REF-952 · Foster et al., 1995 · European Journal of Applied Physiology and Occupational Physiology, Vol. · Paper · PubMed

Effects of specific versus cross-training on running performance

B+moderate evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

The cross-training (XT) hypothesis says athletes may get better in one exercise mode by training in a different one.

Thirty well-trained people (10 men, 20 women) were randomised either to 8 weeks of training raised by +10%/week — through extra running (R) or extra swimming (XT) on top of baseline running — or to baseline running as before (C). 3.2 km time-trial performance improved with both R (-26.4s) and XT (-13.2s) but not with C (-5.4s).

No group showed a significant change in treadmill VO2peak, submaximal VO2, velocity at VO2peak or accumulated O2 deficit. Velocity at a 4 mmol.l-1 blood lactate rose in R, not in XT or C. Only XT raised arm-crank VO2peak, and likewise arm-crank VO2 at 4 mmol.l-1.

The authors conclude that cross-training that uses non-similar muscles may contribute to better running performance, though to a lesser degree than increased specific training.

Participants

30 well-trained individuals (10 men, 20 women), randomised to R, XT or C. Group sizes and ages are not given in the abstract.

Protocol

8 weeks of enhanced training (+10%/week): extra running (R) or extra swimming (XT) on top of baseline running, compared with continued baseline running (C).
Tests before and after: 3.2 km time trial; treadmill VO2peak, steady-state submaximal VO2 at 2.68 m.s-1, velocity at VO2peak, accumulated O2 deficit, velocity at 4 mmol.l-1 blood lactate; arm-crank VO2peak and VO2 at 4 mmol.l-1.

Key Results

3.2 km time trial: R -26.4s and XT -13.2s both improved; C did not (-5.4s).
Treadmill running, R / XT / C, no significant change: VO2peak (-0.2, -6.0, +2.7%), submaximal VO2 (-1.2, -3.3, +0.2 ml.kg-1.min-1 at 2.68 m.s-1), velocity at VO2peak (+0.05, +0.25, +0.09 m.s-1), accumulated O2 deficit (+11.2, -6.1, +9.4%).
The velocity reached at 4 mmol.l-1 blood lactate went up in R only, not in XT or C (R +0.32, XT +0.07, C +0.08 m.s-1).
In XT, arm-crank VO2peak rose +5%, while arm-crank VO2 at 4 mmol.l-1 rose +6.4%, both significant; R and C showed no significant change (VO2peak +1.3 / -7.7%; VO2 at 4 mmol.l-1 +0.8 / +0.4%).

Practical Application

"Does training in another sport help my running?" In 30 well-trained individuals on baseline running, 8 weeks of extra training done as swimming improved a 3.2 km time trial by 13.2 s, extra running improved it by 26.4 s, and unchanged training did not improve it — cross-training in a non-similar sport helped, about half as much as the same extra load done as running. Extra work in another mode is better than no extra work and worse than specific work.
Use cross-training where more running is not available or not wise (off-season variety, injury risk, a runner at their impact limit); as the race nears, the extra load should be specific (SYNTH-76).
Only added running raised the lactate-threshold velocity (+0.32 m.s-1); swimming raised arm measures. The adaptation followed the muscles trained.
Not from this paper: cycling as the cross-training mode, with a muscularly closer pattern, is in REF-953; transfer between the triathlon disciplines is in REF-785.

Limitations

Whether R and XT differed significantly from each other on the time trial is not stated in the abstract.
Cross-training was added to running, not substituted for it.
8 weeks; group sizes not given in the abstract; swimming only.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

parallel randomised trial with 30 in three groups; between-group significance for the time trial is not stated in the abstract

Content correction 2026-09-27
2026-10-01: independent check against the abstract — wording brought back to what the abstract states.