The decision. An obese older adult is going to lose weight by eating less. Caloric restriction (CR) removes fat, but it also removes muscle — and in an older adult, who is already losing lean mass with age, that muscle loss is a cost against function and independence. Does adding resistance training (RT) to the diet stop the muscle loss? This is the RESISTANCE-TRAINING lever for lean-mass preservation in a deficit — the exercise counterpart to the PROTEIN lever on Protein Intake During Energy Restriction, for the same decision.
The single source is Sardeli 2018, a gold-tier meta-analysis of 6 RCTs that each randomized obese/overweight
older adults to CR alone vs CR + resistance training (CRRT)
(Sardeli et al., 2018). confidence: low — one MA, 6
trials, surrogate outcomes only, and the population is narrowly obese-ELDERLY (mean >57 y).
The finding: RT preserves lean mass, but does NOT change fat or weight loss
(Sardeli et al., 2018) Adding RT to CR prevents roughly 93.5% of the lean-body-mass loss that dieting alone causes: «the LBM loss in the CRRT group was 93.5% less than the CR group (RMD = 0.819 kg, 95% CI = 0.364 to 1.273, p < 0.001)».
In absolute terms (mean +/- SD change, pre-to-post):
| Compartment | CR + RT (CRRT) | CR alone | Added by RT |
|---|---|---|---|
| Lean body mass | +0.05 +/- 0.3 kg | -0.76 +/- 0.1 kg | +0.82 kg preserved (RMD 0.819 [0.364-1.273]) |
| Fat body mass | -3.86 +/- 1.3 kg | -3.73 +/- 1.2 kg | no significant difference |
| Body mass | -4.16 +/- 1.2 kg | -4.73 +/- 1.2 kg | no significant difference |
The decision-shaping point is the asymmetry: RT changes the COMPOSITION of the weight loss, not its total. Fat loss (~3.8 kg) and total weight loss (~4-4.7 kg) are statistically unchanged by adding RT; all the benefit lands on lean mass, where CR-alone loses ~0.76 kg and CR+RT holds near zero. RT during a diet is a muscle preserver, not a fat-loss or weight-loss amplifier. This is the same structural pattern the Protein Intake During Energy Restriction lever shows (higher protein reshapes the composition of the loss, not its total) and that Exercise Modality for Body Composition in Obesity shows for exercise generally (exercise alone barely moves the scale).
Surrogate boundary — lean mass is not function, and function was not measured
(Sardeli et al., 2018) Lean body mass is a surrogate -> Surrogate Outcomes. The reason an older adult cares about keeping muscle is function, strength, and independence — and the study was not designed to test function. The transmission from the preserved kilogram to a patient-important outcome is a claim to be evidenced, not an assumption, and here it is weak:
- Muscle quality trended up, but non-significantly. «we showed a trend towards the increase in whole body muscle quality when RT was added to CR (p = 0.07), suggesting that in humans, RT improves muscle function regardless of muscle mass changes during CR». p = 0.07 is a tendency, not an established effect.
- The quality metric is itself limited. «muscle group strength was related to whole LBM, instead of local muscle mass, which is a limitation of the method» — so even the strength signal is tied to the surrogate, not to the muscle doing the work.
- One cited single trial reports mobility benefit: «when CRRT was compared to RT alone, the addition of CR to RT improved mobility (400 m of walk time) without compromising other functional adaptations of RT alone [14]» — cited secondary evidence, not pooled here.
So: the lean-mass preservation is well-measured and significant; the step from there to better function is directional and under-powered. Related mortality/function predictors sit on Low Muscle Mass and Mortality, Grip Strength and Mortality and Muscle-Strengthening Activity and Mortality — but those are observational/interventional on different populations, so they do not close this loop.
Mechanism (proposed, directional)
(Sardeli et al., 2018) The authors propose the protection «relies on the energy costs of protein synthesis. During CR without RT, the blunted muscle protein synthesis with elevated proteolysis might allow energy maintenance» — RT reverses this by driving muscle protein synthesis, shifting energy toward lean-mass maintenance while fat is depleted for fuel. (inferred from Sardeli et al., 2018)
Protein is a complementary lever, not a substitute
(Sardeli et al., 2018) The one pooled trial with higher protein (30% vs 15% of energy) «found lower LBM loss than the others (RMD 1.3 kg compared to the main effect of RMD 0.6 kg)», and Longland «CRRT with high protein consumption induces an increase in LBM and promotes larger fat loss» vs low protein. So RT and higher protein appear to stack — single-trial evidence, directional only. The two levers for the same decision:
- RT (this page): preserves ~0.82 kg LBM under CR in obese elderly.
- Protein (-> Protein Intake During Energy Restriction): a modest protein increase (~1.0-1.3 vs ~0.7 g/kg/d) preserves ~0.4-0.8 kg lean mass under a deficit, including in adults >=50 (Kim) — and more protein stacks on top of the RT stimulus.
For an older adult dieting, both levers point the same way; neither replaces the other.
Stratum and transportability — obese ELDERLY, not general obese
(Sardeli et al., 2018) The pool is «older adults or elderly people (mean > 57 years old), including men, women, obese, sedentary, healthy, dyslipidemic, hyperglycemic, and diabetic individuals». The metabolic strata inside are broad (healthy through diabetic), but the AGE band is old and narrow:
- The finding is for the obese/overweight sedentary OLDER adult. This is exactly the stratum where the lever matters most — baseline lean mass is already falling with age, so a dieting-induced ~0.76 kg loss is a larger relative hit, and the ~0.82 kg preserved is a route-(a) high-baseline-risk benefit (the absolute stake is larger in the old).
- Transportability to the middle-aged/younger obese adult is a stated caveat, not an assumption. Younger dieters lose less lean mass per se and respond to RT differently; whether the ~93.5% attenuation transports down the age range is untested in this pool. Do not read the number as a general-obese result.
The sarcopenic-obesity thread (resolved 2026-10-04 — a distinction, not a tension)
(Sardeli et al., 2018) Sardeli contrasts its 93.5% (RT-only) attenuation against a prior MA: «A previous meta-analysis showed only 50% LBM loss attenuation when different types of exercise were added to CR in sarcopenic obese individuals over 50 years old [10]. However, since endurance, resistance, and combined types of training were included in this analysis, it was not possible to identify which type of exercise led to the preventive effect.» Sardeli reads the gap as RT being the active modality — but the prior MA could not isolate modality, and the population differs (sarcopenic obese vs obese).
Resolution (sibling ingest of Resistance Training for Sarcopenic Obesity, Polo-Ferrero 2025). The
sibling does not answer the same question and does not carry the ~50% figure — the AWAITS premise
was a mis-identification. Polo-Ferrero is RT alone vs no treatment, in sarcopenic obesity, without any
caloric restriction (searched: “caloric restriction”/“weight loss”/“energy restriction” = zero in both
chunks), reporting a composite body-composition SMD (0.35), not a lean-mass-retention-under-deficit
fraction. The parameter table (on the sibling page) shows a different population, exposure, comparator, and
outcome-unit — not-joined check (ii): the two claims are consistent once matched, so this is a
distinction, not a [[tension]] and not type-E.
The ~50% referent belongs to Sardeli’s own cited ref [10] (a different mixed-modality exercise+CR MA in
sarcopenic obese), which the wiki does not hold, so the 93.5%-vs-50% comparison stays open against that
un-held source:.
(inferred from Polo-Ferrero et al., 2025; Sardeli et al., 2018)
Decision relevance
- For an obese older adult losing weight, include resistance training to protect muscle. Expect it to preserve roughly a kilogram of lean mass that the diet would otherwise remove (RMD 0.819 kg [0.364-1.273]), at essentially no cost to fat loss. The diet does the fat loss; RT keeps the muscle.
- Do not expect RT to speed weight or fat loss. Fat and total weight loss are unchanged by adding RT here — RT earns its place on composition and (directionally) function, not on the scale. The weight-loss lever is the energy deficit -> Exercise Modality for Body Composition in Obesity, Does Weight Loss Reduce Cardiovascular Events.
- Stack protein on top. RT + a modest protein increase appear complementary for lean-mass preservation -> Protein Intake During Energy Restriction. This is also the exercise arm of the muscle-preservation strategy for drug-induced weight loss -> GLP-1 and Lean Mass.
- Rank within Layer 1. For a sedentary obese older adult, the big rock is the energy deficit (fat loss) plus doing resistance activity at all; RT-during-CR is a high-value co-prescription because the harm it prevents (accelerated sarcopenia during a diet, in a stratum already losing muscle) is patient-important -> Big Rocks (Elderly).
Limits
- Single gold source,
confidence: low— 6 RCTs, one MA; the cross-links are the wiki’s weave, not replication of the pooled estimate. - Surrogate outcomes only. LBM/FBM/BM; no function endpoint was designed-for, no mortality or CV event measured. The muscle-quality transmission is a p = 0.07 trend tied to whole LBM, not local muscle.
- Obese-elderly only (mean >57 y) — transportability to middle-aged/younger obese adults untested.
- Bone mass is a named gap — CR causes bone loss, which exercise prevents «at least in rodent models of male senile osteoporosis [47]»: rodent-only in this pool, not human evidence (Sardeli et al., 2018).
- Heterogeneous CR protocols (one higher-protein, one reduced-fat arm among the six), and whether CR’s benefits depend on protein restriction is unresolved (authors’ own note).
Self-critique [run 2026-10-04, before commit]
- Overclaim guard: the headline (RT preserves ~0.82 kg lean mass, ~93.5% attenuation) is quoted with its CI and p; fat/weight unchanged is stated as the symmetric non-finding; the function step is explicitly demoted to a p = 0.07 trend plus one cited single trial, not asserted as established.
- Surrogate-as-outcome guard: LBM named a surrogate throughout; the transmission to function/strength is marked a claim-to-be-evidenced and shown weak (non-significant, tied to whole LBM).
- Fake-tension guard: the 93.5%-vs-50% contrast with the prior sarcopenic-obese MA is left as an OPEN THREAD with an AWAITS handle to the Polo-Ferrero sibling, NOT filed as a tension — the two answer a related question in different populations/modality mixes, and the cross-source weave belongs to the sibling ingest (per the pair instruction: do not force a tension now).
- Stratum guard: population pinned as obese-ELDERLY (mean >57 y); transportability to younger obese stated as a caveat, not assumed; the high-baseline-risk (route-a) reading of why the lever matters most in the old is marked as the reasoning, not a subgroup-interaction claim.
- Homelessness check: vsearch + title/alias grep found no page holding the RT-during-CR lean-mass finding (closest: GLP-1 and Lean Mass = drug-induced loss; Protein During Energy Restriction = the protein lever; Exercise Modality for Obesity = modality ranking, exercise-alone-not-weight-loss) — minting, not extending.
- Attribution guard: every effect size chunk-pinned via cite.py; magnitudes carry CIs; the mechanism and the Layer-1/route-a readings are marked.