CoChat
Sign Up
Cited by 18Nutrition and Health in AgingBody Composition Measurement TechniquesDiabetes, Cardiovascular Risks, and Lipoproteins

A diagnostic proposal for sarcopenic obesity in adults based on body composition phenotypes

Open access
Zambon Azevedo, Vittoria, Ponnaiah, Maharajah, Bel Lassen, Pierre, Ratziu, Vlad, Oppert, Jean-Michel·2022·Clinical Nutrition ESPEN
DOI 10.1016/j.clnesp.2022.10.010
Open this paper in CoChat — free account

Abstract

BACKGROUND & AIMS Sarcopenic obesity (SO) associates a decrease in lean body mass (LBM) with an excessive increase in fat mass (FM). A number of diagnostic methods, definitions criteria, and thresholds have been proposed for SO resulting in markedly discordant prevalence estimates in populations with obesity. In this study, we first assessed several previously described SO diagnostic criteria and their limitations, and then we propose an innovative approach for identifying SO. METHODS Data were from a cross-sectional study of a cohort of overweight/obese patients who underwent clinical, laboratory, and body composition assessments by dual-energy X-ray absorptiometry (DXA). We performed unsupervised machine learning through clustering analysis to discriminate lean and fat compartments, and multivariate logistic regressions which provided prognostic variables applied on sex-specific models for SO diagnosis evaluation based on a training dataset (80% of total sample, n = 1165). The predicted models were validated by random forest (RF) machine learning algorithm in the validation dataset (20% of total sample, n = 262). RESULTS Data from 1427 subjects were analyzed, 79.8% women, mean (±s.d.) age 45.0 (±12.9) years, grade III obesity (BMI over 40 kg/m2) in 42.7%, diabetes in 20.7%, dyslipidemia in 86.3%, and arterial hypertension in 30.3%. Patients with grade III obesity had higher amounts of LBM, FM, and bone mass than subjects with overweight (BMI between 25.0 and 29.9 kg/m2) (p-values < 0.001). When published definitions of SO were applied to this cohort, the prevalence ranged from 0.6% to 96.6%. We built a model that identified 62 (4.3%) individuals as SO, 1125 (78.9%) as non-SO, and 240 (16.8%) as borderline-SO. SO patients showed higher body weight, FM, bone mass, leptin levels, and hepatic steatosis index, but lower LBM and all muscle indexes than non-SO subjects (p-values ≤ 0.001). Patients in the SO and borderline-SO categories were more often females than males (4.5% vs. 3.8% and 16.9% vs. 16.7% respectively, p-value < 0.001) and had significantly higher prevalence of metabolic syndrome and hypertension than non-SO subjects. Males with SO also had higher cardiovascular risk score, while females had higher prevalence of respiratory disorders (p-values < 0.05 for all). CONCLUSIONS Current diagnostic criteria for SO result in widely discrepant prevalence values leading to diagnosis uncertainty. We developed and validated diagnostic criteria based on body composition phenotypes, specifically for overweight/obese subjects, which identified patients at risk of cardio-metabolic complications. This approach may improve the identification of sarcopenia in subjects with obesity.

Integrity signals

Registry

Crossref · 10.1016/j.clnesp.2022.10.010

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The limitations of these cut-off values in detecting persons with both low muscle function, low muscle mass and high fat mass might overlook a risk of SO.

    “As a result, the limitations of these cut‐off values in detecting persons with both low muscle function, low muscle mass and high fat mass might overlook a risk of SO [17].”

    From Sarcopenic Obesity Phenotype Index (SOP <i>i</i> ): A Population‐Based Study (2025) · cites this paper as [17]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Zambon Azevedo, V., Ponnaiah, M., Bel Lassen, P., Ratziu, V., & Oppert, J.-M.
Affiliations
  • Fondation pour l’innovation en Cadiométabolisme et Nutrition (FR)
  • Inserm (FR)
  • Sorbonne Université (FR)
ORCID
Venue
Clinical Nutrition ESPEN
Publisher
Elsevier BV
Published
2022-12
Volume
52
Pages
119-130
ISSN
2405-4577 (ISSN-L)
DOI
10.1016/j.clnesp.2022.10.010
OpenAlex
W4306978133
Semantic Scholar
16eda051b81812ebf1003440815928a9dcc2f222
Cited by
18
References
38
Citations by year
2026: 1 · 2025: 13 · 2024: 2 · 2023: 2
FWCI
2.65
Type
journal article
Language
en
Open access
Licence
Locations
  • Clinical Nutrition ESPEN (acceptedVersion)
  • PubMed (publishedVersion)
Topics
  • Health Sciences › Medicine › Physiology › Nutrition and Health in Aging
  • Health Sciences › Medicine › Physiology › Body Composition Measurement Techniques
  • Health Sciences › Medicine › Endocrinology, Diabetes and Metabolism › Diabetes, Cardiovascular Risks, and Lipoproteins
Keywords
sarcopenic obesity, diagnostic criteria, body composition phenotypes, lean body mass, fat mass, metabolic syndrome, cardiovascular risk, hepatic steatosis, leptin, respiratory diseases, overweight and obesity, prevalence estimates, dual-energy X-ray absorptiometry, unsupervised clustering
MeSH terms
Body Composition, Cross-Sectional Studies, Female, Humans, Hypertension, Male, Obesity, Phenotype, Body Mass Index, Overweight, Sarcopenia

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review