REF-800 · Rønnestad et al., 2010 · European Journal of Applied Physiology, Vol. · Paper · PubMed

In-Season Strength Maintenance Training Increases Well-Trained Cyclists' Performance

Blimited evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

Well-trained cyclists either added heavy strength training twice a week to their usual endurance training for a 12-week preparatory period and then kept one strength session a week through the first 13 weeks of the competition period (E + S, n = 6), or did usual endurance training throughout (E, n = 6).

After the preparatory period only E + S had gained thigh muscle cross-sectional area and 1RM; both groups improved maximal oxygen consumption and 40-min all-out power.

At 13 weeks into the season E + S had kept the muscle and strength gains, and over the whole intervention they improved Wingate peak power, power at 2 mmol/L lactate, maximal aerobic power (Wmax) and 40-min all-out power, the last two by more than the endurance-only group, whose Wmax and power at 2 mmol/L did not change.

Participants

n=6 per group, well-trained cyclists:

E + S (endurance + strength): n = 6 (6 men)
E (endurance only): n = 6 (5 men, 1 woman)

Age and physiological values are not given in the abstract.

Protocol

Preparatory period, 12 weeks: E + S added heavy strength training twice a week to usual endurance training; E did usual endurance training only.
Competition period, first 13 weeks: E + S did strength maintenance training once a week; E continued endurance training only.
Measures: thigh muscle cross-sectional area (CSA), 1RM leg strength, Wingate peak power, power at 2 mmol/L blood lactate, Wmax, maximal oxygen consumption, and mean power in a 40-min all-out trial.

Key Results

After the preparatory period: thigh muscle CSA and 1RM rose in E + S (p < 0.05) but not in E. Maximal oxygen consumption and mean power over the 40-min all-out trial rose in both groups (p < 0.05).
At 13 weeks into the competition period: the CSA and strength gains were still intact in E + S.
Across the whole intervention (start of preparatory period to week 13 of competition), E + S improved Wingate peak power, power at 2 mmol/L lactate, Wmax and 40-min all-out mean power (p < 0.05); the relative improvements in Wmax and 40-min power were larger than in E (p < 0.05).
In E, Wmax and power at 2 mmol/L lactate were unchanged.
The abstract gives no magnitudes (percent or watts) for any change.

Practical Application

In well-trained cyclists (6 per group), heavy strength training twice a week for a 12-week preparatory period followed by one session a week through the first 13 weeks of racing kept the gains in thigh muscle size and strength and, over the whole intervention, improved Wingate peak power, power at 2 mmol/L lactate, Wmax and 40-min all-out power, the last two more than in endurance-only controls. To "can I drop to one gym session a week in season?": in this trial one a week was enough to keep the strength gains, and performance kept improving on top.
Do not quote a percentage gain from this entry; the abstract reports none.
The once-a-week dose was the only maintenance dose tested, and the controls never lifted. Not from this paper: what happens when lifting stops altogether is REF-744 (same group, gains decayed within two months of stopping).
Not from this paper; the base's own practice: this is the primary trial behind the in-season maintenance default in SYNTH-43; keep the load heavy and cut the frequency, not the other way round (REF-587).

Limitations

n=6 per group; randomisation is not stated in the abstract; the control group included one woman.
The control group did no strength training at any point, so the design compares lifting (then maintenance) with never lifting, not maintenance with stopping.
No magnitudes in the abstract; laboratory 40-min all-out trial rather than a race.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

parallel controlled trial, n=12 (6 vs 6); randomisation not stated in the abstract