Sarcopenia and metabolomics: mind the gap between biomarker discovery and clinical application

To the Editor,

We were pleased to read the article by Kim et al. in The Journal of Nutrition, Health and Aging, exploring the relationship between plasma metabolites and incident sarcopenia in a large prospective cohort from the UK Biobank [1]. The identification of 38 metabolites associated with sarcopenia and the observed improvement in predictive models by adding these metabolites represent a meaningful contribution to the field. The study advances our understanding of the metabolic pathways that may underlie sarcopenia and highlights potential avenues for early risk assessment.That said, several methodological and clinical aspects warrant closer scrutiny to better contextualize the findings.

First, the diagnostic approach for sarcopenia combined ICD-10 codes with anthropometric thresholds, which may not adequately reflect the functional decline central to the condition. Current consensus definitions, such as those from EWGSOP2, stress the importance of measuring physical performance—for instance, gait speed—alongside muscle strength and mass. The lack of functional metrics in the UK Biobank could lead to incomplete case identification, potentially attenuating the true association between metabolites and sarcopenia [2].

Second, the metabolomic profile available in the UK Biobank is skewed toward lipids and lipoprotein subclasses, with limited data on metabolites closely tied to mitochondrial function and muscle-specific metabolism. Key intermediates such as acylcarnitines or TCA cycle markers—which are implicated in muscle energy production and aging—are underrepresented. This narrow focus may obscure important metabolic disturbances involved in sarcopenia [3].

Furthermore, although the authors adjusted for a range of covariates, residual confounding cannot be ruled out. Lifestyle factors such as detailed dietary habits—especially protein distribution across meals—and structured resistance exercise were not fully captured. Self-reported dietary intake from a single 24-h recall is susceptible to bias and may not represent long-term nutritional behavior, which is more relevant to progressive muscle loss.

Conclusively, the clinical utility of the metabolite panel remains uncertain. The absolute gains in predictive accuracy—though statistically significant—were modest. Moving the AUC from 0.891 to 0.898 may not justify incorporating metabolomic screening into routine practice, particularly given the associated costs and operational complexity. Moreover, the absence of external validation in an independent cohort limits confidence in the generalizability of the model.

Looking ahead, future studies would benefit from integrating functional performance measures and repeated metabolomic assessments to better capture dynamic changes and causal pathways. From a public health standpoint, combining metabolomic data with real-world mobility metrics from wearables could support more personalized and timely interventions.

In summary, Kim et al. have provided a valuable resource for understanding the metabolic dimensions of sarcopenia. Their findings underscore the potential of metabolomics to reveal novel biology and refine risk prediction. However, further validation and a stronger emphasis on functional outcomes are necessary before these insights can be translated into clinical practice. We commend the authors on their work and look forward to future research in this evolving area.

Statement

There are no tables or figures from another resource in this manuscript. This work is not under consideration for publication elsewhere, and I have read, understood and followed the instruction for authors while preparing my manuscript.

Funding

There is no funding, equipment, medication, or support available in any form.

Declaration of competing interest

The authors have no conflict of interest to disclose.