Abstract
Background
Although dysglycemia and dyslipidemia contribute to hypertension, the role of stress hyperglycemia ratio (SHR), a dynamic glycemic marker, and its interaction with lipids remain unclear. Currently, the independent and synergistic effects of these factors on hypertension in older adults have not been fully elucidated in a large cohort. This study investigated the SHR-lipid interplay and quantified the mediating pathways, addressing a critical gap in understanding the metabolic drivers of hypertension.
Methods
The analysis included 4,546 adults (aged ≥ 45 years, normotensive) from the 2011 to 2015 China Health and Retirement Longitudinal Study. Missing data were subjected to multiple imputations. Cox models were used to assess the association of SHR with incident hypertension, and Kaplan–Meier curves depicted disparities in risk accumulation across SHR quartiles. Restricted cubic splines were used to evaluate the dose–response relationships. The subgroup analyses included age, sex, smoking status, and comorbidities. Causal mediation analysis measured lipid-mediated (total cholesterol [TC] and triglyceride [TG]) effects. Sensitivity analysis utilized methodologies including the exclusion of diabetic patients and the implementation of interval-censored COX multivariate analysis.
Results
Over the 4-year follow-up period, 1,717 participants (37.8%) developed hypertension. Multivariate Cox proportional hazards regression analyses indicated a significant increase in the risk of hypertension associated with elevated SHR, with the highest SHR quartile exhibiting a 16% greater risk than the lowest. Restricted cubic spline models corroborated a linear dose–response relationship between increments in SHR and hypertension risk. The Kaplan–Meier survival curves showed the highest cumulative incidence of hypertension in the highest SHR quartile group. Subgroup analyses stratified by age, sex, race, smoking status, and comorbidities, along with sensitivity analyses excluding individuals with diabetes, confirmed the robustness of the associations. Causal mediation models further revealed that TC and TG partially mediated the relationship between SHR and hypertension, suggesting that metabolic dysregulation may be a contributing mechanism.
Conclusions
Elevated SHR are independently associated with a risk of hypertension in older adults, which is partially mediated by lipids. Integrating SHR into routine metabolic assessments and lipid management may enhance prevention of hypertension in high-risk groups.
Graphical abstract
Supplementary Information
The online version contains supplementary material available at 10.1186/s12944-025-02681-9.
Keywords: Stress hyperglycemia ratio, Hypertension, Cohort, Mediation analysis, CHARLS
Research insights
What is currently known about this topic?
Over 40% of Chinese adults aged ≥ 50 years have hypertension, with dyslipidemia and glucose dysregulation as modifiable drivers.
Static markers (HbA1c, fasting glucose, TG, TC) correlate with hypertension but do not adequately reflect acute metabolic stress.
Acute glycemic variability exacerbates endothelial dysfunction, yet its role in hypertension development remains unquantified in aging populations.
What is the key research question?
This study investigated whether SHR, a dynamic marker of glycemic status, is independently associated with the onset of hypertension among middle-aged and older Chinese adults. Additionally, it examined the extent to which lipid levels mediate this association.
What is new?
This study represents the first large-scale cohort investigation to identify SHR as an independent risk factor for hypertension, revealing a linear dose–response relationship.
Partial lipid mediation (TC and TG) elucidated mechanistic interplay between acute glycemic stress and vascular outcomes.
Asian-specific insights from CHARLS cohort, bridging gaps in global hypertension guidelines by integrating dynamic glycemic-lipid metrics.
How might this study influence clinical practice?
SHR-based risk stratification and lipid-lowering interventions (e.g., statins and dietary adjustments) may optimize hypertension prevention in aging populations, particularly in regions with high diabetes comorbidity.
Introduction
Hypertension affects more than 1.3 billion adults globally, with a prevalence exceeding 30% in high-income countries and rapidly increasing among aging populations worldwide [1]. In China, the impact is particularly pronounced, with over 45% of adults aged 50 years and older experiencing hypertension, which contributes to 40% of stroke-related deaths and 25% of cardiovascular mortality within this demographic [2–4]. The early identification of modifiable metabolic precursors, such as glucose variability and lipid dysregulation, is crucial for delaying the onset of hypertension. However, traditional risk models inadequately account for dynamic biomarkers such as the stress hyperglycemia ratio (SHR), which may indicate transient metabolic disturbances preceding chronic blood pressure elevation [5, 6].
SHR, the acute-to-chronic glycemic ratio (e.g., fasting glucose to glycated hemoglobin [HbA1c]-derived average glucose), is a validated indicator of acute cardiovascular outcomes [7–9]. Although its relationship with new-onset hypertension, a complex condition frequently associated with metabolic dysregulation, has not been explored, epidemiological data indicate common pathways linking glycemic variability, lipid abnormalities, and elevated blood pressure [10, 11]. Previous studies have primarily focused on specific static indicators, such as the triglyceride-glucose (TyG) index coupled with the TyG body mass index (BMI), which are primarily used to assess insulin resistance and generalized obesity [12], whereas measures such as the cardiometabolic index and body shape index are focused on central obesity and visceral adiposity [13, 14]. In contrast, SHR specifically reflects acute glycemic stress, a transient yet pathophysiologically significant condition that may contribute to endothelial dysfunction and vascular inflammation [15, 16]. The interplay between this stress and lipid disturbances (e.g., elevated total cholesterol [TC] and triglyceride [TG] levels) promotes the development of hypertension [17, 18]. It has been posited that SHR is independently correlated with incident hypertension in middle-aged and older Chinese adults, with TC and TG serving as partial mediators of this relationship.
Data from the China Health and Retirement Longitudinal Study (CHARLS, 2011–2015), a nationally representative prospective cohort study, were used. A longitudinal analysis was then conducted to evaluate the association between baseline SHR and incident hypertension and to quantify the mediating effects of TC and TG using causal mediation models. By harnessing extensive epidemiological data, this study elucidated the interactions among acute glycemic stress, lipid metabolism, and the development of hypertension, thereby providing valuable insights for early risk stratification.
Methods
Study participants
The China Health and Retirement Longitudinal Study (CHARLS), initiated in 2011, is an ongoing prospective cohort study encompassing 17,708 participants from 28 provinces across mainland China [19]. A multistage stratified probability-proportional-to-size sampling method was used to ensure national representation. The eligibility criteria included individuals aged ≥ 45 years and their spouses. Written informed consent was obtained from all participants. Follow-ups occurred every 2–3 years until 2020. The 2011 baseline survey achieved a response rate of 80.5%. This substantially mitigated potential selection bias and enhanced cohort representativeness. Trained personnel collected anthropometric measurements and venous blood samples. Ethics approval for the CHARLS protocol was granted by the Ethics Review Committee of Peking University (IRB00001052-11015).
In this study, the initial survey conducted in 2011 was used as the baseline, followed by two follow-up surveys conducted in 2013–2014 and 2015–2016. These follow-up surveys were selected because they included blood pressure measurements obtained during physical examinations. The final cohort consisted of 4,546 individuals who were consistently monitored throughout the 4 years. During this period, 11,687 participants were excluded based on the following criteria: (1) incomplete baseline data, including missing measurements of hypertension, fasting blood glucose, and glycated hemoglobin (N = 5,960); (2) diagnosis of hypertension (N = 5,129); (3) availability of postprandial blood glucose data (N = 598); and (4) missing data for waves 2 and 3 (N = 827). The missing covariate data were subjected to multiple imputations. Subsequently, the participants were categorized into quartiles according to the SHR. The data processing procedure is illustrated in Fig. 1.
Fig. 1.
Participant inclusion flowchart (China health and retirement longitudinal study [CHARLS])
Exposures
The serum hyperglycemia ratio was calculated using the formula: 
[20]. Fasting venous blood samples were collected by medical personnel from the Chinese Center for Disease Control and Prevention in accordance with established protocols. Subsequent assays were conducted at the Youanmen Center for the Clinical Laboratory of the Capital Medical University. Plasma glucose levels were assessed using an enzymatic colorimetric method, and HbA1c levels were measured using boronate-affinity high-performance liquid chromatography. The coefficients of variation for the blood marker measurements were 0.9% for glucose and 1.9% for HbA1c.
Outcome
Hypertension was defined as systolic blood pressure (SBP) of ≥ 140 mmHg or diastolic blood pressure (DBP) ≥ 90 mmHg [1], or current use of antihypertensive medication, as ascertainable from clinical records or self-reports. Blood pressure was measured using the Omron HEM-7112 blood pressure monitor (Omron Healthcare Co., Ltd., Dalian, China). The participants sat with their feet flat, left arm resting, palm up, and the cuff directly on the skin, half an inch above their elbow. An air tube is placed in the middle of each arm. The interviewer pressed the start, and the cuff was inflated, displaying the SBP, DBP, and pulse before deflating. Readings and times were recorded. An interval of 45–60-s was observed before measurements were repeated. Blood pressure readings were acquired three times from the left arm. The mean of three consecutive readings was defined as the baseline blood pressure value [19].
Covariates
Covariates encompassed both categorical and continuous variables. Categorical variables obtained through questionnaires included sex, residence, marital status, race, educational level, smoking status, and alcohol consumption. Clinically diagnosed conditions (via self-report or medical records) included diabetes, heart disease, cancer, stroke, arthritis, kidney disease, liver disease, and asthma. Continuous variables included self-reported age, sleep duration, and mental health status. Blood assay measurements revealed blood urea nitrogen (BUN), uric acid (UA), hemoglobin (HB), low-density lipoprotein (LDL), high-density lipoprotein (HDL), creatinine, and cystatin C levels. Waist circumference (WC) and BMI were measured. The selection of chronic diseases as covariates was based on their well-established associations with hypertension-relevant conditions. Detailed protocols are available at: https://charls.charlsdata.com/index/zh-cn.html.
Statistical analysis
The participants were divided into four groups based on the SHR quartiles. Continuous variables were reported as mean (standard deviation) or median (interquartile range). Frequencies and percentages (n, %) were used to characterize categorical variables. Statistical analysis of baseline characteristics involved one-way analysis of variance and the Kruskal–Wallis test for normally and non-normally distributed continuous variables, respectively. The chi-squared test was used for categorical variables. Missing covariate values were subjected to multiple imputation (five iterations). Supplemental Table 1 provides a detailed account of the missing values. Following the evaluation and confirmation of the proportional hazards assumption via analysis of Schoenfeld residuals (Supplemental Table 2), multivariate Cox proportional hazard regression models were applied to calculate the hazard ratios (HR) and 95% confidence intervals (95% CIs) for the association between SHR and hypertension. Model 1 was not adjusted for. Model 2 was adjusted for demographic variables and lifestyle factors, such as sex, age, race, marital status, residence, educational status, sleep duration, smoking status, drinking status, WC, and BMI. Model 3 also integrated chronic diseases, including cancer, heart disease, stroke, arthritis, liver disease, kidney disease, asthma, and mental health. Model 4 incorporated serum biomarkers including creatinine, BUN, UA, HB, HDL, LDL, and cystatin C levels.
Table 1.
Baseline characteristics of the study participants stratified by SHR quartile
| Variables | Total (N = 4546) | Q1 (N = 1137) < 0.92 | Q2 (N = 1136) 0.92–1.01 | Q3 (N = 1135) 1.01–1.12 | Q4 (N = 1138) > 1.12 | P-value |
|---|---|---|---|---|---|---|
| Demographic | ||||||
| Sex, n (%) | 0.151 | |||||
| Female | 2369 (52.1) | 614 (54.0) | 601 (52.9) | 592 (52.2) | 562 (49.4) | |
| Male | 2177 (47.9) | 523 (46.0) | 535 (47.1) | 543 (47.8) | 576 (50.6) | |
| Age, y | 57.5 ± 8.6 | 57.9 ± 8.7 | 57.6 ± 8.7 | 57.2 ± 8.5 | 57.4 ± 8.4 | 0.283 |
| Race, n (%) | 0.670 | |||||
| Non-Han | 289 (7.1) | 79 (7.7) | 67 (6.6) | 76 (7.4) | 67 (6.5) | |
| Han | 3806 (92.9) | 948 (92.3) | 950 (93.4) | 952 (92.6) | 956 (93.5) | |
| Marital status, n (%) | 0.609 | |||||
| No | 393 (8.6) | 102 (9.0) | 92 (8.1) | 92 (8.1) | 107 (9.4) | |
| Yes | 4153 (91.4) | 1035 (91) | 1044 (91.9) | 1043 (91.9) | 1031 (90.6) | |
| Residence, n (%) | 0.157 | |||||
| Urban community | 1551 (34.1) | 379 (33.3) | 362 (31.9) | 405 (35.7) | 405 (35.6) | |
| Rural village | 2995 (65.9) | 758 (66.7) | 774 (68.1) | 730 (64.3) | 733 (64.4) | |
| Educational status, n (%) | 0.062 | |||||
| Illiterate | 2062 (45.4) | 557 (49.0) | 532 (46.8) | 491 (43.4) | 482 (42.4) | |
| Primary school | 966 (21.3) | 231 (20.3) | 228 (20.1) | 245 (21.6) | 262 (23.1) | |
| Junior high school | 996 (21.9) | 224 (19.7) | 247 (21.7) | 255 (22.5) | 270 (23.8) | |
| High school or above | 517 (11.4) | 125 (11.0) | 129 (11.4) | 141 (12.5) | 122 (10.7) | |
| Lifestyle | ||||||
| Sleep duration, h | 6.4 ± 1.8 | 6.3 ± 1.9 | 6.4 ± 1.9 | 6.4 ± 1.8 | 6.5 ± 1.7 | 0.150 |
| Smoking status, n (%) | 0.696 | |||||
| Never a | 2733 (61.6) | 675 (61.0) | 689 (62.1) | 699 (62.9) | 670 (60.5) | |
| Past b | 346 (7.8) | 79 (7.1) | 94 (8.5) | 82 (7.4) | 91 (8.2) | |
| Current c | 1357 (30.6) | 353 (31.9) | 327 (29.5) | 331 (29.8) | 346 (31.3) | |
| Drinking status, n (%) | < 0.001 | |||||
| Never | 2632 (59.5) | 701 (63.3) | 648 (58.5) | 646 (58.3) | 637 (58.0) | |
| Past | 324 (7.3) | 79 (7.1) | 103 (9.3) | 83 (7.5) | 59 (5.4) | |
| Current | 1466 (33.2) | 327 (29.5) | 356 (32.2) | 380 (34.3) | 403 (36.7) | |
| Health Status | ||||||
| Diabetes, n (%) | < 0.001 | |||||
| No | 3965 (87.8) | 1064 (94.3) | 1069 (95.1) | 1049 (92.8) | 783 (69.1) | |
| Yes | 550 (12.2) | 64 (5.7) | 55 (4.9) | 81 (7.2) | 350 (30.9) | |
| Cancer, n (%) | 0.967 | |||||
| No | 4485 (99.1) | 1121 (99.2) | 1121 (99.1) | 1122 (99.0) | 1121 (99.0) | |
| Yes | 41 (0.9) | 9 (0.8) | 10 (0.9) | 11 (1.0) | 11 (1.0) | |
| Heart Diseases, n (%) | 0.523 | |||||
| No | 4126 (91.2) | 1026 (90.7) | 1034 (91.3) | 1041 (92.2) | 1025 (90.6) | |
| Yes | 397 (8.8) | 105 (9.3) | 98 (8.7) | 88 (7.8) | 106 (9.4) | |
| Stroke, n (%) | 0.856 | |||||
| No | 4478 (98.7) | 1119 (98.6) | 1119 (98.8) | 1117 (98.6) | 1123 (98.9) | |
| Yes | 58 (1.3) | 16 (1.4) | 14 (1.2) | 16 (1.4) | 12 (1.1) | |
| Arthritis, n (%) | 0.170 | |||||
| No | 3017 (66.5) | 724 (63.8) | 758 (66.8) | 763 (67.3) | 772 (68.0) | |
| Yes | 1522 (33.5) | 410 (36.2) | 377 (33.2) | 371 (32.7) | 364 (32.0) | |
| Liver Diseases, n (%) | 0.064 | |||||
| No | 4349 (96.2) | 1082 (95.9) | 1076 (95.1) | 1095 (96.9) | 1096 (97.0) | |
| Yes | 170 (3.8) | 46 (4.1) | 55 (4.9) | 35 (3.1) | 34 (3.0) | |
| Kidney Diseases, n (%) | 0.338 | |||||
| No | 4250 (94.0) | 1049 (93) | 1066 (94.2) | 1059 (94) | 1076 (94.8) | |
| Yes | 271 (6.0) | 79 (7.0) | 66 (5.8) | 67 (6.0) | 59 (5.2) | |
| Asthma, n (%) | 0.333 | |||||
| No | 4339 (95.8) | 1082 (95.5) | 1080 (95.2) | 1082 (95.8) | 1095 (96.6) | |
| Yes | 192 (4.2) | 51 (4.5) | 55 (4.8) | 48 (4.2) | 38 (3.4) | |
| Mental health d, score | 8.3 ± 6.2 | 9.2 ± 6.5 | 8.3 ± 6.1 | 8.1 ± 6.3 | 7.7 ± 6.0 | < 0.001 |
| Anthropometric Measures | ||||||
| WC, cm | 82.1 ± 11.7 | 81.5 ± 10.5 | 81.5 ± 12.5 | 83.0 ± 11.9 | 82.6 ± 11.6 | 0.006 |
| BMI, kg/m2 | 22.5 (20.4, 24.8) | 22.1 (20.1, 24.5) | 22.4 (20.4, 24.8) | 22.9 (20.7, 25.1) | 22.6 (20.5, 24.7) | < 0.001 |
| BUN, mg/dl | 15.67 ± 4.58 | 15.75 ± 4.84 | 15.49 ± 4.30 | 15.36 ± 4.24 | 16.07 ± 4.87 | 0.001 |
| Creatinine, mg/dl | 0.76 ± 0.20 | 0.77 ± 0.26 | 0.75 ± 0.17 | 0.76 ± 0.17 | 0.77 ± 0.18 | 0.282 |
| Cystatin C, mg/dl | 0.98 ± 0.25 | 1.02 ± 0.32 | 0.98 ± 0.21 | 0.97 ± 0.21 | 0.96 ± 0.23 | < 0.001 |
| UA, mg/dl | 4.28 ± 1.17 | 4.26 ± 1.13 | 4.22 ± 1.12 | 4.26 ± 1.20 | 4.39 ± 1.22 | 0.002 |
| FBG, mg/dl | 106.4 ± 29.4 | 91.8 ± 14.7 | 100.5 ± 14.3 | 106.4 ± 22.5 | 126.6 ± 43.5 | < 0.001 |
| Glycated hemoglobin, % | 5.24 ± 0.74 | 5.49 ± 0.64 | 5.26 ± 0.51 | 5.13 ± 0.74 | 5.07 ± 0.95 | < 0.001 |
| SHR | 1.03 ± 0.20 | 0.83 ± 0.08 | 0.97 ± 0.03 | 1.06 ± 0.03 | 1.28 ± 0.20 | < 0.001 |
| HB, g/dl | 14.32 ± 2.20 | 14.28 ± 2.26 | 14.34 ± 2.12 | 14.36 ± 2.20 | 14.32 ± 2.19 | 0.852 |
| TC, mg/dl | 191.4 ± 37.4 | 189.5 ± 36.4 | 193.0 ± 35.4 | 190.8 ± 37.5 | 192.5 ± 40.3 | 0.107 |
| TG, mg/dl | 97.3 (70.8, 140.7) | 88.5 (68.1, 124.8) | 95.6 (69.9, 133.6) | 97.3 (71.7, 139.8) | 108.9 (75.2, 170.6) | < 0.001 |
| HDL, mg/dl | 52.4 ± 15.4 | 53.5 ± 15.0 | 53.4 ± 15.1 | 52.7 ± 15.1 | 49.9 ± 16.1 | < 0.001 |
| LDL, mg/dl | 116.0 ± 33.3 | 116.5 ± 31.7 | 118.7 ± 31.0 | 115.6 ± 33.3 | 113.0 ± 36.8 | < 0.001 |
Data are presented as the mean (SD), (IQR) or number (%), as appropriate
SHR stress hyperglycemia ratio, WC waist circumference, BMI body mass index, BUN blood urea nitrogen, UA uric acid, FBG fasting blood glucose, HB Hemoglobin, TC total cholesterol, TG triglyceride, HDL High-Density Lipoprotein, LDL low-density lipoprotein, SD standard deviation, IQR Interquartile Range
aDefined as individuals with no history of smoking, including cigarette smoking, pipe use, or chewing tobacco, where smoking is classified as having consumed more than 100 cigarettes or an equivalent amount of tobacco in their lifetime
bDefined as individuals with a history of smoking cigarettes, smoking a pipe, or chewing tobacco, either currently or in the past, with smoking classified as having consumed more than 100 cigarettes in their lifetime
cDefined as an individual with a history of smoking cigarettes, using a pipe, or chewing tobacco, with smoking defined as having consumed more than 100 cigarettes or equivalent, but who has now quit
dDefined as the CES-D scale assesses depressive symptoms over the past week via 10 items (8 negative, 2 positive). Scores range from 0 to 30, with higher scores indicating more severe depression
A restricted cubic spline (RCS) was used to explore the potential linear association between SHR and hypertension after adjusting for the variables in Model 4. Interaction and subgroup analyses were performed using Cox proportional hazards regression models. Adjustments for these analyses incorporated the variables from Model 4. Kaplan–Meier curves were used to illustrate variations in the prevalence of hypertension among distinct SHR groups over the same period.
A mediation analysis was conducted to investigate the potential contribution of dysregulated lipid metabolism, precipitated by variations in blood glucose levels, to the onset of hypertension. Specifically, the bootstrap method with 1,000 resamples was used to estimate the mediating effects of TG and TC on the association between SHR and development of hypertension.
Five sensitivity analyses were performed to ensure the robustness of the findings. To address potential inaccuracies in the timing of new-onset hypertension during the follow-up period, an interval-censored Cox model was used to evaluate the association between the SHR and new-onset hypertension. Subsequently, to validate the association assessment after converting the SHR into a categorical variable, the analyses were replicated using SHR quintiles. To mitigate bias arising from extreme values, the SHR distribution was trimmed at the 0.5th and 99.5th percentiles before modeling. Moreover, to ensure the robustness of the findings, particularly considering that patients with diabetes may experience elevated blood glucose levels and unstable glucose control, the data were reanalyzed using multivariate Cox proportional hazard regression models after excluding the cohort with diabetes. Finally, to further confirm the robustness of the results, the effect of unmeasured confounders on the relationship between SHR and hypertension was quantified by calculating the E-values, which provided an estimate of the strength of the confounding factors required to explain the observed association.
Statistical analyses were performed using R Statistical Software (v4.2.3; R Foundation for Statistical Computing; http://www.R-project.org) and a Free Statistical Analysis Platform (v2.1.1; Beijing Free Clinical Medical Technology Co., Ltd.). Descriptive statistics were calculated for the entire cohort. A two-tailed test was used to determine statistical significance, which was set at P < 0.05 considered significant.
Results
Characteristics of the population according to the SHR
A total of 4,546 individuals were included in this study, exhibiting a mean age of 57.5 ± 8.6 years. The cohort's composition featured 2,177 (45.2%) males, and the average SHR stood was 1.03 ± 0.20. Participant characteristics categorized by SHR quartiles are presented in Table 1. Higher SHRs were typically associated with alcohol consumption, diabetes, poor mental health, as well as lower cystatin C, HDL, and LDL levels and higher WC, BMI, BUN, UA, and TG levels.
Association between the SHR and risk of hypertension
Multivariate adjustment revealed a significant association between SHR and hypertension (Table 2). An increment of one unit in SHR corresponded to a 48% greater prevalence of hypertension (HR = 1.48, 95% CI: 1.17–1.88, P = 0.001). Individuals in the highest SHR quartile exhibited a 16% higher risk of hypertension than those in the lowest SHR quartile (P = 0.028). Figure 2 illustrates the multivariate-adjusted RCS analysis examining the link between the SHR and hypertension risk. A positive linear relationship was observed (P for nonlinearity > 0.05).
Table 2.
Association of SHR with Hypertension risk (Multiple imputation)
| Categories | Model 1 | Model 2 | Model 3 | Model 4 | ||||
|---|---|---|---|---|---|---|---|---|
| HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P-value | |
| SHR | 1.50 (1.19–1.89) | 0.001 | 1.45 (1.16–1.83) | 0.001 | 1.44 (1.14–1.81) | 0.002 | 1.48 (1.17–1.88) | 0.001 |
| SHR Quartile | ||||||||
| Q1 | 1(Ref) | 1(Ref) | 1(Ref) | 1(Ref) | ||||
| Q2 | 1.02 (0.89–1.17) | 0.779 | 1.08 (0.92–1.27) | 0.341 | 1.02 (0.89–1.17) | 0.759 | 1.01 (0.88–1.16) | 0.843 |
| Q3 | 1.03 (0.90–1.18) | 0.649 | 0.98 (0.84–1.16) | 0.846 | 1.01 (0.88–1.16) | 0.843 | 1.00 (0.87–1.15) | 0.947 |
| Q4 | 1.18 (1.04–1.35) | 0.013 | 1.22 (1.04–1.42) | 0.014 | 1.16 (1.01–1.32) | 0.033 | 1.16 (1.02–1.33) | 0.028 |
| P for trend | 0.015 | 0.026 | 0.042 | 0.040 | ||||
Model 1: not adjusted
Model 2: adjusted for sex, age, race, marital status, residence, educational status, sleep duration, smoking status, drinking status, waist circumference, BMI
Model 3: adjusted for model 2, additionally adjusted for cancer, heart diseases, stroke, arthritis, liver diseases, kidney diseases, asthma, and mental health
Model 4: adjusted for model 3, additionally adjusted for creatinine, BUN, UA, HB, HDL, LDL, cystatin C
SHR Stress hyperglycemia ratio, WC Waist circumference, BMI Body mass index, BUN Blood urea nitrogen, UA Uric acid, FBG Fasting blood glucose, HB Hemoglobin, HDL High-Density Lipoprotein, LDL Low-density lipoprotein, HR Hazard ratios, CI Confidence interval
Fig. 2.
The dose-response relationship between SHR and the risk of hypertension. SHR: stress hyperglycemia ratio. This hazard ratio plot, constructed with restricted cubic splines, depicts the association between SHR and the risk ratio of hypertension. The knots for the splines are positioned at the 5th, 35th, 65th, and 95th percentiles of SHR. The reference point is set at a SHR value of 1.008. The solid red line indicates the estimated hazard ratio, with the red shaded area representing the 95% confidence interval. The dashed line at a hazard ratio of 1 serves as the reference for no effect. The overall association is statistically significant (P = 0.025), while the P-value for nonlinearity is 0.381. The plot displays data within the range of 0.95% to 99.60% of the distribution. The analysis is adjusted for covariates as per Model 4
Subgroup analysis and Kaplan–Meier curves
Subgroup analyses assessed potential variations in the SHR-hypertension association across demographic and clinical categories (Fig. 3). Overall, 1,769 (36.5%) patients developed hypertension. Interactions were not significant across subgroups stratified by age, sex, race, smoking status, or history of heart disease (P > 0.05). Figure 4 shows the Kaplan–Meier curves depicting the cumulative incidence of hypertension across the entire study population. Notably, individuals with the highest SHR had a significantly higher prevalence of hypertension (P = 0.03).
Fig. 3.
Subgroup Analysis of Hypertension Risk Using Multivariable Cox Regression Model. SHR: stress hyperglycemia ratio, WC: waist circumference, BMI: body mass index, BUN: blood urea nitrogen, UA: uric acid, FBG: fasting blood glucose, HB: Hemoglobin, HDL: High-Density Lipoprotein, LDL: low-density lipoprotein, HR: hazard ratios; CI: confidence interval. Forest plot displaying the adjusted HR and 95% CI for hypertension across various subgroups. The analysis was conducted using a multivariable Cox regression model, adjusting for marital status, residence, educational status, sleep duration, drinking status, WC, BMI, cancer, stroke, arthritis, liver diseases, kidney diseases, asthma, mental health, creatinine, BUN, UA, HB, HDL, LDL, and cystatin C. The overall crude and adjusted models included 4844 subjects with 1769 cases of hypertension, presenting an adjusted HR of 1.48 (95% CI: 1.17–1.88). Subgroup analyses by age, race, sex, smoking status, and presence of heart diseases are shown. The P-values for interaction indicate no significant differences in the effect of hypertension across subgroups (P for interaction > 0.05 for all). The diamond represents the overall effect size, and the horizontal lines are the 95% CIs
Fig. 4.
Kaplan-Meier Analysis of Cumulative Hypertension Incidence by Quartile Groups. Kaplan–Meier curves showing the non-disease prevalence rate over time for four quartile groups (Q1 to Q4). The number of individuals at risk for each quartile group at non-disease status is indicated at specific time points (20, 30, 40, 50 months). The log-rank test for intergroup comparison yields a P-value of 0.03, suggesting a statistically significant difference in non-disease prevalence among the quartile groups. Notably, the highest SHR group (Q4) exhibits the highest cumulative hypertension prevalence rate
Mediation analysis
Figure 5 illustrates the potential of TC and TG (lipid parameters) to mediate the relationship between SHRs and hypertension. In the fully adjusted analyses, TG demonstrated a mediation proportion of 12.76% (P = 0.04), whereas TC exhibited a mediation proportion of 14.84% (P = 0.028). These findings indicated that TG and TC may partially mediate SHR's influence on human metabolic processes, thereby promoting hypertension.
Fig. 5.
Mediation Analysis of the Relationship Between SHR and Hypertension. SHR: stress hyperglycemia ratio, WC: waist circumference, BMI: body mass index, BUN: blood urea nitrogen, UA: uric acid, FBG: fasting blood glucose, HB: Hemoglobin, TC: total cholesterol, TG: triglyceride, HDL: High-Density Lipoprotein, LDL: low-density lipoprotein, HR: hazard ratios; CI: confidence interval. The pathway diagram illustrates the mediating effects of TG and TC on the relationship between SHR and hypertension risk, adjusted for age, race, sex, marital status, residence, educational status, sleep duration, smoking status, drinking status, WC, BMI, cancer, stroke, arthritis, heart diseases, liver diseases, kidney diseases, asthma, mental health, creatinine, BUN, UA, HB, HDL, LDL, and cystatin C. The proportion of the mediating effect, the P-value of the mediating effect, and the proportion of the mediating effect are all annotated. The mediating effect proportion for TG is 14.84%, and for TC, it is 12.76%, indicating that both have partial mediating effects
Sensitivity analyses
In multifactorial Cox regression analysis performed after interval censoring, no significant alterations were observed (Supplemental Table 3). A one-unit rise in SHR exhibited a 51% higher hypertension prevalence (HR = 1.51, 95% CI: 1.18–1.94, P = 0.001). According to the SHR quartiles, individuals in the highest SHR quartile demonstrated an 18% greater risk of hypertension than those in the lowest SHR quartile (P = 0.021). No significant changes were observed in the SHR quintiles (Supplemental Table 4). Additionally, the findings remained essentially unchanged when the adjusted models with distributions of SHR at the 0.5th and 99.5th percentiles were used (Supplemental Table 5). Moreover, the findings remained robust even after excluding patients with diabetes (Supplemental Table 6). Finally, E-values were computed to evaluate the influence of the unmeasured confounders. The association between SHR and risk of hypertension was robust unless unmeasured confounding factors were associated with both the SHR and risk of hypertension at a relative risk exceeding 1.95 (Supplementary Fig. 1).
Discussion
This extensive population-based cohort study reveals several novel findings. First, an elevated SHR was independently correlated with a higher incidence of hypertension. This association remained significant even after excluding of individuals with diabetes. Second, these associations were independent of variables such as age, sex, ethnicity, and smoking status and exhibited a linear relationship. Furthermore, TC and TG levels served as mediating variables, partially mediating the effects of SHR on the risk of hypertension. These findings imply that fluctuations or stress-induced increases in blood glucose levels increase the risk of hypertension by affecting lipid metabolism.
Previous research on metabolic risk factors for hypertension, including studies focusing on the Chinese population, has left two critical issues unresolved. First, although existing cohort studies have established associations between static glycemic markers (such as fasting blood glucose and HbA1c) or individual lipid parameters (TC and TG) and hypertension [21–25], they have not yet assessed dynamic glycemic stress as quantified by the SHR (a metric that integrates acute and chronic glucose ratios) despite the proven utility of the SHR in predicting acute cardiovascular events. Second, methodological limitations such as reliance on a cross-sectional design or insufficient adjustment for confounding factors have constrained causal inference and obscured the potential mediating pathways. For example, a cross-sectional study conducted in 2019 identified associations between indices such as TyG and hypertension [22]; however, the potential mediating role of lipids was not investigated. Similarly, a study conducted in 2023 failed to consider temporal factors within cohort studies and did not exclude populations with diabetes [23]. Additionally, a study using the Digital Repository of Information and Data (DRYAD) database established an association between lipid levels and the risk of hypertension [24]; however, it relied on single lipid measurements and ignored the longitudinal variability in those measurements. This study contributes to the advancement of this field through three key innovations. (1) By using a prospective design, 1,717 incident cases of hypertension were rigorously identified from the CHARLS cohort, thereby ensuring sufficient statistical power to detect subtle effects. (2) A causal mediation model was utilized to quantify the association between the SHR and incident hypertension via the TC/TG pathway, providing mechanistic insights previously not reported in Chinese cohorts. (3) Robust bias mitigation techniques were implemented, including multiple imputation, interval-censored Cox regression, and sensitivity analyses (e.g., excluding baseline prediabetes). By addressing these limitations, the SHR has been established as a dynamic biomarker that uniquely captures the interplay between acute glycemic stress and lipid dysregulation during the development of hypertension, offering significant implications for risk stratification beyond traditional static markers.
TC and TG partially mediated the association between SHR and the development of hypertension. This relationship is potentially elucidated through interconnected pathways that encompass acute glycemic stress and lipid-induced vascular dysfunction. Elevated SHR indicates transient hyperglycemic events that induce oxidative stress and release of inflammatory cytokines [26–28], thereby impairing endothelial nitric oxide bioavailability and promoting arterial stiffness [29, 30]. Concurrently, acute hyperglycemia enhances hepatic lipogenesis and the mobilization of free fatty acids, leading to increased circulating levels of TG and TC [31, 32]. These lipid abnormalities play significant roles in the development of hypertension through various mechanisms [33, 34]. Hyperlipidemia exacerbates insulin resistance, thereby establishing a detrimental cycle that perpetuates glycemic instability and lipid dysregulation [35]. This synergistic interaction may account for the partial mediation observed in this study, wherein stress induced by SHR initiated vascular damage. In contrast, elevated TC and TG levels maintained and intensified the pathways leading to hypertension. Consequently, SHR may serve as a clinical indicator of metabolic status because acute glycemic fluctuations and chronic lipid imbalances collectively increase the risk of hypertension.
Strengths and limitations
Our study possesses several key strengths, particularly its three innovations, that significantly advance hypertension prevention strategies. (1) SHR, a simple metric derived from routine glucose and HbA1c measurements, was identified as an independent risk factor, enabling cost-effective risk stratification during routine care without the need for specialized testing. (2) By quantifying TG and TC as mediators, modifiable targets were highlighted; lipid-lowering therapies (e.g., statins) or dietary interventions may attenuate the risk of SHR-associated hypertension. (3) Focusing on China’s aging population addresses a critical gap as Western-derived guidelines may inadequately address Asian-specific dyslipidemia patterns. By leveraging a nationally representative cohort and rigorous methods (interval censoring and multiple imputation), the results advocate the integration of SHR and lipid profiling into prevention protocols, particularly in regions with a high diabetes burden. These insights refine risk prediction models and inform dual-target interventions (glycemic variability and lipid control), offering actionable pathways to curb hypertension-related morbidity in aging societies.
This study had some limitations. First, the cohort comprised middle-aged and older Chinese adults, limiting direct extrapolation of the findings to other populations; however, nationally representative sampling enhances the relevance of these findings to similarly aging societies. Second, despite multivariable adjustments and E-value analyses (minimum E-value: 1.95), residual confounding from unmeasured variables (e.g., dietary sodium intake and genetic factors) cannot be excluded. Nevertheless, the E-value suggests that unmeasured confounders would need risk ratios > 1.95 to negate the findings, which is unlikely given the known biological effect sizes. Third, missing data were addressed using multiple imputations; however, selection bias may have persisted because of incomplete follow-up. Sensitivity analyses across the imputed and complete case datasets yielded consistent results, thereby supporting the robustness. Future multi-ethnic cohort studies that incorporate real-time glucose monitoring and genetic data can further validate these findings. Despite these limitations, this study provides novel and methodologically rigorous evidence that SHR-associated hypertension risk is partially mediated by lipids, thereby contributing to the development of targeted hypertension prevention strategies in high-risk populations.
Conclusions
This extensive cohort study robustly established the SHR as a clinically accessible indicator of incident hypertension risk, with its effects partially mediated by lipid dysregulation. These findings directly impact patient care by advocating the integration of SHR into routine screening and emphasizing targeted lipid management (TG and TC) to enhance hypertension prevention efforts in aging populations, particularly in those with cardiometabolic diseases.
Supplementary Information
Supplementary Material 1: Supplemental Table 1. Sample Missing Data Overview
Supplementary Material 2: Supplemental Table 2. Proportional Hazards Test. This table presents the results of the proportional hazards assumption test conducted prior to the multivariable Cox regression analysis
Supplementary Material 3: Supplemental Table 3. Association of SHR with Hypertension risk (interval-censored). Interval-censored multivariate Cox regression analysis was performed on the data
Supplementary Material 4: Supplemental Table 4. Association of SHR with risk of Hypertension (SHR Quintile)
Supplementary Material 5: Supplemental Table 5. Association of SHR with Hypertension risk (Minorized distributions of SHR at the 0.5 and 99.5 percentiles). Adjusted models with distributions of SHR at the 0.5th and 99.5th percentiles were used
Supplementary Material 6: Supplemental Table 6. Association of SHR with Hypertension risk (Exclude diabetic patients). After excluding the diabetic cohort, the data were reanalyzed using a multivariate Cox proportional hazards regression model
Supplementary Material 7: Supplementary Fig. 1 E-value Analysis for Unmeasured Confounding in the Association Between SHR and Hypertension Risk. SHR: stress hyperglycemia ratio. Bias plot illustrates the calculated E-values to assess the potential impact of unmeasured confounding factors on the association between SHR and the risk of hypertension. The solid line represents the risk ratio for the exposure-confounder relationship (RR_UD), while the dashed line indicates the risk ratio for the exposure-disease relationship (RR). The plot shows that the association between SHR and hypertension risk remains robust unless unmeasured confounding factors are associated with both SHR and hypertension risks at a magnitude exceeding 1.95. The point (1.95, 1.95) is marked to highlight this threshold. The risk ratio for the exposure-confounder relationship is plotted against the risk ratio for the exposure-disease relationship, with the curve approaching the dashed line as the magnitude of potential unmeasured confounding increases
Acknowledgements
The authors thank the CHARLS staff, investigators, and participants. Thanks to the Free Statistics team for providing technical assistance and valuable tools for data analysis and visualization. We thank Dr. Liu jie (People’s Liberation Army of China General Hospital, Beijing, China) for helping in this revision.
Abbreviations
- BMI
Body mass index
- BUN
Blood urea nitrogen
- CHARLS
China Health and Retirement Longitudinal Study
- CI
Confidence interval
- DBP
Diastolic blood pressure
- HB
Hemoglobin
- HbA1c
Glycated hemoglobin
- HDL
High-density lipoprotein
- HR
Hazard ratio
- LDL
Low-density lipoprotein
- RCS
Restricted cubic spline
- SBP
Systolic blood pressure
- SHR
Stress hyperglycemia ratio
- TC
Total cholesterol
- TG
Triglycerides
- TyG
Triglyceride-glucose
- UA
Uric acid
- WC
Waist circumference
Authors’ contributions
GSJ: data collection, data analysis, manuscript writing. HBW: perform interval-censored analysis, XXQ: data analysis, manuscript editing. JZ, CYF,YXL: data collection. JPL: project development, manuscript editing. All authors have read and approved this manuscript.
Funding
No funding was received for conducting this study.
Data availability
The data that support the findings of this study are available in China Health and Retirement Longitudinal Study. https://charls.charlsdata.com/pages/data/111/zh-cn.html.
Declarations
Ethics approval and consent to participate
Ethical approval was obtained from the Institutional Review Boards of Peking University (No. IRB00001052-11015) for the CHARLS, and all participants gave informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Mills KT, Stefanescu A, He J. The global epidemiology of hypertension. Nat Rev Nephrol. 2020;16(4):223–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lu J, Lu Y, Wang X, et al. Prevalence, awareness, treatment, and control of hypertension in China: data from 1·7 million adults in a population-based screening study (China PEACE Million Persons Project). Lancet. 2017;390(10112):2549–58. [DOI] [PubMed] [Google Scholar]
- 3.Wang JG, Zhang W, Li Y, Liu L. Hypertension in China: epidemiology and treatment initiatives. Nat Rev Cardiol. 2023;20(8):531–45. [DOI] [PubMed] [Google Scholar]
- 4.Zhao D, Liu J, Wang M, Zhang X, Zhou M. Epidemiology of cardiovascular disease in China: current features and implications. Nat Rev Cardiol. 2019;16(4):203–12. [DOI] [PubMed] [Google Scholar]
- 5.Cheng Y, Gu K, Ji W, Hu Z, Yang Y, Zhou Y. Two-year hypertension incidence risk prediction in populations in the desert regions of Northwest China: prospective cohort study. J Med Internet Res. 2025;27:e68442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Qin C, Peng L, Liu Y, et al. Development and validation of a nomogram-based model to predict primary hypertension within the next year in children and adolescents: retrospective cohort study. J Med Internet Res. 2024;26:e58686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Song Y, Cui K, Yang M, et al. High triglyceride-glucose index and stress hyperglycemia ratio as predictors of adverse cardiac events in patients with coronary chronic total occlusion: a large-scale prospective cohort study. Cardiovasc Diabetol. 2023;22(1):180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chen X, Yang Z, Shi R, Wang X, Li X. Stress hyperglycemia ratio association with all-cause mortality in critically ill patients with coronary heart disease: an analysis of the MIMIC-IV database. Sci Rep. 2024;14(1):29110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Liu S, Wu Z, Yan G, et al. Relationship between stress hyperglycemia ratio and progression of non target coronary lesions: a retrospective cohort study. Diabetol Metab Syndr. 2025;17(1):27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ormazabal V, Nair S, Elfeky O, Aguayo C, Salomon C, Zuñiga FA. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bjornstad P, Eckel RH. Pathogenesis of lipid disorders in insulin resistance: a brief review. Curr Diab Rep. 2018;18(12):127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen Y, Zhong Z, Gue Y, et al. Impact of surrogates for insulin resistance on mortality and life expectancy in primary care: a nationwide cross-sectional study with registry linkage (LIPIDOGRAM2015). Lancet Reg Health Eur. 2024;49:101182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Petrovic A, Jovicic S, Dodevska M, Djordjevic B, Milinkovic N, Ivanovic ND. Effects of specially designed energy-restricted diet on anthropometric parameters and cardiometabolic risk in overweight and obese adults: pilot study. Nutrients. 2024;16(20):3453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ji M, Zhang S, An R. Effectiveness of a body shape index (ABSI) in predicting chronic diseases and mortality: a systematic review and meta-analysis. Obes Rev. 2018;19(5):737–59. [DOI] [PubMed] [Google Scholar]
- 15.Khan AW, Jandeleit-Dahm KAM. Atherosclerosis in diabetes mellitus: novel mechanisms and mechanism-based therapeutic approaches. Nat Rev Cardiol. 2025;22(7):482–496. [DOI] [PubMed]
- 16.Kim DY, Piao J, Hong HS. Substance-P inhibits cardiac microvascular endothelial dysfunction caused by high glucose-induced oxidative stress. Antioxidants. 2021;10(7):1084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Niu Z, Wu Q, Luo Y, et al. Plasma lipidomic subclasses and risk of hypertension in middle-aged and elderly Chinese. Phenomics. 2022;2(5):283–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sangrós FJ, Torrecilla J, Giráldez-García C, et al. Association of general and abdominal obesity with hypertension, dyslipidemia and prediabetes in the PREDAPS Study. Rev Esp Cardiol (Engl Ed). 2018;71(3):170–7. [DOI] [PubMed] [Google Scholar]
- 19.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Roberts GW, Quinn SJ, Valentine N, et al. Relative hyperglycemia, a marker of critical illness: introducing the stress hyperglycemia ratio. J Clin Endocrinol Metab. 2015;100(12):4490–7. [DOI] [PubMed] [Google Scholar]
- 21.Xin F, He S, Zhou Y, Jia X, Zhao Y, Zhao H. The triglyceride glucose index trajectory is associated with hypertension: a retrospective longitudinal cohort study. Cardiovasc Diabetol. 2023;22(1):347 (Published 2023 Dec 15). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bala C, Gheorghe-Fronea O, Pop D, et al. The association between six surrogate insulin resistance indexes and hypertension: a population-based study. Metab Syndr Relat Disord. 2019;17(6):328–33. [DOI] [PubMed] [Google Scholar]
- 23.Li Y, Gui J, Zhang X, et al. Predicting hypertension by obesity- and lipid-related indices in mid-aged and elderly Chinese: a nationwide cohort study from the China health and retirement longitudinal study. BMC Cardiovasc Disord. 2023;23(1):201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Chen S, Cheng W. Relationship between lipid profiles and hypertension: a cross-sectional study of 62,957 Chinese adult males. Front Public Health. 2022;10:895499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Dong X, Han B, Huang W, et al. Association of TyG index with hypertension in Chinese adults: the China Health Examination Collaborative Study (CHEC Study). Asia Pac J Clin Nutr. 2023;32(3):362–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.González P, Lozano P, Ros G, Solano F. Hyperglycemia and oxidative stress: an integral, updated and critical overview of their metabolic interconnections. Int J Mol Sci. 2023;24(11):9352 (Published 2023 May 27). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gu X, Shi Y, Chen X, et al. Isoliquiritigenin attenuates diabetic cardiomyopathy via inhibition of hyperglycemia-induced inflammatory response and oxidative stress. Phytomedicine. 2020;78:153319. [DOI] [PubMed] [Google Scholar]
- 28.Zhao M, Wang S, Zuo A, et al. HIF-1α/JMJD1A signaling regulates inflammation and oxidative stress following hyperglycemia and hypoxia-induced vascular cell injury. Cell Mol Biol Lett. 2021;26(1):40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Minari TP, Pisani LP. Exploring sodium nitrate supplementation in enhancing nitric oxide bioavailability and reducing oxidative stress: implications for blood pressure and endothelial dysfunction in hypertension. Eur J Pharmacol. 2025;999:177702. [DOI] [PubMed]
- 30.Feenstra L, Kutikhin AG, Shishkova DK, et al. Calciprotein particles induce endothelial dysfunction by impairing endothelial nitric oxide metabolism. Arterioscler Thromb Vasc Biol. 2023;43(3):443–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Newman CB, Blaha MJ, Boord JB, et al. Lipid management in patients with endocrine disorders: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2020;105(12): dgaa674. [DOI] [PubMed] [Google Scholar]
- 32.Xiao Y, Yu B, Chao C, et al. Chinese expert consensus on blood lipid management in patients with diabetes (2024 edition). J Transl Intern Med. 2024;12(4):325–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ramírez-Melo LM, Estrada-Luna D, Rubio-Ruiz ME, et al. Relevance of lipoprotein composition in endothelial dysfunction and the development of hypertension. Int J Mol Sci. 2025;26(3): 1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Soppert J, Lehrke M, Marx N, Jankowski J, Noels H. Lipoproteins and lipids in cardiovascular disease: from mechanistic insights to therapeutic targeting. Adv Drug Deliv Rev. 2020;159:4–33. [DOI] [PubMed] [Google Scholar]
- 35.Gepstein V, Weiss R. Obesity as the main risk factor for metabolic syndrome in children. Front Endocrinol (Lausanne). 2019;10: 568. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1: Supplemental Table 1. Sample Missing Data Overview
Supplementary Material 2: Supplemental Table 2. Proportional Hazards Test. This table presents the results of the proportional hazards assumption test conducted prior to the multivariable Cox regression analysis
Supplementary Material 3: Supplemental Table 3. Association of SHR with Hypertension risk (interval-censored). Interval-censored multivariate Cox regression analysis was performed on the data
Supplementary Material 4: Supplemental Table 4. Association of SHR with risk of Hypertension (SHR Quintile)
Supplementary Material 5: Supplemental Table 5. Association of SHR with Hypertension risk (Minorized distributions of SHR at the 0.5 and 99.5 percentiles). Adjusted models with distributions of SHR at the 0.5th and 99.5th percentiles were used
Supplementary Material 6: Supplemental Table 6. Association of SHR with Hypertension risk (Exclude diabetic patients). After excluding the diabetic cohort, the data were reanalyzed using a multivariate Cox proportional hazards regression model
Supplementary Material 7: Supplementary Fig. 1 E-value Analysis for Unmeasured Confounding in the Association Between SHR and Hypertension Risk. SHR: stress hyperglycemia ratio. Bias plot illustrates the calculated E-values to assess the potential impact of unmeasured confounding factors on the association between SHR and the risk of hypertension. The solid line represents the risk ratio for the exposure-confounder relationship (RR_UD), while the dashed line indicates the risk ratio for the exposure-disease relationship (RR). The plot shows that the association between SHR and hypertension risk remains robust unless unmeasured confounding factors are associated with both SHR and hypertension risks at a magnitude exceeding 1.95. The point (1.95, 1.95) is marked to highlight this threshold. The risk ratio for the exposure-confounder relationship is plotted against the risk ratio for the exposure-disease relationship, with the curve approaching the dashed line as the magnitude of potential unmeasured confounding increases
Data Availability Statement
The data that support the findings of this study are available in China Health and Retirement Longitudinal Study. https://charls.charlsdata.com/pages/data/111/zh-cn.html.