Non-linear association between metabolic score for insulin resistance and diabetes mellitus in normal-weight middle-aged and older Chinese adults: a multicenter retrospective cohort study

Background: Although the link between the metabolic score for insulin resistance (METS-IR) and diabetes mellitus (DM) is well-established in general populations, it remains underexplored in normal-weight middle-aged and old adults—an often-overlooked group in metabolic research. Therefore, this study investigated the association between METS-IR and diabetes in this specific population.

Methods: A retrospective cohort study was conducted using data from the Rich Healthcare Group database, including 23,692 normal-weight (body mass index: 18.5–23.9 kg/m²) Chinese adults aged ≥45 years. Cox proportional hazards models were applied with adjustments for demographic, clinical, and biochemical confounders. Non-linearity was examined using smooth curve fitting and threshold effect analysis.

Results: Higher METS-IR values were significantly correlated with greater diabetes risk during follow-up. In the model 3, each 1-unit rise in METS-IR corresponded to a 12% increase in diabetes risk (HR: 1.12, 95% CI: 1.10, 1.14). The analysis revealed a non-linear association, with a critical inflection point at a METS-IR of 37.24. Below this threshold, risk rose more sharply (HR: 1.18, 95% CI: 1.14, 1.21 per unit), whereas above the threshold, the association was attenuated (HR: 0.96, 95% CI: 0.89, 1.04). Subgroup analyses showed largely consistent findings across strata.

Conclusions: Among normal-weight middle-aged and older Chinese adults, METS-IR exhibits a positive, non-linear relationship with DM risk. Maintaining a lower METS-IR may help prevent the development of DM.

Introduction

Diabetes mellitus (DM) poses a substantial and escalating global health burden, with its prevalence increasing rapidly, especially among aging populations (1). According to 2021 estimates by the International Diabetes Federation, approximately 537 million adults (20–79 years) had diabetes, a number expected to rise to 783 million by 2045. This steep increase underscores the critical need for effective public health strategies (2, 3). Such a progressive increase in diabetes cases also has major socioeconomic consequences, including rising healthcare costs and loss of productivity, marking DM as one of the most pressing health issues of the 21st century (4). A key pathophysiological feature preceding type 2 DM (T2DM) is insulin resistance (IR), which typically presents as diminished cellular responsiveness to insulin, due to chronic inflammation, excess adiposity, and metabolic dysfunction (5–7). Although obesity, particularly visceral fat accumulation, is a well-documented risk factor contributing to IR and subsequently the development of T2DM, it is essential to note that a significant proportion of individuals with diabetes maintain a normal body weight, particularly among middle-aged and older adults (8, 9). T2DM and non-alcoholic fatty liver disease (NAFLD) are closely linked in lean individuals (10). On the other hand, normal-weight individuals, including middle-aged and elderly adults, can develop NAFLD, and it may even present with more severe liver disease and higher mortality risks compared to obese individuals. This condition, known as lean NAFLD, is often overlooked because it lacks the typical signs of obesity (11). This pattern reinforces the notion that relying solely on body mass index (BMI) for metabolic risk assessment is inadequate.

Calculated based on BMI high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), and fasting plasma glucose (FPG), the metabolic score for IR (METS-IR) serves as a useful non-insulin-based surrogate for assessing IR (12, 13). This composite index provides a cost-effective and accessible tool, especially beneficial in large-scale epidemiological studies where comprehensive insulin assays may be unfeasible (14). METS-IR is known to correlate significantly with several pathological components of IR, including visceral and intrahepatic fat content (15), thus highlighting its utility in identifying at-risk individuals even before diabetes becomes clinically evident (16). Notably, METS-IR has been associated with an increased risk of diabetes in a wide range of populations, including those with varying body weights (17). However, greater attention must be directed toward the often-overlooked normal-weight demographic, as evidence suggests that these individuals still harbor metabolic abnormalities (18, 19). Normal-weight middle-aged and elderly adults can still have increased carotid intima-media thickness (CIMT) if they have other risk factors like central obesity, high blood pressure, or high blood sugar (20), importantly, CIMT and NAFLD remain closely associated, contributing substantially to cardiovascular disease risk (21).

This is clinically significant because normal-weight adults, particularly those in middle-age or older, often show latent metabolic dysfunction despite having a BMI in the normal range. Age-related changes, including sarcopenia and visceral fat accumulation, predispose these adults to IR without overt obesity (22, 23). However, there is a paucity of data on the link between METS-IR and diabetes risk in normal-weight middle-aged and older adults. Therefore, we leveraged a large retrospective cohort from China's Rich Healthcare Group to examine the link between METS-IR and DM in a cohort focused specifically on normal-weight adults aged ≥45 years. Our findings are expected to inform early preventive strategies tailored to normal-weight individuals who may be overlooked in routine diabetes screening, and could also improve understanding of their broader metabolic health, including NAFLD and cardiovascular risk.

Methods

Study design and population

This secondary analysis used data from a publicly available cohort of 211,833 Chinese adults originally compiled by Chen et al. (24), and obtained from the Dryad Digital Repository (https://datadryad.org/stash/dataset/ doi: 10.5061/dryad.ft8750v). The dataset includes detailed medical records derived from health examination programs. As an open-access repository, Dryad permits the free download and reuse of data for secondary analyses in accordance with its terms of service. The original study received ethical approval from the Rich Healthcare Group Review Board, and thus no additional review was required for this secondary analysis. Both the original investigation and the present analysis adhered to the ethical principles of the Declaration of Helsinki.

The baseline population was drawn from 685,277 Chinese adults (≥20 years) who completed health screenings between 2010 and 2016. After applying standard exclusions for missing data, extreme values, baseline diabetes, and insufficient follow-up, the final dataset contained 211,833 individuals. For the current analysis of normal-weight middle-aged and elderly adults, we further excluded participants aged < 45 years, those with missing triglyceride or HDL-C measurements, extreme HDL-C values (HDL-C = 0 mmol/L), and BMI outside the normal-weight range (< 18.5 or ≥24 kg/m²), resulting in a final analytical sample of 23,692 participants (Figure 1).

Figure 1

Figure 1. Participants selection flowchart. BMI, body mass index; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol.

METS-IR evaluation

The METS-IR value for each participant was calculated based on their clinical characteristics using the following formula: METS-IR = Ln [(2 × FPG (mmol/L) × 18) + TG (mmol/L) × 88.5] × BMI (kg/m²)/Ln [HDL-C (mmol/L) × 38.7] (12).

DM assessment

DM was defined as FPG ≥ 7.00 mmol/L and/or a self-reported physician diagnosis. Follow-up for each participant was terminated upon diabetes diagnosis or at their last visit, whichever occurred first (24, 25).

Data collection

As described previously (24), trained research staff, who completed standardized protocol training before data collection, performed all procedures to ensure accuracy and consistency. Data were collected on demographic characteristics, including smoking status, age, family history of diabetes, gender, and drinking status. Physical examinations included standardized measurements of height, weight, and blood pressure. After at least a 10-h fast, venous blood samples were obtained from all participants. Biochemical parameters—total low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC), FPG, TG, serum creatinine (Scr), alanine aminotransferase (ALT), HDL-C, and blood urea nitrogen (BUN)—were analyzed using an automated biochemistry analyzer (Beckman 5800).

Statistical analysis

In this study, baseline characteristics of the study population were first described in detail. Continuous variables were presented as means ± standard deviation, while categorical variables were summarized as frequencies and percentages. Survival rates and cumulative event incidence were estimated using the Kaplan–Meier method.

Cox proportional hazards regression was performed to quantify the relationship between METS-IR and incident diabetes, generating hazard ratios (HRs) and 95% confidence intervals (CIs). Three sequentially adjusted models were constructed: Model 1 remained unadjusted; Model 2 was adjusted for diastolic blood pressure (DBP), age, systolic blood pressure (SBP), and gender; and Model 3 was further adjusted for LDL-C, family history of diabetes, BUN, TC, Scr, and ALT. To validate the findings from the continuous METS-IR variable and assess trends, METS-IR was categorized into quartiles, and P-value for trend was calculated.

To examine potential non-linearity in the METS-IR-DM relationship, a Cox proportional hazards model with smooth curve fitting (cubic splines) was applied for graphical assessment. When non-linearity was detected, the inflection point was identified using a recursive algorithm, followed by construction of a piecewise Cox model. A likelihood ratio test was then applied to compare the linear and non-linear models and to determine optimal fit (26).

To assess the robustness of the primary results, stratified and sensitivity analyses were performed. Stratified analyses were conducted across subgroups defined by gender, age, and family history of diabetes. Subgroup analyses by smoking and drinking status were not performed due to insufficient data resulting from a high degree of missing values. In the primary regression models, missing data were handled using complete-case analysis. The robustness of the findings was evaluated through two sensitivity analyses: first, by including variables with high missing rates (smoking and drinking status) without imputation; and second, by assigning dummy variables to all covariates with more than 5% missing values, including smoking and drinking status. Statistical analyses were performed using R and EmpowerStats, with significance defined as P < 0.05.

Results

Characteristics of study participants

Baseline characteristics of the 23,692 enrolled normal-weight middle-aged and older participants, stratified by METS-IR quartiles (Q1–Q4, n = 5,923 each), are shown in Table 1. With rising METS-IR quartile, participants tended to be older and displayed progressive increases in SBP, DBP, ALT, Scr, FPG, TG, and BMI, whereas HDL-C levels decreased. The proportions of current smokers, males, and current drinkers also rose across higher METS-IR quartiles.

Table 1

Table 1. The baseline characteristics of participants.

Tendency of DM across METS-IR quartiles in normal-weight middle-aged and older Chinese adults

Kaplan–Meier survival curves showing cumulative diabetes-free survival across METS-IR quartiles are presented in Figure 2. The curves are clearly separated, indicating a significant association between higher METS-IR and reduced diabetes-free survival (log-rank test, p < 0.0001). Participants in the highest quartile (Q4) exhibited the lowest survival probability, whereas those in the lowest quartile (Q1) maintained the highest probability throughout the follow-up period. The sample counts at each follow-up interval, indicated below the figure, clearly show a steady attrition of participants over time, which aligns with the Kaplan-Meier survival curve.

Figure 2

Figure 2. Kaplan–Meier event-free survival curve. Kaplan-Meier analysis of incident diabetes mellitus based on METS-IR quartiles in normal-weight middle-aged and older Chinese adults (log-rank, P < 0.0001).

Relationship of METS-IR with DM in normal-weight middle-aged and older Chinese adults

A significant positive association was observed between METS-IR and DM in normal-weight middle-aged and older Chinese adults (Table 2). In the Model 3, each 1-unit increase in METS-IR corresponded to a 12% higher diabetes risk (HR 1.12, 95% CI 1.10, 1.14, P < 0.0001). Quartile analyses showed a clear dose-response relationship (P for trend < 0.0001). Participants in the highest quartile (Q4) had a substantially elevated risk (HR 3.34, 95% CI 2.56, 4.36) compared to those in the lowest quartile (Q1) in Model 3. A significant risk increase was also evident for the third quartile (Q3; HR 1.86, 95% CI 1.39, 2.49), whereas the risk for the second quartile (Q2) remained non-significant. This graded association was consistent across progressively adjusted models.

Table 2

Table 2. Association between METS-IR and diabetes mellitus in normal-weight middle-aged and older Chinese adults.

Threshold effect of METS-IR on DM in normal-weight middle-aged and older Chinese adults

The dose-response relationship between METS-IR and diabetes risk was significantly non-linear (P < 0.001), characterized by a steep increase in risk below the inflection point of 37.24 (HR 1.18 per unit) that plateaued to non-significance above this threshold (Table 3 and Figure 3).

Table 3

Table 3. Threshold effect analysis between METS-IR and diabetes mellitus in normal-weight middle-aged and older Chinese adults.

Figure 3

Figure 3. The non-linear relationship between METS-IR and diabetes mellitus in normal-weight middle-aged and older Chinese adults. A non-linear relationship between them was detected after adjusting for gender, age, SBP, DBP, family history of diabetes, TC, LDL-C, Scr, BUN, and ALT. The red curve on the graph represents the smoothed fit between the variables, while the blue lines indicate the 95% confidence intervals surrounding the fit.

To assess the robustness of the results, sensitivity analyses were performed to address covariates with missing proportions exceeding 5%. First, analyses were conducted by including variables with high rates of missing data (smoking and drinking status) without imputation. Second, all covariates with more than 5% missing values (including smoking and drinking status) were assigned dummy variables, and the models were re-fit accordingly. The non-linear association between METS-IR and DM in normal-weight, middle-aged and older Chinese adults remained evident, confirming the stability of our primary results (Supplementary Tables S1, S2).

Subgroup analyses

Subgroup analyses explored potential effect modification by age, family history of diabetes, and gender (Table 4). The positive association was consistently observed in all predefined subgroups. The strength of this association appeared to be modified by age (P for interaction = 0.0119), with a slightly higher hazard ratio observed in the middle-aged group (45 to < 60 years; HR 1.14, 95% CI 1.11–1.17) compared to the older group (≥60 years; HR 1.09, 95% CI 1.06–1.12). In contrast, no significant effect modification was detected for sex (P for interaction = 0.8618) or family history of diabetes (P for interaction = 0.5050).

Table 4

Table 4. Subgroup analysis.

Discussion

This multicenter, retrospective cohort study demonstrated a robust, non-linear, and positive association between METS-IR and incident diabetes in normal-weight middle-aged and older Chinese adults. Following comprehensive adjustment for demographic, clinical, and biochemical confounders, each 1-unit rise in METS-IR corresponded to a 12% increase in diabetes risk. A clear inflection point was observed at a METS-IR of 37.24, below which the risk increased more sharply (HR: 1.18 per unit), while above this point, the association was attenuated and lost statistical significance.

METS-IR serves as a promising non-insulin-based indicator by incorporating critical components such as FPG, BMI, HDL-C, and TG (12). This integrative approach enables a simple and cost-effective assessment of IR, which is particularly useful in large epidemiological studies (27, 28). Our findings in a normal-weight population align with prior studies in mixed-weight and general cohorts, confirming a positive link between METS-IR and diabetes risk (29, 30). Notably, Chen et al. identified a non-linear association between METS-IR and diabetes risk, similar to our findings; however, compared with our METS-IR inflection point of 37.24, theirs occurred at a higher METS-IR level (30). This difference may reflect varying metabolic risk patterns across study cohorts. Our research specifically evaluated normal-weight adults aged ≥45 years (BMI 18.5–23.9 kg/m²), a demographic that may exhibit a heightened susceptibility to the effects of IR, even at comparatively lower METS-IR values. Research has indicated that metabolic risk factors exert different influences depending on an individual's age and body composition (24, 31). Moreover, it is plausible that this subgroup presents heightened metabolic vulnerability; factors such as progressive β-cell dysfunction and adverse body composition, despite a normal BMI (including possible sarcopenia or greater visceral adiposity), could exacerbate diabetes risk and allow earlier detection of an inflection point (32, 33). Furthermore, the saturation effect observed beyond the inflection point in our cohort suggests that at higher METS-IR levels, additional pathophysiological processes likely increase diabetes risk. The saturation effect noted beyond this inflection point—where the association with diabetes risk significantly attenuates—emphasizes the need for a nuanced understanding in specific populations. Thus, recognizing these distinctions underscores the importance of population- tailored risk stratification and reinforces the argument that universal cut-off values for metabolic indices, such as METS-IR may not suit all demographic groups (34, 35).

The underlying mechanisms linking METS-IR to the risk of DM, particularly in normal-weight middle-aged and older Chinese adults, may involve several interconnected pathways. Even within the normal BMI range, elevated METS-IR levels may signal ectopic fat accumulation, especially in the liver and muscles, thus thereby impairing insulin signaling and adversely affecting glucose metabolism. This suggests an early onset of diabetes risk, even in ostensibly healthy individuals (30, 34). Dyslipidemia manifests as high TG and low HDL-C levels, hallmark features of IR. These lipid abnormalities indicate a predisposition to metabolic dysfunction regardless of body weight (36, 37). Furthermore, elevated FPG reflects declining glucose regulation, signaling a chronic state of reduced insulin sensitivity. These observations highlight the inadequacy of BMI as a sole health metric. Its inability to capture harmful body composition changes, particularly visceral adiposity and age-related sarcopenia, renders it a misleading indicator, in older adults (38, 39). The relationship between METS-IR and diabetes is not strictly linear; evidence suggests that a metabolic threshold exists below which, incremental increases in METS-IR correspond to a significant rise in diabetes risk, reflecting a progressive deterioration in IR and metabolic stability (40). Once this threshold is surpassed, impaired beta-cell secretion may become the primary limiting factor in diabetes progression, emphasizing the value of early intervention before clinical symptoms appear. These findings underscore the limitations of relying solely on traditional metrics, such as BMI, and instead support broader evaluations of metabolic health, as increased METS-IR levels can precede diabetes diagnosis (41).

METS-IR, as an integrated measure of glucose, lipid, and adiposity parameters, may also be closely related to liver health and vascular structural changes (42). For instance, elevated ALT levels often indicate NAFLD, which is not uncommon in normal-weight individuals—a condition termed lean NAFLD (11, 43). Moreover, an increase in CIMT can also occur in normal-weight middle-aged and elderly individuals who exhibit cardiometabolic risk factors, including central obesity, hypertension, or elevated blood glucose (20). While our study lacked data on NAFLD or subclinical atherosclerosis, the METS-IR components are known correlates of these conditions. Future studies integrating imaging or biomarker data could explore whether the METS-IR-DM relationship is mediated through hepatic or vascular pathways.

Our study has several notable strengths. First, the findings are supported by data from a sizable multicenter cohort in China, ensuring robust and reliable evidence. Second, our investigation focuses on middle-aged and older individuals with normal body weight—a demographic often considered “low-risk” and overlooked in routine metabolic screening. Moreover, our analysis reveals a non-linear positive association between METS-IR and DM within this population, thereby contributing a novel perspective to the existing literature.

Nevertheless, certain limitations should be acknowledged. First, the observational cohort design precludes definitive causal inference between METS-IR and DM. Second, because this study involves a secondary analysis of a publicly available dataset, data on NAFLD and CIMT were unavailable, restricting comprehensive evaluation. Third, as the evidence stems from health examination data of a Chinese cohort, caution is warranted when extrapolating these findings to other populations.

Finally, despite adjustments for known confounders, this observational study cannot eliminate potential residual bias from unmeasured factors, such as diet and physical activity. Most critically, A major limitation is the very high rate of missing data for key lifestyle confounders (smoking, drinking). This prevents adequate control for these factors and limits the interpretability of models that include them.

Based on the novel demonstration of a non-linear relationship between METS-IR and diabetes risk, with a distinct threshold at METS-IR = 37.24 in normal-weight middle-aged and older Chinese adults, we recommend that researchers: (i) initiate prospective studies to establish causality between METS-IR and diabetes incidence, (ii) clarify dynamic changes in insulin sensitivity and β-cell compensation underlying the observed METS-IR threshold, and (iii) enhance integration of this metric into precision–prevention frameworks to address gaps in current metabolic risk assessment for normal-weight individuals, (iv) validate the identified METS-IR threshold (37.24) in independent, ethnically diverse cohorts of normal-weight individuals.

Conclusions

This large cohort study identified a significant non-linear relationship between METS-IR and DM risk among normal-weight middle-aged and older Chinese adults, with a threshold effect around 37.24 below which the risk increased markedly. These findings highlight the potential of METS-IR to improve early risk stratification and inform prevention strategies in this population.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://datadryad.org/stash/dataset/doi:10.5061%2Fdryad.ft8750v.

Ethics statement

Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants' legal guardians/next of kin in accordance with the national legislation and the institutional requirements.

Author contributions

HC: Conceptualization, Data curation, Formal analysis, Software, Visualization, Writing – original draft. ZC: Conceptualization, Data curation, Formal analysis, Software, Visualization, Writing – original draft. QC: Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Science and Technology Major Program on the Prevention and Control of Emerging, Sudden and Major Infectious Diseases (Grant No. 2025ZD01906000, Grant No. 2025ZD01906003), and the National Key Research and Development Program of China (Grant No. 2022YFC2304402).

Acknowledgments

We greatly appreciate the raw data provided by Chen, Ying et al.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2025.1717792/full#supplementary-material

Abbreviations

METS-IR, metabolic score for insulin resistance; DM, diabetes mellitus; IR, insulin resistance; T2DM, type 2 diabetes mellitus; NAFLD, non-alcoholic fatty liver disease; CIMT, carotid intima-media thickness; CIs, confidence intervals; HRs, hazard ratios; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; ALT, alanine aminotransferase; Scr, serum creatinine; BUN, blood urea nitrogen; FPG, fasting plasma glucose.

References

1. Sheen Y-J, Hsu C-C, Jiang Y-D, Huang C-N, Liu J-S, Sheu WH-H. Trends in prevalence and incidence of diabetes mellitus from 2005 to 2014 in Taiwan. J Formos Med Assoc. (2019) 118(Suppl. 2):S66–73. doi: 10.1016/j.jfma.2019.06.016

PubMed Abstract | Crossref Full Text | Google Scholar

2. Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. Diabetes atlas: global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. (2018) 138:271–81. doi: 10.1016/j.diabres.2018.02.023

Crossref Full Text | Google Scholar

3. Raoufinia R, Rahimi HR, Saburi E, Moghbeli M. Advances and challenges of the cell-based therapies among diabetic patients. J Transl Med. (2024) 22:435. doi: 10.1186/s12967-024-05226-3

PubMed Abstract | Crossref Full Text | Google Scholar

4. Azzam AY, Mora LM, Morsy MM, Essibayi MA, Altschul DJ, Nassar M. Epidemiological Patterns of Diabetes Mellitus in The United States of America: An Observational Multicenter Analysis From 1990 to 2024. ASIDE Intern Med. (2025) 1:1–7. doi: 10.71079/ASIDE.IM.02202517

PubMed Abstract | Crossref Full Text | Google Scholar

5. Yaribeygi H, Sathyapalan T, Atkin SL, Sahebkar A. Molecular Mechanisms Linking Oxidative Stress and Diabetes Mellitus. Oxid Med Cell Longev. (2020) 2020:8609213. doi: 10.1155/2020/8609213

PubMed Abstract | Crossref Full Text | Google Scholar

6. Iovino S, Burkart AM, Warren L, Patti ME, Kahn CR. Myotubes derived from human-induced pluripotent stem cells mirror in vivo insulin resistance. Proc Natl Acad Sci USA. (2016) 113:1889–94. doi: 10.1073/pnas.1525665113

PubMed Abstract | Crossref Full Text | Google Scholar

7. Mukherjee S, Im S-S. Decoding health: exploring essential biomarkers linked to metabolic dysfunction-associated steatohepatitis and type 2 diabetes mellitus. Biomedicines. (2025) 13:359. doi: 10.3390/biomedicines13020359

PubMed Abstract | Crossref Full Text | Google Scholar

8. Wang Q, Zhang X, Fang L, Guan Q, Guan L, Li Q. Prevalence, awareness, treatment and control of diabetes mellitus among middle-aged and elderly people in a rural Chinese population: a cross-sectional study. PLoS ONE. (2018) 13:e0198343. doi: 10.1371/journal.pone.0198343

PubMed Abstract | Crossref Full Text | Google Scholar

9. Rattarasarn C. Dysregulated lipid storage and its relationship with insulin resistance and cardiovascular risk factors in non-obese Asian patients with type 2 diabetes. Adipocyte. (2018) 7:71–80. doi: 10.1080/21623945.2018.1429784

PubMed Abstract | Crossref Full Text | Google Scholar

10. Li Y, Wang J, Tang Y, Han X, Liu B, Hu H, et al. Bidirectional association between nonalcoholic fatty liver disease and type 2 diabetes in Chinese population: evidence from the Dongfeng-Tongji cohort study. PLoS ONE. (2017) 12:e0174291. doi: 10.1371/journal.pone.0174291

PubMed Abstract | Crossref Full Text | Google Scholar

11. Kuchay MS, Martínez-Montoro JI, Choudhary NS, Fernández-García JC, Ramos-Molina B. Non-alcoholic fatty liver disease in lean and non-obese individuals: current and future challenges. Biomedicines. (2021) 9:1346. doi: 10.3390/biomedicines9101346

PubMed Abstract | Crossref Full Text | Google Scholar

12. Bello-Chavolla OY, Almeda-Valdes P, Gomez-Velasco D, Viveros-Ruiz T, Cruz-Bautista I, Romo-Romo A, et al. METS-IR, a novel score to evaluate insulin sensitivity, is predictive of visceral adiposity and incident type 2 diabetes. Eur J Endocrinol. (2018) 178:533–44. doi: 10.1530/EJE-17-0883

PubMed Abstract | Crossref Full Text | Google Scholar

13. Duan M, Zhao X, Li S, Miao G, Bai L, Zhang Q, et al. Metabolic score for insulin resistance (METS-IR) predicts all-cause and cardiovascular mortality in the general population: evidence from NHANES 2001–2018. Cardiovasc Diabetol. (2024) 23:243. doi: 10.1186/s12933-024-02334-8

PubMed Abstract | Crossref Full Text | Google Scholar

14. Alogaili MH, Alsallami AA. The association of the metabolic score for insulin resistance (METS-IR) with prehypertension in normoglycemic individuals. Int J Endocrinol Metab. (2024) 22:e145894. doi: 10.5812/ijem-145894

PubMed Abstract | Crossref Full Text | Google Scholar

15. Liu X-H, Xu Q, Zhang L, Liu H-J. Association between metabolic score for insulin resistance and regression to normoglycemia from prediabetes in Chinese adults: a retrospective cohort study. PLoS ONE. (2024) 19:e0308343. doi: 10.1371/journal.pone.0308343

PubMed Abstract | Crossref Full Text | Google Scholar

16. Bello-Chavolla OY, Antonio-Villa NE, Vargas-Vázquez A, Martagón AJ, Mehta R, Arellano-Campos O, et al. Prediction of incident hypertension and arterial stiffness using the non-insulin-based metabolic score for insulin resistance (METS-IR) index. J Clin Hypertens (Greenwich). (2019) 21:1063–70. doi: 10.1111/jch.13614

PubMed Abstract | Crossref Full Text | Google Scholar

17. Hou Y, Li R, Xu Z, Chen W, Li Z, Jiang W, et al. Association of METS-IR index with type 2 diabetes: a cross-sectional analysis of national health and nutrition examination survey data from 2009 to 2018. PLoS ONE. (2024) 19:e0308597. doi: 10.1371/journal.pone.0308597

PubMed Abstract | Crossref Full Text | Google Scholar

18. Tang A, Coster ACF, Tonks KT, Heilbronn LK, Pocock N, Purtell L, et al. Longitudinal changes in insulin resistance in normal weight, overweight and obese individuals. J Clin Med. (2019) 8:623. doi: 10.3390/jcm8050623

PubMed Abstract | Crossref Full Text | Google Scholar

19. Lee S-H, Han K, Yang HK, Kim H-S, Cho J-H, Kwon H-S, et al. novel criterion for identifying metabolically obese but normal weight individuals using the product of triglycerides and glucose. Nutr Diabetes. (2015) 5:e149. doi: 10.1038/nutd.2014.46

Crossref Full Text | Google Scholar

20. Yang Q, Lin Q, Guo D, Wang H, Liu J, Zhang X, et al. Association of carotid intima media thickness with metabolic syndrome among low-income middle-aged and elderly Chinese: a population-based cross-sectional study. Front Cardiovasc Med. (2021) 8:669245. doi: 10.3389/fcvm.2021.669245

PubMed Abstract | Crossref Full Text | Google Scholar

21. Tarantino G, Costantini S, Finelli C, Capone F, Guerriero E, La Sala N, et al. Carotid intima-media thickness is predicted by combined eotaxin levels and severity of hepatic steatosis at ultrasonography in obese patients with nonalcoholic fatty liver disease. PLoS ONE. (2014) 9:e105610. doi: 10.1371/journal.pone.0105610

PubMed Abstract | Crossref Full Text | Google Scholar

22. Lv X, Zhou W, Sun J, Lin R, Ding L, Xu M, et al. Visceral adiposity is significantly associated with type 2 diabetes in middle-aged and elderly Chinese women: a cross-sectional study. J Diabetes. (2017) 9:920–8. doi: 10.1111/1753-0407.12499

PubMed Abstract | Crossref Full Text | Google Scholar

23. Verkouter I, Noordam R, le Cessie S, van Dam RM, Lamb HJ, Rosendaal FR, van Heemst D, de Mutsert R. The association between adult weight gain and insulin resistance at middle age: mediation by visceral fat and liver fat. J Clin Med. (2019) 8:1559. doi: 10.3390/jcm8101559

PubMed Abstract | Crossref Full Text | Google Scholar

24. Chen Y, Zhang X-P, Yuan J, Cai B, Wang X-L, Wu X-L, et al. Association of body mass index and age with incident diabetes in Chinese adults: a population-based cohort study. BMJ Open. (2018) 8:e021768. doi: 10.1136/bmjopen-2018-021768

PubMed Abstract | Crossref Full Text | Google Scholar

25. ElSayed NA, Aleppo G, Aroda VR, Bannuru RR, Brown FM, Bruemmer D, et al. 2. Classification and diagnosis of diabetes: standards of care in diabetes-−2023. Diabetes Care. (2023) 46:S19–40. doi: 10.2337/dc23-S002

Crossref Full Text | Google Scholar

26. Zheng X, Zhang X, Han Y, Hu H, Cao C. Nonlinear relationship between atherogenic index of plasma and the risk of prediabetes: a retrospective study based on Chinese adults. Cardiovasc Diabetol. (2023) 22:205. doi: 10.1186/s12933-023-01934-0

PubMed Abstract | Crossref Full Text | Google Scholar

27. Yang W, Cai X, Hu J, Wen W, Mulalibieke H, Yao X, et al. The Metabolic Score for Insulin Resistance (METS-IR) Predicts Cardiovascular Disease and Its Subtypes in Patients with Hypertension and Obstructive Sleep Apnea. Clin Epidemiol. (2023) 15:177–89. doi: 10.2147/CLEP.S395938

PubMed Abstract | Crossref Full Text | Google Scholar

28. Cai X, Hu J, Zhu Q, Wang M, Liu S, Dang Y, et al. Relationship of the metabolic score for insulin resistance and the risk of stroke in patients with hypertension: a cohort study. Front Endocrinol (Lausanne). (2022) 13:1049211. doi: 10.3389/fendo.2022.1049211

PubMed Abstract | Crossref Full Text | Google Scholar

29. Lu S, Wang Q, Lu H, Kuang M, Zhang M, Sheng G, et al. Lipids as potential mediators linking body mass index to diabetes: evidence from a mediation analysis based on the NAGALA cohort. BMC Endocr Disord. (2024) 24:66. doi: 10.1186/s12902-024-01594-5

PubMed Abstract | Crossref Full Text | Google Scholar

30. Chen Z, Huang C, Zhou Z, Zhang Y, Xu M, Tang Y, et al. nonlinear associations of metabolic score for insulin resistance index with incident diabetes: a retrospective Chinese cohort study. Front Clin Diabetes Healthc. (2022) 3:1101276. doi: 10.3389/fcdhc.2022.1101276

Crossref Full Text | Google Scholar

31. Fu W, Wang C, Zou L, Jiang H, Miller M, Gan Y, et al. Association of adiposity with diabetes: a national research among Chinese adults. Diabetes Metab Res Rev. (2021) 37:e3380. doi: 10.1002/dmrr.3380

PubMed Abstract | Crossref Full Text | Google Scholar

32. Tamehri Zadeh SS, Cheraghloo N, Masrouri S, Esmaeili F, Azizi F, Hadaegh F. Association between metabolic score for insulin resistance and clinical outcomes: insights from the Tehran lipid and glucose study. Nutr Metab (Lond). (2024) 21:34. doi: 10.1186/s12986-024-00808-w

PubMed Abstract | Crossref Full Text | Google Scholar

33. Zhang X, Liu F, Li W, Zhang J, Zhang T, Yu X, et al. Metabolic score for insulin resistance (METS-IR) predicts adverse cardiovascular events in patients with type 2 diabetes and ischemic cardiomyopathy. Diabetes Metab Syndr Obes. (2023) 16:1283–95. doi: 10.2147/DMSO.S404878

PubMed Abstract | Crossref Full Text | Google Scholar

34. Sheng G, Kuang M, Yang R, Zou Y. Association of metabolic score for insulin resistance with progression or regression of prediabetes: evidence from a multicenter Chinese medical examination cohort study. Front Endocrinol (Lausanne). (2024) 15:1388751. doi: 10.3389/fendo.2024.1388751

PubMed Abstract | Crossref Full Text | Google Scholar

35. Gonzalez-Gay MA, Ferraz-Amaro I, Castañeda S, Pinto Tasende JA, Uriarte-Ecenarro M, Plaza Z, et al. The metabolic score for insulin resistance predicts the risk of cardiovascular disease in patients with psoriatic arthritis: results from the 10-year prospective CARMA cohort. RMD Open. (2025) 11:e005352. doi: 10.1136/rmdopen-2024-005352

PubMed Abstract | Crossref Full Text | Google Scholar

36. Pan L, Zou H, Meng X, Li D, Li W, Chen X, et al. Predictive values of metabolic score for insulin resistance on risk of major adverse cardiovascular events and comparison with other insulin resistance indices among Chinese with and without diabetes mellitus: results from the 4C cohort study. J Diabetes Investig. (2023) 14:961–72. doi: 10.1111/jdi.14024

PubMed Abstract | Crossref Full Text | Google Scholar

37. Elrayess MA, Rizk NM, Fadel AS, Kerkadi A. Prevalence and predictors of insulin resistance in non-obese healthy young females in Qatar. Int J Environ Res Public Health. (2020) 17:5088. doi: 10.3390/ijerph17145088

PubMed Abstract | Crossref Full Text | Google Scholar

38. Fkh S, Z K, M J. Insulin resistance and high molecular weight adiponectin in obese and non-obese patients with Polycystic Ovarian Syndrome (PCOS). BMC Endocr Disord. (2021) 21:45. doi: 10.1186/s12902-021-00710-z

PubMed Abstract | Crossref Full Text | Google Scholar

39. Scott RA, Fall T, Pasko D, Barker A, Sharp SJ, Arriola L, et al. Common genetic variants highlight the role of insulin resistance and body fat distribution in type 2 diabetes, independent of obesity. Diabetes. (2014) 63:4378–87. doi: 10.2337/db14-0319

PubMed Abstract | Crossref Full Text | Google Scholar

40. McLaughlin T, Abbasi F, Lamendola C, Yee G, Carter S, Cushman SW. Dietary weight loss in insulin-resistant non-obese humans: metabolic benefits and relationship to adipose cell size. Nutr Metab Cardiovasc Dis. (2019) 29:62–8. doi: 10.1016/j.numecd.2018.09.014

PubMed Abstract | Crossref Full Text | Google Scholar

41. Lim SM, Choi DP, Rhee Y, Kim HC. Association between obesity indices and insulin resistance among healthy Korean adolescents: the JS high school study. PLoS ONE. (2015) 10:e0125238. doi: 10.1371/journal.pone.0125238

PubMed Abstract | Crossref Full Text | Google Scholar

42. Zou H, Ma X, Pan W, Xie Y. The association of the metabolic score for insulin resistance and metabolic dysfunction-related steatosis liver disease: a population-based observational study. Eur J Med Res. (2025) 30:657. doi: 10.1186/s40001-025-02936-2

PubMed Abstract | Crossref Full Text | Google Scholar

43. Elizondo-Montemayor L, Ugalde-Casas PA, Lam-Franco L, Bustamante-Careaga H, Serrano-González M, Gutiérrez NG, et al. Association of ALT and the metabolic syndrome among Mexican children. Obes Res Clin Pract. (2014) 8:e79–87. doi: 10.1016/j.orcp.2012.08.191

PubMed Abstract | Crossref Full Text | Google Scholar