immersion.
In practice. Adds to evidence that cold-water immersion is linked to less post-exercise soreness and muscle damage markers on average, with no sign that any water temperature is superior.
Trust level: Moderate. A large pool of 52 studies supports the average effect, but heterogeneity was substantial, the review was not prospectively registered, and the temperature comparison was exploratory and not significant.