REF-1736 · Feng, Ying & Jun, 2026 · Frontiers in Sports and Active Living, · Paper · PubMed

Order of strength and endurance work in concurrent training: a narrative (semi-systematic) review of 42 studies (26 human trials, 3 rat studies, 13 reviews) found no consistent order effect on endurance, muscle size or maximal strength in human trials and favoured strength first for explosive power and relative strength (one trial in 30 older men: relative strength +27% vs +15%); its rule of more than 3 h between an endurance session and a following strength session comes from the time course of a muscle signalling molecule, not from a training trial

Main caveat: The authors' own: varied protocols across trials; human and animal results disagree; most trials had fewer than 30 per group and lasted 8-12 weeks.
thin evidenceMeasured on non-athletes to highly trained athletesRead in full

Feng, Ying and Jun reviewed what the order of strength and endurance work within a session, or with only a short gap between, does to adaptation, and tried to explain any order effect through nervous-system and molecular responses.

They searched four databases, including the Chinese CNKI, from 1980 to 2024 and kept 42 records: 26 human trials, 3 rat studies and 13 reviews, of which 6 were systematic reviews or meta-analyses. Synthesis was narrative, with human and animal work kept apart.

In the human trials the order made no consistent difference to gains in VO2max, muscle size or maximal strength, a reading the authors share with two earlier meta-analyses (REF-330, REF-588).

Lifting first went with larger gains in explosive power and relative strength, which the authors attribute to fresher nervous-system output; their main numerical support is one trial in older men (relative strength +27% with strength first vs +15% with endurance first).

Acute signalling responses (mTOR, AMPK, S6, MuRF-1) differed with order, but the human studies were small and inconsistent, and the rat studies showed larger order effects than any human study.

The authors recommend strength first when power or relative strength is the goal, order chosen by goal and person otherwise, and at least 3 h between an endurance session and a following strength session, reasoning that AMPK takes at least 3 h to return to baseline after high-intensity endurance exercise.

Participants

Human trials (26): samples named in the text include young men or young adults (n = 42, described as untrained; n = 36; n = 34), older adults (n = 26 twice, n = 36 men, n = 30 men; the same Cadore trial is given as n = 26 and as n = 30), middle-aged adults (n = 56), physical-education students (n = 48), young soccer players called 'elite-level' (n = 57), young soccer players (n = 36, n = 42), prepubescent boys (n = 128), athletes in a swimming-versus-running inflammation study (n = 18) and resistance-untrained men (n = 12).Most trials had fewer than 30 per group, by the authors.
Animal studies (3): rats; strength work simulated by electrical stimulation.
Reviews (13): 7 narrative, 6 systematic reviews or meta-analyses, including the 43-study pooling by Schumann and colleagues (REF-330), the 15-study review by Lundberg and colleagues, Eddens and colleagues (REF-588) and the 21-study pooling by Wilson and colleagues (REF-512).
Sex split, training status per trial and endurance athletes as such: not reported in the text.

Protocol

Search: PubMed, Web of Science, EBSCO and CNKI; terms for concurrent, simultaneous or combined strength and endurance training combined with sequence or order; January 1980 to December 2024. 2,292 records, 1,395 duplicates removed, 897 screened, 128 reports sought, 88 not retrieved, 40 assessed, 2 added from other sources, 42 included. Some non-core sources from 2025 were added as background references.
Inclusion: peer-reviewed original studies, systematic reviews or meta-analyses; human or animal subjects grouped by training order; protocols described (intensity, volume, duration, gap between modes); outcomes such as maximal strength, explosive power, VO2max, mTOR, AMPK, MuRF-1, EMG or H-reflex.
Quality: two reviewers, a third to settle disputes; Newcastle-Ottawa Scale (maximum 9) for the human trials, a modified version for the animal work, AMSTAR 2 for systematic reviews; narrative reviews not scored.
Synthesis: narrative; consistent findings summarised, conflicting ones discussed by population and protocol; no effect sizes pooled, no heterogeneity statistics, no registration stated.
Funding: National Key Research and Development Program of China (2023YFC3305801); the authors declare no commercial conflicts.

Key Results

QuestionWhat the review reportsWhat it rests on
Order and maximal strength, muscle sizeNo consistent order effect in human trialsSchumann REF-330 (43 studies); trials in older adults (n = 26, 26, 36), students (n = 48), soccer players (n = 57); p > 0.05 or "no significant difference" between orders
Order and explosive powerStrength first favouredCountermovement jump and power in several trials, including 128 prepubescent boys (medicine-ball throw, jumps) and two trials in older adults; no pooled size
Order and relative strengthStrength first: +27% vs +15%One trial, 30 older men (Cadore)
Order and VO2maxNo order effectEddens REF-588 and trials comparing both orders; Hickson 1980 for concurrent vs endurance-only: VO2max +20% vs +17%
Order and anaerobic powerStrength before plyometrics raised Wingate power more than the reverseOne trial, 36 young soccer players
Adults vs adolescentsAdults gained more dynamic strength with strength first; adolescents with aerobic work firstTwo meta-analyses, one of them Eddens REF-588
Acute signallingOrder changed S6, mTOR, TSC2 and MuRF-1 responses, inconsistently between studiesHuman biopsy studies of 8 and 6 young men; rat studies with larger effects
Gap after endurance workAt least 3 h before strength workAMPK takes at least 3 h to return to baseline after hard endurance exercise (one cited source); mTORC1 activity lasts 18 h after strength work
Strength before runningRunning economy fell for up to 8 h after lower-body liftingOne trial, 12 resistance-untrained men
Quality ratings of the included work (authors): human trials scored 6-9 on the Newcastle-Ottawa Scale (median 7.5): 15 (57.69%) high (8 or more), 10 (38.46%) moderate to high (6-7), 1 low; animal studies 7-8; of the 6 systematic reviews, 2 high and 4 moderate by AMSTAR 2. The authors put 33 of 42 records (78.6%) at moderate to high quality and dropped the one low-quality trial from the synthesis of key outcomes.
Modality: the review repeats Wilson's finding REF-512 that running interfered with hypertrophy more than cycling, and one study of 18 athletes in which sprint-interval swimming caused less inflammation than running.

Practical Application

When strength and endurance share a day and power or relative strength is the goal, lift first; for VO2max, maximal strength and muscle size the order did not show a consistent effect in human trials. This is a narrative reading of mostly small 8-12-week trials in untrained people, older adults, students and young soccer players, and the same direction as SYNTH-21 and the 10-study pooling in REF-588 (+6.91% lower-body dynamic strength with strength first)."No consistent effect" means differences were not significant in small trials, not that order is shown to be irrelevant.
The 3-h gap is a molecular rule, not a tested training rule. It comes from AMPK taking at least 3 h to return to baseline after high-intensity endurance work, a signalling marker; no trial in this review compared a 3-h gap with a shorter one on strength or performance. SYNTH-21's ≥3 h rests on a pooled performance outcome instead: the explosive-strength penalty in REF-330 was concentrated in same-session designs and not significant with sessions 3 h or more apart.
Not independent support. Its order conclusions retell the meta-analyses already in the base (REF-330, REF-588, REF-512) together with trials those pool; cite those entries for the numbers.
Inference, not tested here: for an endurance athlete, a hard lift scheduled after a key run or ride on the same day may cost the lift's quality, and a lower-body lift before running may raise the oxygen cost of running for hours (REF-1300, up to 8 h in trained runners); SYNTH-31 sets the spacing.
The review's advice to adolescents (aerobic work first) and to swap running for cycling or swimming rests on subgroup splits and single studies, and SYNTH-21 treats modality as unsettled (REF-512 against REF-330); not settled by this review.

Limitations

The authors' own: varied protocols across trials; human and animal results disagree; most trials had fewer than 30 per group and lasted 8-12 weeks; most mechanistic work is in animals; few studies in elite athletes.
Narrative synthesis: no effect sizes, confidence intervals or heterogeneity; "no consistent association" and "favoured" are the authors' readings of significance across small trials.
Primary trials and the reviews that pool them are counted together in the 42, so the same trials can be weighed more than once.
Quality tool: the Newcastle-Ottawa Scale is built for cohort and case-control studies, not for randomised training trials; the per-study scores sit in a supplement that was not available. The text gives the one low-quality trial a score of 6, which overlaps the "moderate to high" band of 6-7.
Internal inconsistencies: the Cadore trial behind the relative-strength figure is given as n = 26 and as n = 30; the flow diagram shows 88 of 128 sought reports not retrieved, a large loss that the text does not discuss.
The power and relative-strength claims rest on a few trials, much of it in older adults and prepubescent boys, with no pooled estimate; the 3-h interval rests on a signalling marker.
Chinese-language sources (CNKI) were included; their weight in the synthesis is not stated.
How this entry was checked and changed
Why this level

a semi-systematic review with a systematic search but a narrative synthesis that is not a systematic review of controlled trials in the rubric's sense (trials, rat studies and reviews counted together, per-study appraisal in an unavailable supplement); it mixes primary trials with the reviews that already pool them, rated trials with a cohort tool, and its interval rule rests on a signalling marker