Association between metabolic score for insulin resistance and hypertension in middle-aged and older adults: a nationwide cross-sectional and longitudinal study

Abstract

Introduction

Metabolic score for insulin resistance (METS-IR) is a promising and reliable non-insulin-based method for assessing insulin resistance and cardiovascular disease risk. Through cross-sectional and longitudinal analyses, we evaluated the predictive ability of METS-IR for hypertension risk, aiming to provide a basis for early screening and intervention strategies to reduce hypertension-related complications.

Methods

Data were utilized from the China Health and Retirement Longitudinal Study (CHARLS), involving participants aged 45 and above with complete METS-IR and self-reported hypertension records. A combination of cross-sectional (waves 1 and 3) and longitudinal designs was used to track hypertension incidence, with 10,738 participants in the cross-sectional analysis and 6,788 participants in the longitudinal analysis. Data collection included health behaviors, physical measurements, and blood tests. METS-IR was calculated using a standardized formula, and hypertension was defined according to established criteria. Statistical analyses assessed associations, with a significance level set at p < 0.05.

Results

The cross-sectional analysis, after adjusting for confounders, showed that each unit increase in METS-IR was associated with a threefold increase in the risk of hypertension (adjusted OR = 3.48, 95% CI: 2.87, 4.21, P < 0.0001). A significant nonlinear relationship between METS-IR and hypertension risk was observed, particularly beyond a value of 2.0, where the risk significantly increased. Subgroup analysis revealed that smoking and sex significantly affected this association. Longitudinal analysis demonstrated that each unit increase in METS-IR was significantly associated with increased hypertension incidence at 2 years (OR = 1.54, 95% CI: 1.09, 2.18, P < 0.05), 5 years (OR = 1.52, 95% CI: 1.17, 1.97, P < 0.05), and 7 years (OR = 1.71, 95% CI: 1.38, 2.12,, P < 0.0001), respectively.

Conclusion

Using CHARLS data, we found that higher METS-IR independently predicts hypertension incidence and prevalence in Chinese older adults.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40520-025-03099-0

Introduction

Hypertension is a global public health issue and a major risk factor for cardiovascular diseases (CVD), with its pathogenesis closely related to metabolic abnormalities [1]. With the development of the socio-economic environment and accelerated aging of the population, the prevalence and incidence of hypertension continued to rise, particularly among middle-aged and older adults. In recent years, metabolic syndrome and its related insulin resistance have become key areas of research in hypertension. The metabolic score for insulin resistance (METS-IR), an indicator used to assess individual insulin resistance, can better reflect a person’s metabolic health [2]. METS-IR is calculated by combining metabolic parameters such as fasting blood glucose (FBG), triglycerides (TG), body mass index (BMI), and high-density lipoprotein cholesterol (HDL-C) to measure the degree of insulin resistance [3]. Insulin resistance is a key driver of hypertension, while elevated blood glucose, triglycerides, and reduced HDL-C levels further exacerbate the process of atherosclerosis, vascular endothelial dysfunction, and arterial stiffness, leading to elevated blood pressure [4]. Therefore, studying the relationship between METS-IR and hypertension is critical for understanding the impact of metabolic abnormalities on hypertension.

Although previous studies have shown a significant association between METS-IR and metabolic diseases such as cardiovascular disease and obesity, the relationship between METS-IR and the prevalence and incidence of hypertension, particularly the variations in different age groups, remains insufficiently explored in large-scale epidemiological data [5]. Most existing research has focused on specific populations or small sample sizes, limiting the understanding of the broader applicability of METS-IR. This is especially relevant for the middle-aged and elderly population, where the cumulative effect of metabolic abnormalities becomes increasingly apparent, and fluctuations in METS-IR may be closely linked to the development of hypertension [6]. Thus, the primary aim of this study is to investigate the association between METS-IR and hypertension using nationally representative data, and to further analyze the predictive capacity of METS-IR for hypertension risk under varying metabolic conditions, providing valuable insight into personalized prevention and intervention strategies for hypertension.

Therefore, this study aims to investigate the association between METS-IR and both the prevalence and incidence of hypertension in a nationally representative sample of middle-aged and older Chinese adults, using data from the China Health and Retirement Longitudinal Study (CHARLS). By evaluating its predictive value under varying metabolic conditions, we seek to provide evidence for the potential application of METS-IR as a population-level screening tool to support early identification and prevention strategies for hypertension.

Methods

Study design and population

This study utilized data from the China Health and Retirement Longitudinal Study (CHARLS), which covers participants from 28 provinces, 150 counties or districts, and 450 villages across China. The dataset includes demographic, socioeconomic, health status, blood tests, and functional information. Baseline data were collected in Wave 1(2011), with follow-up surveys conducted every two years in Wave 2 (2013), Wave 3 (2015), Wave 4 (2018) and Wave 5 (2020). Blood tests were only carried out in 2011 and 2015.

To explore the relationship between METS-IR and hypertension, we employed a combined cross-sectional and longitudinal study design. In the cross-sectional analysis, we used data from Wave 1 and Wave 3. The inclusion criteria were: (1) having fasting blood glucose data; (2) aged 45 years or older with complete sociodemographic information; (3) complete hypertension diagnostic data and METS-IR measurements. The exclusion criteria were: (1) participants younger than 45 years or with incomplete data; (2) individuals diagnosed with diabetes at baseline or those with unclear diabetes status during follow-up, to maintain the focus on pre-diabetic insulin resistance. Diabetes represents a state of advanced insulin resistance accompanied by complex metabolic disturbances, including those that directly affect blood pressure regulation. Including diabetic individuals would have introduced additional heterogeneity and made it difficult to isolate the independent association between METS-IR and hypertension; (3) participants taking glucose-lowering or lipid-lowering medications, as well as those with incomplete systolic and diastolic blood pressure information. Ultimately, 10,738 participants met the criteria for cross-sectional analysis, as illustrated in Fig. 1.

Fig. 1.

Flowchart of participant selection

In the longitudinal analysis, we selected participants who did not have hypertension in 2011 and tracked the incidence of hypertension during follow-up. The inclusion criteria required complete METS-IR values in 2011 and hypertension diagnostic data during follow-up, while the exclusion criteria were similar to those in the cross-sectional analysis. After screening, a total of 6,788 participants met the criteria for inclusion in the longitudinal analysis (Fig. 2).

Fig. 2.

Flowchart of participant selection for longitudinal analysis

Ethical approval

Access to the CHARLS dataset is available via its official website at charls.ccer.edu.cn/en. The project and the protocol for biomarker sample collection were approved by the Biomedical Ethics Review Committee of Peking University (IRB00001052-11014) and the Institutional Review Board of the National School of Development at Peking University (IRB00001052-11015), with informed consent obtained from all participants.

Data collection and potential covariates

The CHARLS study collected data using professionally trained personnel using structured questionnaires to gather sociodemographic information. Health-related behaviors, such as smoking and drinking status, medical history (heart disease, dyslipidemia, and stroke), and medication used for hypertension were also recorded. Physical measurements, including height, weight, and blood pressure, were conducted by trained professionals. Blood pressure was measured using an Omron HEM-7200 electronic sphygmomanometer, and the average of three readings was recorded. BMI was calculated as weight (kg) divided by height squared (m²).

Fasting venous blood samples were collected in the morning to measure FBG, glycated hemoglobin (HbA1c), TG, total cholesterol (TC), HDL-C, low-density lipoprotein cholesterol (LDL-C), serum creatinine (Scr), C-reactive protein (CRP), blood urea nitrogen (BUN), and serum uric acid (SUA) levels.

Covariates included sex, age, education level, smoking status, drinking status, TC, CRP, HbA1c, Scr, LDL-C, BUN, and SUA levels.

Measurement of METS-IR

The METS-IR was calculated [7] by (Ln[(2×FBG (mg/dL))+TG(mg/dL)]×BMI(kg/m²))/[Ln(HDL-C(mg/dL))], expressed in mmol/L. In subsequent analysis, we examined METS-IR as both a continuous variable and categorized it into two groups (Q1: METS-IR ≤ 2.03; Q2: METS-IR > 2.03) to enhance the analytical strength.

Definition of hypertension

According to established medical diagnostic criteria, participants meeting any of the following conditions were classified as hypertensive: systolic blood pressure (SBP) ≥ 140 mmHg or diastolic blood pressure (DBP) ≥ 90 mmHg without the use of antihypertensive medication. In this study, hypertension was defined as a binary variable [8]. Hypertension cases were identified through the following two questions: (1) whether the participant responded affirmatively to “Are you currently managing or treating hypertension through Traditional Chinese Medicine or Western medicine?” or (2) whether the participant answered, “Have you ever been diagnosed with hypertension by a doctor?” in the affirmative.

Statistical analysis

Statistical analyses were conducted using Empower (version 2.0) and R software (version 4.3.1). Continuous variables are expressed as means ± standard deviations, while categorical variables are presented as frequencies and percentages. Differences between variables across METS-IR groups were compared using one-way ANOVA, Kruskal-Wallis H test, or chi-square test. Three models were applied: Model 1 was unadjusted; Model 2 adjusted for age, sex, education level, smoking status and drinking status; Model 3 further adjusted for LDL-C, Scr, BUN, SUA, CRP, HbA1c, and TC based on Model 2. The association between METS-IR and hypertension prevalence was assessed, and results were reported as odds ratios (ORs) with 95% confidence intervals (CIs).

This study utilized linear trend analysis and restricted cubic spline (RCS) curves to explore the nonlinear relationship between METS-IR and hypertension prevalence. Subgroup analyses were conducted for hypertension, SBP, DBP, and pulse pressure (PP), with additional stratification by sex, age, alcohol consumption, smoking status, glycated hemoglobin, BMI, and dyslipidemia. Interactions were assessed using multivariable logistic regression models, and statistical significance was defined as p < 0.05.

To address missing data, we employed Multiple Imputation by Chained Equations (MICE) [9]. Seven continuous covariates—BUN, TC, LDL-C, Scr, CRP, HbA1c, and SUA—had missing rates ranging from approximately 5–40%. A total of five imputed datasets were generated under the assumption of missing at random (MAR), incorporating all relevant variables included in the multivariable analysis. Each dataset was analyzed separately, and results were pooled using Rubin’s rules to obtain valid statistical inferences. Sensitivity analyses were also performed and yielded consistent results, supporting the robustness of the findings.

In the longitudinal analysis, follow-up data from the 2011 CHARLS cohort were utilized to examine hypertension incidence over different periods (2, 5, and 7 years). Results were presented as ORs with corresponding 95% CIs, highlighting the long-term impact of METS-IR on the risk of developing hypertension.

Results

Characteristics of the participants

The CHARLS chart initially enrolled 25,873 participants. After excluding missing and abnormal data, a total of 10,738 participants were included in the cross-sectional analysis (Fig. 1). Based on the ROC curve threshold [10], METS-IR was divided into two groups: Q1 group (METS-IR ≤ 2.03, N = 5,829) and Q2 group (METS-IR > 2.03, N = 4,909). The two groups were compared, and participants in the Q2 group were younger (p < 0.001). The prevalence of hypertension, dyslipidemia, cardiovascular disease, and stroke was significantly higher in the Q2 group (p < 0.001). Additionally, SBP, DBP, SUA, BUN, Scr, CRP, and HbA1c levels were significantly higher in the Q2 group, indicating higher levels of metabolic markers (p < 0.001). The Q2 group had a significantly lower drinking rate (p < 0.001), while smoking rates showed slight differences. (Table 1).

Table 1.

Baseline characteristics of participants METS-IR in cross-sectional study

Variable	Total	Q1 ≤ 2.03	Q2 > 2.03	P value
Participants, sample size (N)	10,738	5829	4909	NA
Age, years	59.07 ± 9.34	59.84 ± 9.73	58.30 ± 8.94	< 0.001
Gender (%)				0.187
Male	5123 (47.65)	2815 (48.29)	2308 (47.02)
Female	5615 (52.35)	3014 (51.71)	2601 (52.98)
Education level (%)				< 0.001
Below high school or vocational school	9670 (89.93)	5326 (91.37)	4344 (88.49)
High school or vocational school	939 (8.83)	452 (7.75)	487 (9.92)
Above high school or vocational school	129 (1.23)	51 (0.87)	78 (1.59)
Marital status (%)				< 0.001
Married	8944 (83.43)	4775 (81.92)	4169 (84.93)
Married separation	450 (4.20)	244 (4.19)	206 (4.20)
Separated	46 (0.43)	26 (0.45)	20 (0.41)
Divorced	74 (0.69)	39 (0.67)	35 (0.71)
Widowed	1142 (10.51)	697 (11.96)	445 (9.06)
Never married	82 (0.76)	48 (0.82)	34 (0.69)
Hypertension (%)				< 0.001
yes	2729 (25.93)	1165 (19.99)	1564 (31.86)
no	8009 (74.07)	4664 (80.01)	3345 (68.14)
Dyslipidemia (%)				< 0.001
yes	608 (5.84)	218 (3.74)	390 (7.95)
no	10,127 (94.16)	5610 (96.26)	4517 (92.05)
Cardiovascular Disease (%)				< 0.001
yes	1194 (11.33)	542 (9.34)	652 (13.33)
no	9503 (88.67)	5264 (90.66)	4239 (86.67)
Stroke (%)				0.014
yes	227 (2.15)	105 (1.81)	122 (2.49)
no	10,489 (97.85)	5712 (98.19)	4777 (97.51)
SBP, mmHg	128.41 ± 20.60	126.45 ± 20.57	130.36 ± 20.63	< 0.001
DBP, mmHg	75.37 ± 11.86	73.68 ± 11.69	77.05 ± 12.03	< 0.001
PP, mmHg	53.04 ± 14.87	52.77 ± 14.93	53.31 ± 14.80	0.036
Antihypertensive medication (%)				< 0.001
yes	1843 (17.75)	735 (12.73)	1108 (22.76)
no	8800 (82.25)	5039 (87.27)	3761 (77.24)
Drinking (%)				< 0.001
yes	4719 (43.88)	2670 (45.82)	2049 (41.75)
no	6016 (56.12)	3157 (54.18)	2859 (58.25)
Smoking (%)				0.065
yes	3965 (40.39)	2206 (41.31)	1759 (39.47)
no	5831 (59.61)	3134 (58.69)	2697 (60.53)
TC, mg/dL	190.60 ± 38.09	189.38 ± 35.78	191.83 ± 40.39	< 0.001
LDL-C, mg/dL	112.56 ± 33.61	112.52 ± 31.76	112.60 ± 35.45	0.514
BUN, mg/dL	15.57 ± 4.49	16.01 ± 4.65	15.13 ± 4.32	< 0.001
SUA, mg/dL	4.62 ± 1.29	4.41 ± 1.23	4.82 ± 1.35	< 0.001
Scr, mg/dL	0.79 ± 0.23	0.78 ± 0.25	0.79 ± 0.21	< 0.001
CRP, mg/dL	2.69 ± 7.01	2.43 ± 7.01	2.94 ± 7.00	< 0.001
HbA1c (%)	5.40 ± 0.68	5.33 ± 0.59	5.46 ± 0.76	< 0.001
METS-IR	2.06 ± 0.17	1.86 ± 0.12	2.26 ± 0.22	< 0.001

Annotation: Data are presented as mean ± standard deviation or number (%)

Abbreviations: BMI– body mass index; BUN– blood urea nitrogen; CRP– C-reactive protein; DBP– diastolic blood pressure; HDL-C– high-density lipoprotein cholesterol; LDL-C– low-density lipoprotein cholesterol; METS-IR– metabolic score for insulin resistance; PP– pulse pressure; SBP– systolic blood pressure; Scr– serum creatinine; SUA– serum uric acid; TC– total cholesterol; TG– triglycerides; HbA1c–

Association between METS-IR and hypertension prevalence

Multivariate regression analysis demonstrated the association between METS-IR and hypertension, reporting OR values and their 95% confidence intervals across three models. In Model 1, each one-unit increase in METS-IR was associated with a 227% increase in hypertension risk, with an OR of 3.27 (95% CI: 2.78, 3.85, P < 0.0001). In Models 2 and 3, after adjusting for additional covariates, the OR values were 3.87 (95% CI: 3.25, 4.16, P < 0.0001) and 3.48 (95% CI: 2.87, 4.21, P < 0.0001), respectively, indicating the robustness of this association across different models. Furthermore, compared to the low METS-IR group, the OR values for the high METS-IR group were 1.87 (95% CI: 1.71, 2.04), 2.05 (95% CI: 1.87, 2.26), and 1.85 (95% CI: 1.68, 2.04), all showing a significantly increased risk of hypertension (P < 0.0001). These findings suggest that METS-IR is significantly associated with higher hypertension prevalence, with a consistent impact observed across various statistical models (Table 2).

Table 2.

Multivariate regression analysis of the association between METS-IR and hypertension prevalence

Variable	Model 1		Model 2		Model 3
	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value
METS-IR	3.27 (2.78, 3.85)	< 0.0001	3.87 (3.25, 4.16)	< 0.0001	3.48 (2.87, 4.21)	< 0.0001
Low METS-IR (METS-IR ≤ 2.03)	Ref.		Ref.		Ref.
High METS-IR (METS-IR > 2.03)	1.87 (1.71, 2.04)	< 0.0001	2.05 (1.87, 2.26)	< 0.0001	1.85 (1.68, 2.04)	< 0.0001

Annotation: Model 1 adjusted for none. Model 2 adjusted for gender, age, education level, smoking status and drinking status. Model 3 adjusted for LDL-C, Scr, BUN, SUA, CRP, HbA1c, TC on the basis of Model 2

METS-IR as a continuous variable and grouped by ROC threshold

Non-linear relationship between METS-IR and hypertension

Using linear trend analysis and RCS curves, after adjusting for all covariates, the results showed that as METS-IR values increased, the prevalence of hypertension rose. The analysis revealed a significant nonlinear relationship between METS-IR and the prevalence of hypertension (p for nonlinear < 0.001). This increase was particularly significant when the METS-IR value approached 2.0 (as shown in Fig. 3).

Fig. 3.

Restricted cubic splines were used to evaluate the linear relationship between METS-IR and hypertension prevalence. The blue solid line represents the probability of hypertension prevalence, and the light blue area represents the 95% confidence interval curve. The RCS model was adjusted for sex, age, education level, smoking status, drinking status, LDL-C, Scr, BUN, SUA, CRP, HbA1c, and TC

Additionally, the threshold effect analysis (as shown in Table S1) further supports the presence of a nonlinear association between METS-IR and hypertension prevalence. Using a restricted cubic spline (RCS) model with three knots placed at the 25th, 50th, and 75th percentiles of the METS-IR distribution, we identified a threshold point at METS-IR = 2.29. Below this threshold, the risk of hypertension increased significantly (OR = 5.37, 95% CI: 4.20–6.88, p < 0.0001). In contrast, above this threshold, the association was not statistically significant (OR = 1.32, 95% CI: 0.91–1.91, p = 0.1443). The log-likelihood ratio test (p < 0.001) confirmed that the threshold model provided a significantly better fit than the standard linear model. These findings reinforce the presence of a nonlinear relationship between METS-IR and hypertension prevalence and highlight the importance of considering threshold effects when evaluating metabolic predictors of hypertension.

Subgroup analysis

To ensure the reliability of our findings, we conducted subgroup analyses. As shown in Fig. 4, the figure contains four forest plots (A, B, C, D) illustrating the associations between different subgroups and hypertension (A), systolic blood pressure (SBP, B), diastolic blood pressure (DBP, C), and pulse pressure (PP, D). Panel A demonstrates that sex and smoking status significantly influence hypertension, with a notable interaction effect between these two factors (P < 0.05). Panel B reveals significant associations between sex, smoking status, and glycated hemoglobin (HbA1c) levels with SBP, along with a prominent interaction effect between age and smoking status (P < 0.05). Panel C indicates significant correlations between sex, smoking status, and DBP, with smoking status showing a particularly strong interaction effect (P < 0.05). Panel D highlights the associations of PP with drinking and smoking, while also identifying significant interaction effects between sex and smoking status. These findings suggest that the risk factors for hypertension, SBP, DBP, and PP vary across subgroups, particularly in terms of sex, smoking status, and BMI. Notably, smoking status emerged as a key determinant across all indicators, with significant interaction effects, underscoring its critical role in blood pressure-related conditions.

Fig. 4.

Subgroup analysis of the association between METS-IR and hypertension prevalence. Associations between METS-IR and (A) Hypertension; (B) SBP; (C) DBP and (D) PP were examined in subgroups defined by demographic and metabolic factors. ORs and 95% CIs were derived from logistic regression models adjusted for potential confounders. P for interaction indicates subgroup effect modification

Association between METS-IR and hypertension incidence

To further understand the comparative results between newly developed hypertension (new HTN) and non-hypertension (non-HTN) at different follow-up time points (2 years, 5 years, and 7 years), the OR for New HTN after 2 years of follow-up was 1.54 (95% CI: 1.09, 2.18, P = 0.0138), indicating a significantly higher risk for New HTN compared to non-HTN. During 5 years of follow-up, the OR for New HTN was 1.52 (95% CI: 1.17, 1.97, P = 0.0015), also showing a marked increase in risk. By the 7-year follow-up, the risk further escalated, with an OR for New HTN of 1.71 (95% CI: 1.38, 2.12, P < 0.0001), demonstrating that the extended follow-up period is significantly associated with an increased risk of New HTN. These results highlight the importance of monitoring hypertension over long-term follow-up, suggesting that the risk of New HTN cases continues to rise over time. (Table 3).

Table 3.

Multivariate regression analysis of the association between METS-IR and hypertension incidence

Variables	β	OR (95%CI)	P value
2 years followup			0.0138
Non-HTN	Ref.	Ref.
New HTN	0.43	1.54(1.09,2.18)
5 years followup			0.0015
Non-HTN	Ref.	Ref.
New HTN	0.42	1.52(1.17,1.97)
7 years followup			< 0.0001
Non-HTN	Ref.	Ref.
New HTN	0.54	1.71(1.38,2.12)

OR: Odds Ratio, CI: Confidence Interval, HTN: Hypertension

Association between METS-IR and blood pressure measurement

This study comprehensively explored the relationship between METS-IR and blood pressure (SBP and DBP). According to the 2018 ESC/ESH hypertension diagnostic criteria [11], the thresholds for SBP and DBP were set at 140 mmHg and 90 mmHg, respectively. Meanwhile, the 2017 ACC/AHA guidelines and the China Rural Hypertension Control Project lowered the diagnostic thresholds to 130/80 mmHg [12, 13], underscoring the cardiovascular risk within this range. To investigate the association of METS-IR under these two diagnostic standards, our analysis revealed that within the SBP range of 130–140 mmHg, each one-unit increase in METS-IR was associated with an OR of 1.39 (95% CI: 1.11, 1.73, p = 0.0035) in Model 3, indicating a significant positive correlation between METS-IR and mildly elevated SBP. Similarly, within the DBP range of 80–90 mmHg, each one-unit increase in METS-IR corresponded to an OR of 1.68 (95% CI: 1.38, 2.04, p < 0.0001), also demonstrating a significant positive association. Furthermore, compared to the low METS-IR group (Q1), the high METS-IR group (Q2) exhibited significantly higher ORs for both SBP and DBP across different models, further corroborating this relationship. These findings suggest that higher METS-IR levels substantially increase the risk of mildly elevated blood pressure and may serve as a potential marker for hypertension risk (Table 4).

Table 4.

Multivariate regression analysis of METS-IR status and indicators of blood pressure

Variable	Model 1		Model 2		Model 3
	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value
140 mmHg > Systolic pressure ≥ 130 mmHg
METS-IR	1.26 (1.04, 1.53)	0.0187	1.36 (1.12, 1.66)	0.0022	1.39 (1.11, 1.73)	0.0035
Q1 (METS-IR ≤ 2.03)	Ref.		Ref.		Ref.
Q2 (METS-IR > 2.03)	1.17 (1.06, 1.30)	0.0028	1.22 (1.10, 1.36)	0.0003	1.21 (1.09, 1.36)	0.0007
90 mmHg > Diastolic pressure ≥ 80 mmHg
METS-IR	1.88 (1.59, 2.23)	< 0.0001	1.83 (1.54, 2.19)	< 0.0001	1.68 (1.38, 2.04)	< 0.0001
Q1 (METS-IR ≤ 2.03)	Ref.		Ref.		Ref.
Q2 (METS-IR > 2.03)	1.38 (1.26, 1.52)	< 0.0001	1.36 (1.23, 1.50)	< 0.0001	1.30 (1.18, 1.44)	< 0.0001

Annotation: Model 1 adjusted for none. Model 2 adjusted for gender, age, education level, smoking status, drinking status

Model 3 adjusted for LDL-C, Scr, BUN, SUA, CRP, HbA1c, TC on the basis of Model 2

METS-IR as a continuous variable and grouped by ROC threshold

Discussion

In recent years, the prevalence of metabolic syndrome has been steadily increasing, becoming a major global public health concern, particularly in the context of rising obesity rates. Various biomarkers and indicators associated with metabolic syndrome have garnered growing attention, especially METS-IR [14]. METS-IR has been widely applied for its effectiveness in predicting cardiovascular disease risk. It integrates several factors closely related to insulin resistance and metabolic abnormalities, including fasting blood glucose, triglycerides, and HDL-C. These factors not only reflect an individual’s metabolic health status but are also strongly associated with cardiovascular risk factors such as hypertension, obesity, and dyslipidemia, further validating the clinical utility of METS-IR.

A large body of research has demonstrated the critical role of METS-IR in assessing cardiovascular disease risk [15]. With changes in lifestyle, particularly in diet and physical activity, the incidence of obesity and metabolic syndrome continues to rise, leading to an increased risk of cardiovascular diseases [16]. Therefore, identifying and monitoring effective indicators related to cardiovascular disease has become paramount. In this context, the introduction of METS-IR offers a novel perspective for early intervention and personalized management [17, 18]. Our study aimed to explore the relationship between METS-IR and the incidence of hypertension, addressing gaps in the current literature [19]. The conceptual framework of our study is illustrated in Fig. S1, which outlines the proposed pathway linking METS-IR, metabolic abnormalities, vascular dysfunction, and hypertension prevalence and incidence. We conducted a seven-year follow-up on a cohort of participants who were free from hypertension at baseline. During the follow-up period, we observed a significant association between elevated METS-IR levels and the occurrence of new-onset hypertension, establishing METS-IR as a reliable predictor of hypertension, with a progressive increase in risk over time. This finding highlights the potential of METS-IR in early identification of high-risk populations, providing valuable insights for clinical practice.

In terms of study design, we utilized data from a large, nationally representative Chinese database, conducting a comprehensive cross-sectional analysis. At the start of the analysis, participants underwent thorough health examinations, including blood pressure measurements, height and weight assessments, and laboratory blood tests. Our results indicated a strong correlation between elevated METS-IR levels and the prevalence of hypertension. The unadjusted OR = 3.27 (95% CI: 2.78, 3.85, P < 0.0001), and even after adjusting for multiple traditional cardiovascular risk factors, the OR remained significant (adjusted OR = 3.48, 95% CI: 2.87, 4.21, P < 0.0001), suggesting that METS-IR may play a pivotal role in the pathogenesis of hypertension.

The selection of subgroup variables was informed by both established pathophysiological mechanisms and prior empirical evidence suggesting potential effect modification in the association between metabolic dysfunction and hypertension [20]. Age and sex are fundamental biological variables that influence vascular reactivity, hormonal regulation, and metabolic responses. Behavioral factors such as smoking and alcohol consumption have been shown to alter insulin sensitivity and endothelial function, thereby modifying cardiometabolic risk. Moreover, HbA1c, BMI, and dyslipidemia are key indicators of metabolic status and may interact with METS-IR to influence hypertension risk. Stratified analyses based on these covariates thus provide a more nuanced understanding of differential susceptibility across population subgroups and contribute to the identification of high-risk individuals who may benefit from more targeted preventive strategies.

Moreover, in the longitudinal analysis of participants without hypertension at baseline, after seven years of follow-up, a significant increase in the incidence of new-onset hypertension was observed among those with higher METS-IR levels. The OR for incident hypertension after seven years was 1.71 (95% CI: 1.38, 2.12, P < 0.0001). This result underscores the importance of continuously monitoring METS-IR in patients with known metabolic risk factors, particularly for the prevention of hypertension and its associated complications. This finding is consistent with a retrospective cohort study, which emphasized that persistent insulin resistance is a critical driver of long-term cardiovascular outcomes [21]. Therefore, clinicians should consider regular METS-IR assessments in patients with elevated metabolic risk to prevent the development of hypertension and its complications.

Beyond individual-level risk prediction, the findings underscore the broader public health relevance of METS-IR [22]. As the prevalence of obesity and metabolic syndrome continues to rise globally, particularly in low- and middle-income settings, METS-IR may serve as a cost-effective, integrative tool to identify individuals at elevated hypertension risk [23]. Incorporating METS-IR assessments into routine health screening—especially for populations with high metabolic burden—could improve early intervention strategies, optimize resource allocation, and ultimately reduce the population-level burden of cardiovascular diseases.

Our study found that METS-IR is significantly associated with both the prevalence and incidence of hypertension, highlighting the critical role of metabolic dysfunction in hypertension development [24]. We recommend using METS-IR for hypertension risk assessment to guide personalized prevention strategies. Additionally, policymakers should consider incorporating METS-IR screening into routine health checks, particularly in regions with high obesity rates, to identify high-risk individuals early and reduce the incidence of hypertension and its complications. Promoting healthy lifestyles and early interventions are essential strategies for reducing the burden of cardiovascular disease [25].

As an emerging metabolic index, METS-IR shows strong potential in the field of hypertension risk assessment. Continued investigation across various demographic and clinical contexts is essential to fully establish its utility and to guide its integration into clinical and public health frameworks [26]. Our study contributes new insights into the application of METS-IR and provides a basis for future research in this area.

The strengths of this study lie in the use of a large, nationally representative database and the combination of cross-sectional and longitudinal analyses, enhancing the robustness and generalizability of the findings. However, several limitations should be noted. First, reliance on self-reported data (e.g., smoking and drinking status) may introduce reporting bias, affecting the accuracy of certain variables. Second, the diagnosis of hypertension was based primarily on self-reported physician diagnosis rather than standardized blood pressure measurements over multiple visits. While this reflects the structure of the CHARLS dataset, it may introduce non-differential misclassification, potentially underestimating the association between METS-IR and hypertension. Nonetheless, prior studies have demonstrated acceptable validity of self-reported hypertension in Chinese adults. We also adjusted for covariates and conducted sensitivity analyses to mitigate this bias. Third, although multiple confounders were included, residual confounding cannot be ruled out, limiting causal inference. Fourth, the absence of repeated METS-IR measurements precluded evaluation of temporal changes. Fifth, multiple subgroup analyses were conducted without correction for multiple comparisons, which may increase the risk of Type I error. These findings should thus be interpreted cautiously. Finally, as the study focuses on middle-aged and older Chinese individuals, generalizability to other populations may be limited. Future research should incorporate repeated measurements and diverse cohorts to validate these results.

Conclusion

Higher METS-IR is significantly associated with a higher prevalence and incidence of hypertension in middle-aged and older Chinese adults. This association suggests that METS-IR may be a useful marker for identifying individuals at increased risk of hypertension. Further research is needed to assess the clinical utility of METS-IR for risk stratification and to evaluate the effectiveness of interventions targeting individuals with elevated METS-IR.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Weicheng Lai: Writing– review & editing, Writing − original draft, Validation, Supervision, Project administration, Methodology. Xiao Chen: Writing − review & editing, Writing– original draft, Validation, Supervision, Project administration, Methodology. Lerui Wang: Writing − review &editing, Writing − original draft, Visualization, Validation, Investigation, Formal analysis. Shenshen Du: Writing − review & editing, Validation, Investigation, Formal analysis. Yupeng Wu: Writing − review & editing, Validation, Investigation, Formal analysis. Liangyan Wu: Writing − review & editing, Validation, Investigation, Formal analysis. Huayang Qin: Writing − review & editing, Validation, Investigation, Formal analysis. Xin Li: Writing − review & editing, Validation, Investigation, Formal analysis. Liangxiu Wu: Writing − review & editing, Writing − original draft, Resources, Project administration, Formal analysis, Data curation, Conceptualization. Boda Zhou: Writing − review & editing, Writing − original draft, Resources, Project administration, Formal analysis, Data curation, Conceptualization.

Funding

This work was supported by Elderly Health Research Project from the Jiangsu Provincial Health Commission (LKM2023029), Beijing Municipal Health Commission (2024-3-034), Undergraduate Education Teaching Reform Project of Tsinghua University (ZY01_02), Student Research Training program of Tsinghua University (2511T0682).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The CHARLS protocol was approved by the Biomedical Ethics Review Committee of Peking University (IRB00001052-11014) and the Institutional Review Board of the National School of Development at Peking University (IRB00001052-11015), with informed consent obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.