Will adding cardio cost me muscle after 50?
In 12 RCTs of non-athletes 50+, lean mass did not differ significantly from lifting alone (−0.10 kg, CI −0.61 to 0.41); +0.52 kg under 24 weeks, −0.74 kg at 24+. Trained people excluded.
Main caveat: Population. Mostly non-athletes with a chronic condition or overweight (our tally of Table 1: 12 of 53 trials, 411 participants, labelled simply healthy).
Khalafi and colleagues pooled randomised trials comparing concurrent training with aerobic or resistance training alone on body composition in adults aged 50 and over, searched to March 2024; 53 trials and 2873 participants qualified.
Against aerobic training, concurrent training raised body weight (0.31 kg) and lean body mass (0.51 kg) more; fat measures, muscle mass and fibre or muscle cross-sectional area did not differ significantly. Against resistance training, nothing differed significantly overall.
The authors conclude concurrent training is as effective as aerobic training for reducing fat and as effective as resistance training for building muscle. Our reading: the lean-mass gain over aerobic alone is what you would expect from adding lifting, and it is small.
"As effective as" claims more than a non-significant difference with a confidence interval of about ±0.5 kg can show, and the subgroups run in opposite directions: lean mass higher with concurrent training in trials under 24 weeks and lower in longer ones. No equivalence margin was tested. These are mostly clinical and overweight non-athletes.
Protocol
Design: systematic review and meta-analysis; PRISMA and Cochrane Handbook; PROSPERO CRD42024511774.
Search: PubMed, Web of Science, Scopus, inception to March 2024, plus Google Scholar and reference lists; English language.
Who: n = 2873 in 53 trials; mean ages 51 to 82; BMI 23 to 35; trial sizes 15 to 208. Any health status. Trained or athletic older adults excluded.
Interventions: concurrent training (aerobic plus resistance, same or separate sessions; balance or flexibility allowed; HIIT excluded) vs aerobic or resistance training alone; 6 weeks to 12 months, 2 to 5 sessions a week (3 in most). Trials with only a non-exercise control were excluded.
Measures: fat mass, body fat percentage, waist, visceral fat, lean body mass (fat-free mass where lean was not given), muscle mass or volume, muscle or fibre cross-sectional area, body weight, BMI. DXA, MRI, CT, bioimpedance and skinfolds all accepted, with sensitivity analyses dropping the last two.
Quality: PEDro with two items removed (range 4 to 9 of 9). No GRADE or other certainty rating.
Subgroups: age, sex, BMI, duration (under 24 vs at least 24 weeks), same vs separate sessions, where more than 10 interventions were available.
Funding and interests: no external funding; no conflicts declared.
Key Results
| Outcome | Concurrent vs aerobic: trials, effect (95% CI), p | Concurrent vs resistance: trials, effect (95% CI), p |
|---|---|---|
| Lean body mass, kg | 27, WMD 0.51 (0.29 to 0.73), 0.001 | 12, WMD −0.10 (−0.61 to 0.41), 0.69 |
| Fat mass, kg | 22, WMD −0.01 (−0.35 to 0.31), 0.91 | 13, WMD −0.56 (−1.23 to 0.10), 0.09 |
| Body fat, % | 33, WMD −0.30 (−0.69 to 0.08), 0.13 | 21, WMD −0.25 (−0.74 to 0.23), 0.36 |
| Waist, cm | 20, WMD −0.33 (−0.90 to 0.23), 0.24 | 15, WMD 0.09 (−0.82 to 1.01), 0.83 |
| Visceral fat | 5, SMD −0.02 (−0.27 to 0.22), 0.84 | not analysed |
| Muscle mass or volume | 10, SMD 0.22 (−0.01 to 0.45), 0.06 | 7, SMD −0.09 (−0.38 to 0.19), 0.52 |
| Cross-sectional area | 7, SMD 0.21 (−0.02 to 0.44), 0.07 | 5, SMD −0.15 (−0.40 to 0.10), 0.25 |
| Body weight, kg | 31, WMD 0.31 (0.09 to 0.54), 0.006 | 23, WMD −0.39 (−0.96 to 0.16), 0.16 |
Matched exercise volume only (sensitivity): against aerobic, lean mass WMD 0.45 kg (0.14 to 0.77, p = 0.004; 10 trials) and body fat −0.54% (−0.85 to −0.23, p = 0.001; 12 trials). Against resistance, lean mass 0.08 kg (−0.78 to 0.95, p = 0.84; 4 trials).
Reliable measures only (sensitivity, no bioimpedance or skinfolds): against aerobic, lean mass 0.43 kg (0.19 to 0.67, p = 0.001; 15 trials), fat mass −0.47 kg (p = 0.03; 11 trials). Against resistance, lean mass −0.23 kg (−0.83 to 0.36, p = 0.44; 12 trials).
Duration against resistance alone, the subgroup that matters: lean mass +0.52 kg under 24 weeks (p = 0.002) and −0.74 kg at 24 weeks or more (p = 0.001); older adults +0.49 kg (p = 0.01). Fat mass −1.75 kg in older adults (p = 0.001). The text prints no intervals for these subgroups, and the abstract does not report them.
Same vs separate sessions against aerobic alone: lean mass +0.69 kg separate (p = 0.001) and +0.38 kg same (p = 0.02). No between-subgroup test is printed in the text read.
Heterogeneity: against resistance, lean mass I² 65.34, body fat I² 64.90, fat mass I² 49.22; against aerobic, low except body fat (38.91).
Mechanisms
Measured: body composition only. No signalling, no fibre-type work in the pooled outcomes, no performance outcome.
Authors' interpretation: adding aerobic to resistance training does not appear to cause an interference effect on muscle in middle-aged and older adults; larger volumes and timing may affect the result through fatigue.
Inference: the extra 0.51 kg over aerobic alone is largely the lifting. And the sign change with duration could reflect protocol, health status or volume differences rather than time itself; subgroups are not randomised and are likely confounded.
Practical Application
Will adding cardio cost me muscle after 50? In 12 RCTs of non-athletes 50+, lean mass did not differ significantly from lifting alone (−0.10 kg, CI −0.61 to 0.41); +0.52 kg under 24 weeks, −0.74 kg at 24+. Trained people excluded.
Adding lifting to aerobic work: about half a kilo more lean mass (0.51 kg, 0.43 to 0.45 in the stricter analyses), with no significant difference in fat mass overall. Direct for untrained adults 50+; for a masters endurance athlete, a transfer.
Do not say "as effective as": no equivalence test was run, so a non-significant difference is what can be stated; the interval allows up to about 0.6 kg less lean mass than with lifting alone, and the longer trials point that way.
Not a fat-loss claim against lifting: fat mass −0.56 kg (CI −1.23 to 0.10), not significant overall; significant only in the older-adult subgroup (−1.75 kg), and most trials were in overweight or clinical groups.
Limitations
Population. Mostly non-athletes with a chronic condition or overweight (our tally of Table 1: 12 of 53 trials, 411 participants, labelled simply healthy); trained and athletic people excluded by design.
A null read as equivalence. The conclusion's "as effective as" has no equivalence margin behind it.
Unmatched volume and timing in most trials, by the authors' own statement; concurrent arms often simply do more. Matched-volume subsets are small (4 trials for lean mass against resistance).
Mixed measures. DXA, MRI, CT, bioimpedance and skinfolds pooled; fat-free mass substituted for lean mass where needed. Sensitivity analyses help on some outcomes; none are stable across all of them.
Subgroups. Unadjusted, across trials; opposite signs by duration are unexplained. No certainty grading.
Publication bias. Funnel plots suggested it for several outcomes while Egger's tests mostly did not; the trim-and-fill estimate moved BMI against resistance to −0.40 and muscle mass against aerobic to SMD 0.28 (0.07 to 0.49).
Reporting. The sensitivity CI for fat mass (−0.90 to 0.04) crosses zero with p = 0.03, and the one for body fat (−1.31 to −0.24) excludes zero with p = 0.17, so one of each pair is misprinted. The discussion calls concurrent training superior to aerobic for muscle mass while muscle mass itself had p = 0.06. The introduction's muscle loss of "3–8% per annum" after age 30 is not a figure to carry; it is not this review's result.
Interests. No external funding; no conflicts declared.
How this entry was checked and changed
Why this level
read from the full text (open access, CC BY 4.0). An SR/MA of controlled trials is A by design; held at A- because many trials are small, heterogeneity reaches 65% for the concurrent-vs-resistance outcomes, volume and timing were unmatched in most trials, the measures include skinfolds and bioimpedance, and no certainty grading was done
Full text (checked 2026-10-31)
53 articles with 2873 middle-aged and older adults; mean ages from 51 to 82 years; trial sizes from 15 to 208 (full text)
Studies of trained or athletic older adults and HIIT-based concurrent protocols were excluded (full text)
Against aerobic training: lean body mass WMD 0.51 kg (95% CI 0.29 to 0.73), p = 0.001, 27 trials (full text)
Against resistance training: lean body mass WMD −0.10 kg (95% CI −0.61 to 0.41), p = 0.69, 12 trials; fat mass WMD −0.56 kg (95% CI −1.23 to 0.10), p = 0.09, 13 trials (full text)
Lean mass against resistance training by duration: WMD 0.52 kg, p = 0.002 under 24 weeks; WMD −0.74 kg, p = 0.001 at 24 weeks or more (full text)
Exercise volume and timing were not matched between arms in the majority of trials (full text)
No external funding; no conflicts of interest (full text)