REF-761 · Martínez-Navarro et al., 2021 · European Journal of Sport Science, Vol. · Paper · PubMed

The week after running a marathon: Effects of running vs elliptical training vs resting on neuromuscular performance and muscle damage recovery

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

Sixty-four road-marathon finishers were randomised at the finish line to easy running, easy elliptical training or rest for the following week. The two exercise groups did 40 min at 95-105% of the heart rate at the first ventilatory threshold at 48, 96 and 144 h after the race.

Creatine kinase and lactate dehydrogenase recovered no differently between groups, and the running group's squat jump was better than the rest group's at 96 h. The authors conclude that returning to running at 48 h does not seem to harm muscle-damage recovery up to eight days after the race and could speed neuromuscular recovery.

Participants

n=64 finishers of a road marathon (abstract: "Sixty-four finishers"): 54 men and 10 women, 39 ± 4 years, finish time 3 h 35 min ± 21 min. Group sizes are not given in the abstract.

Protocol

Randomised on arrival at the finish line to RUN (running), ELIP (elliptical training) or REST (resting recovery).
RUN and ELIP: 40 min continuous at moderate intensity, 95-105% of the HR at the first ventilatory threshold, at 48, 96 and 144 h after the marathon.
Blood lactate dehydrogenase (LDH) and creatine kinase (CK): the day before the race, within 15 min of finishing, and at 24, 48, 96, 144 and 192 h post-race.
Squat jump (SJ): before and after the marathon and at 48, 96 and 144 h post-race.

Key Results

Muscle-damage blood markers (CK, LDH): no 'Intervention' effect and no 'Intervention x Time' interaction (p > 0.05). Running or elliptical training from 48 h did not change how these markers recovered compared with rest, through 192 h.
Squat jump at 96 h: RUN 108.29 ± 10.64% vs REST 100.58 ± 9.16% (p = 0.020, d = 0.80). This is the only between-group difference the abstract reports.
The abstract reports no squat-jump comparison for ELIP, and no RUN vs ELIP comparison.
The abstract does not give the marker values themselves or the day on which each returned to baseline.

Practical Application

In sixty-four recreational marathon finishers randomised to running, elliptical or rest for the week after the race, three easy 40-min sessions at 48, 96 and 144 h (95-105% of the HR at the first ventilatory threshold) did not slow the recovery of CK or LDH up to 192 h, and the runners jumped better than the resters at 96 h (108.29% vs 100.58% of pre-race squat jump, p = 0.020, d = 0.80). You can run again from 48 h after a marathon, easily and for about 40 min; on this evidence it does not delay muscle-damage recovery and may bring your legs back sooner than full rest.
The elliptical was no worse than rest on the blood markers, but the abstract reports the neuromuscular advantage only for running — it is a fallback if running hurts, not a proven equal.
The intensity is the condition. Inference, not tested here: 95-105% of the HR at the first ventilatory threshold is an easy aerobic run; an athlete without a measured threshold should run conversationally, and "back running at 48 h" is not "back training at 48 h".
The abstract does not say when CK and LDH returned to baseline or when running performance recovered, so it cannot give a date for the first hard session or race.
Not from this paper; the base's own practice: keep the post-marathon week easy and short, and delay intervals and long runs until jumps, stairs and easy-run HR feel normal; the biomarker time course after a marathon is in REF-762.

Limitations

Group sizes are not given; the sample was mostly men (54 men, 10 women).
The exercise dose was fixed (40 min, moderate, three sessions); longer, harder or earlier (24 h) sessions were not tested.
Neuromuscular recovery was measured by squat jump only, and the difference appeared at one time point (96 h); running performance was not measured.
Blinding is not described and cannot be complete for an exercise intervention.
Written from the abstract and record metadata only (full text not read); marker values and return-to-baseline times are not available here.
How this entry was checked and changed
Why this level

three-arm parallel RCT, n=64 in total but split three ways (group sizes not given in the abstract), so no pairwise comparison reaches the rubric's well-powered bar; the one positive result is a single time point