Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42)

Guoying Jiang; Guobing Wang

doi:10.4093/dmj.2025.0939

Articles

Page Path: HOME > Diabetes Metab J > Volume 50(2); 2026 > Article

Letter Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42): Guoying Jiang¹, Guobing Wang²; Diabetes & Metabolism Journal 2026;50(2):412-413.
DOI: https://doi.org/10.4093/dmj.2025.0939
Published online: March 1, 2026

1,182 Views
27 Download

¹The Preventive Treatment of Disease Department, Yibin Traditional Chinese Medicine Hospital, Yibin, China

²Department of Pathology, Yibin Traditional Chinese Medicine Hospital, Yibin, China

Corresponding author: Guobing Wang

Department of Pathology, Yibin Traditional Chinese Medicine Hospital, No. 2, Dawan Road, Xuzhou District, 644000, Yibin, Sichuan, China E-mail: wgb19891112@163.com

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

prev next

Full Article

Download PDF

NOTES
REFERENCES
About this article

See the reply "Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42)" in Volume 50 on page 425.

See the article "Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study" on page 331.

Upon reviewing the study entitled “Adiponectin as a predictor of metabolic dysfunction-associated steatotic liver disease and non-alcoholic fatty liver disease: a 17-year Korean Cohort Study” by Yang et al. [1]. in the Diabetes & Metabolism Journal, we acknowledge the authors’ thorough prospective approach and the extended follow-up period. However, there are several significant methodological concerns that require attention.

First, there are key discrepancies between the flowchart and text: the 2,493 and 5,090 exclusion categories appear to be swapped, and there are 615 cases with pre-existing liver disease that are unaccounted for. More critically, Table 1 in original article reports implausible person-time values. The non-steatotic liver disease group (n=32,708; mean follow-up 14.56 years) should contribute approximately 476,000 person-years, yet subgroup totals (e.g., body mass index >25 kg/m²: 50 million; non-diabetic: 148 million) exceed this by 100- to 300-fold, suggesting unit mislabeling or extraction errors. As survival models rely on accurate exposure time, this flaw undermines all hazard ratio estimates [2].

Second, outcome definitions rely exclusively on the International Classification of Diseases, 10th Revision (ICD-10) code K76.0, with metabolic dysfunction-associated steatotic liver disease (MASLD) identified retrospectively through the incorporation of metabolic criteria. However, the diagnostic accuracy of K76.0 alone is limited (positive predictive value approximately 0.82), showing only slight improvement with the exclusion of certain codes [3]. Furthermore, the categorization of alcohol intake into a binary variable (drinker vs. non-drinker) does not effectively distinguish MASLD from nonalcoholic fatty liver disease (NAFLD) according to established thresholds (≥210 g/week for men) [4]. We propose validating the algorithm against imaging or biomarkers in a subset of the sample and adopting the updated MASLD/NAFLD classification tree to enhance diagnostic specificity.

Third, the study solely assessed baseline adiponectin levels using a proprietary enzyme-linked immunosorbent assay (ELISA; Adipomark, Mesdia Co., Seoul, Korea) without providing information on inter-assay coefficients of variation or cross-platform calibration. Variability between assays for adiponectin is well-documented and can notably influence risk assessments [5]. Additionally, single-timepoint biomarker measurements are susceptible to regression dilution bias, which typically weakens associations towards null findings. Adiponectin levels are likely to have changed significantly over the 17-year period, although this temporal variability was not accounted for. Utilizing joint modeling techniques or reliability-adjusted analyses would offer a more comprehensive understanding of the longitudinal relationship between the biomarker and the outcome [6].

Fourth, the study is vulnerable to immortal time bias despite excluding patients diagnosed with NAFLD within 1 year before baseline [7]. Patients in lower adiponectin groups need to survive until NAFLD diagnosis to be part of the analysis. To address this issue, time-dependent Cox models, landmark analysis at 6 to 12 months, or left-truncation at diagnosis are necessary for aligning observation times. Besides, model 3 includes variables that are likely on the causal pathway from adiponectin to metabolic outcomes, such as triglycerides, fasting glucose, and liver enzymes, which poses a risk of overadjustment bias. Conditioning on these mediators can lead to biased estimates towards the null and obscure the total effect of adiponectin [8]. Moreover, the NAFLD-cardiometabolic subgroup (n=531) with over 16 covariates has an events-per-variable ratio of less than 3, significantly below the recommended range of 10 to 20, which increases the risk of overfitting and unstable estimates [9]. It is recommended to present minimally adjusted, confounder-adjusted (excluding mediators), and mediation-aware analyses to elucidate the underlying mechanisms.

Clarification is needed regarding the handling of competing risks. While cause-specific Cox models are mentioned, the connection to the reported cumulative incidence functions remains unclear. To predict absolute risk in the presence of death as a competing event, Fine-Gray subdistribution models or Aalen-Johansen estimation is typically necessary [2]. It is important to report tests of proportional hazards for adiponectin itself, along with the use of appropriate competing risk estimators.

In conclusion, Yang et al.’s longitudinal study [1] offers valuable data on adiponectin and metabolic liver disease. However, the study’s multiple methodological issues, including errors in person-time calculations, insufficient diagnostic validation, and analytical limitations, must be promptly resolved before these findings can inform clinical practice. Rigorous reanalysis and validation are necessary to enhance the evidence supporting adiponectin as a biomarker for assessing the risk of metabolic liver disease.

NOTES

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

REFERENCES

1. Yang YS, Zhang HS, Kimm H, Jung KJ, Kim S, Baek JW, et al. Adiponectin as a predictor of metabolic dysfunction-associated steatotic liver disease and non-alcoholic fatty liver disease: a 17-year Korean Cohort Study. Diabetes Metab J 2026;50:331-42.
2. Putter H, Fiocco M, Geskus RB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med 2007;26:2389-430.Article PubMed
3. Astrom H, Wester A, Hagstrom H. Administrative coding for non-alcoholic fatty liver disease is accurate in Swedish patients. Scand J Gastroenterol 2023;58:931-6.Article PubMed
4. Rinella ME, Lazarus JV, Ratziu V, Francque SM, Sanyal AJ, Kanwal F, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. J Hepatol 2023;79:1542-56.PubMed
5. Mavilia MG, Wu GY. Liver and serum adiponectin levels in non-alcoholic fatty liver disease. J Dig Dis 2021;22:214-21.Article PubMed PDF
6. Hutcheon JA, Chiolero A, Hanley JA. Random measurement error and regression dilution bias. BMJ 2010;340:c2289.Article PubMed
7. Suissa S. Immortal time bias in pharmaco-epidemiology. Am J Epidemiol 2008;167:492-9.Article PubMed
8. Schisterman EF, Cole SR, Platt RW. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology 2009;20:488-95.Article PubMed PMC
9. Ogundimu EO, Altman DG, Collins GS. Adequate sample size for developing prediction models is not simply related to events per variable. J Clin Epidemiol 2016;76:175-82.Article PubMed PMC

Figure & Data

References

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite this Article

Cite this Article: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42)

Diabetes Metab J. 2026;50(2):412-413. Published online March 1, 2026

DOI: https://doi.org/10.4093/dmj.2025.0939

XML Download

Related articles

Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42)

About this article

Jiang G, Wang G. Adiponectin as a Predictor of Metabolic Dysfunction-Associated Steatotic Liver Disease and Non-Alcoholic Fatty Liver Disease: A 17-Year Korean Cohort Study (Diabetes Metab J 2026;50:331-42). Diabetes Metab J. 2026;50(2):412-413.

DOI: https://doi.org/10.4093/dmj.2025.0939.

Download citation ↓