Abstract
Although remnant cholesterol has been associated with cardiovascular disease, the risk of remnant cholesterol and blood pressure remains unclear. This study aimed to investigate the association between remnant cholesterol and new-onset hypertension. We used middle-aged and older adults aged ≥ 45 years from the first three waves of the China Health and Retirement Longitudinal Study (CHARLS). Estimated remnant cholesterol was calculated as total cholesterol minus high-density lipoprotein cholesterol minus low-density lipoprotein cholesterol. Multiple linear regression analysis and Cox proportional hazards modeling were used to assess the association between remnant cholesterol levels and blood pressure levels and new-onset hypertensive events, respectively. Nonlinear associations were assessed using restricted cubic spline models. A total of 3,044 participants were included, and 839 new-onset hypertensive events (76.5 events per 1000 person-years) were documented during a median follow-up period of 4.0 years. After adjustment for age, sex, lifestyle factors, and other cardiovascular risk factors, compared with participants with normal-range (< 31 mg/dl), those with significantly elevated (≥ 46 mg/dl) remnant cholesterol had elevated systolic blood pressure by 2.36 (95% confidence interval [95% CI]: 0.55, 4.17) mmHg and diastolic blood pressure by 1.66 mmHg (95% CI: 0.47, 2.84), and the risk of new-onset hypertension was 28% higher (hazard ratio [HR]: 1.28; 95% CI: 1.03, 1.59), with no significant association observed in the mildly elevated group (31–46 mg/dl). This association showed similar results in different subgroups. Elevated remnant cholesterol is significantly associated with blood pressure levels and risk of new-onset hypertension, suggesting that remnant cholesterol might be a potential therapeutic target for hypertension prevention.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-09678-5.
Keywords: Remnant cholesterol, Hypertension, Blood pressure, Cohort study, CHARLS
Subject terms: Hypertension, Dyslipidaemias
Introduction
Hypertension is a major global public health challenge and a leading risk factor for cardiovascular disease and premature death worldwide. According to the World Health Organization, approximately 1.28 billion adults aged 30–79 years have hypertension globally, with two-thirds living in low- and middle-income countries1. In China, the prevalence of hypertension has increased substantially over the past decades, affecting about 245 million adults2. The prevention of hypertension has become a critical public health priority.
Dyslipidemia has been well-established as a significant risk factor for cardiovascular disease. While traditional lipid measurements, such as total cholesterol, low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C), have been the focus of cardiovascular risk assessment, emerging evidence suggests that remnant cholesterol may play an important role in cardiovascular health3. remnant cholesterol, calculated as total cholesterol minus HDL-C minus LDL-C, primarily represents very low-density lipoprotein cholesterol and intermediate-density lipoprotein cholesterol4.
Recent studies have demonstrated that elevated remnant cholesterol levels are associated with an increased risk of cardiovascular events, independent of traditional risk factors5–7. A large-scale study involving 109,574 individuals found that higher remnant cholesterol levels were associated with an increased risk of myocardial infarction, ischemic heart disease, and all-cause mortality8. Furthermore, remnant cholesterol is a better predictor of cardiovascular events than triglycerides in some populations9.
While the relationship between remnant cholesterol and cardiovascular disease has been explored, the potential association between remnant cholesterol and hypertension remains largely unknown. Additionally, identifying novel risk factors for hypertension could help develop more effective prevention strategies. However, to our knowledge, no large-scale longitudinal studies have investigated the association between remnant cholesterol and new-onset hypertension in the Chinese population.
Therefore, we conducted this prospective study using data from the China Health and Retirement Longitudinal Study (CHARLS) to examine the association between remnant cholesterol levels and the risk of new-onset hypertension. We hypothesized that elevated remnant cholesterol levels would be associated with an increased risk of developing hypertension.
Methods
This study followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines (TableS1)10.
Study population
We used data from the CHARLS, a nationally representative longitudinal survey of Chinese adults aged 45 years and older. The baseline survey (Wave 1) was conducted in 2011–2012, with follow-up surveys in 2013 (Wave 2) and 2015 (Wave 3). The study used a multistage probability sampling strategy covering 28 provinces11. Of the 13,565 participants in Wave 1–3, we excluded those with (1) age < 45 years (n = 411), (2) missing lipid measurements (n = 3,772), (3) missing blood pressure measurements (n = 3,690), (4) prevalent hypertension at baseline (n = 2,648). The final analysis included 3,044 participants (Fig. 1).
Fig. 1.
Participants screening flowchart.
Assessment of remnant cholesterol
Blood samples for lipid measurements were collected after an overnight fast of at least 12 h. Serum concentrations of total cholesterol, HDL-C, and LDL-C were analyzed at the Youanmen Center for Clinical Laboratory (Beijing, China), which is certified by the College of American Pathologists. Total cholesterol and HDL-C levels were measured using the enzymatic colorimetric method with an automatic analyzer (Beckman Coulter AU series). LDL-C was calculated using the Friedewald equation for participants with triglyceride levels < 400 mg/dl; for those with higher triglyceride levels, an automatic direct assay was used. All laboratory assays followed strict quality control procedures, and the equipment and reagents used were calibrated prior to testing. remnant cholesterol was calculated as total cholesterol minus HDL-C minus LDL-C12. Based on previous guidelines, participants were categorized into three groups according to remnant cholesterol levels: normal range (< 31 mg/dl), mildly elevated (31–45 mg/dl), and significantly elevated (> 45 mg/dl)13.
Ascertainment of incident hypertension events
Blood pressure was measured three times at 5-minute intervals using an electronic blood pressure monitor (Omron HEM-7200) after participants had rested for at least 5 min in a seated position. The mean of the three measurements was used for analysis. Hypertension was defined as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg, or self-reported use of antihypertensive medications14. Hypertension onset was defined as the follow-up visit in which the participant had an average SBP ≥ 140 mmHg or DBP ≥ 90 mmHg, or self-reported a physician diagnosis of hypertension, or reported use of antihypertensive medication. The date of onset was considered the date of that follow-up visit.
Covariates
Covariates were selected based on: established risk factors for hypertension from previous literature, and clinical expertise15–17. Demographic and lifestyle information was collected through face-to-face interviews using standardized questionnaires. Covariates included age, sex, education level, smoking status, and alcohol consumption. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Diabetes was defined as fasting plasma glucose ≥ 7.0 mmol/L, HbA1c ≥ 6.5%, or self-reported physician diagnosis18. Heart disease was based on self-reported physician diagnosis. Chronic kidney disease was defined as an estimated glomerular filtration rate, calculated using the Chronic Kidney Disease Epidemiology Collaboration equation19, based on cystatin C less than 60 mL/min/1.73m2, or self-reported physician diagnosis. Depressive status was assessed by the Center for Epidemiologic Studies Depression Scale20. The missing proportions for all covariates are shown in Table S2 and were imputed using K-nearest neighbor imputation, with k = 1021.
Statistical analysis
Baseline characteristics were presented as median with interquartile (IQR) for continuous variables and numbers (percentages) for categorical variables, according to remnant cholesterol categories, and compared using the χ2 test and Kruskal-Wallis rank sum test, as appropriate.
Multiple linear regression analysis was used to examine the association between remnant cholesterol levels and blood pressure levels, adjusting for potential confounders. Cox proportional hazard models were used to estimate hazard ratio (HR) and 95% confidence interval (95% CI) for the association between remnant cholesterol categories and incident hypertension. Follow-up time was calculated from the date of the baseline survey to the date of hypertension diagnosis or the last follow-up visit, whichever occurred first. Event rates were reported per 1000 person-years of follow-up. Adjustment variables were selected a priori based on existing literature for their relation to hypertension. All adjustment variables were evaluated for collinearity (Figure S1 and Table S3). Four models were constructed: Model 1 was unadjusted; Model 2 adjusted for sociodemographic factors, including age, gender, residence, education, and marriage; Model 3 additionally adjusted for smoking, alcohol consumption, and sleep duration; Model 4 further adjusted for BMI, diabetes, kidney disease, heart disease, depression status, and baseline blood pressure. We performed E-value analysis to quantify the minimum strength of association that an unmeasured confounder would need to have with both exposure and outcome to fully explain away the observed associations22. The E-value was calculated as: RR + sqrt[RR × (RR-1)], where RR represents the observed risk ratio.
Possible nonlinear relationships between the remnant cholesterol and new-onset hypertension were examined by a Cox regression model with a restricted cubic spline. The knots between 3 and 7 were tested respectively, and the model with the lowest Akaike information criterion value was selected for restricted cubic spline (Table S4). Finally, we used 3 knots at the 10th, 50th, and 90th percentiles. Subgroup analyses were performed to evaluate potential effect modification by age, sex, heart disease, chronic kidney disease, and diabetes status. The proportional hazards assumption was tested using Schoenfeld residuals.
To further evaluate the predictive value of remnant cholesterol for incident hypertension, we compared its predictive performance with that of LDL-C, HDL-C, triglycerides, and total cholesterol. For each lipid marker, a separate multivariable Cox proportional hazards model was constructed, adjusting for the same set of covariates as in the main analysis. The Harrell’s C-index was calculated for each model to assess discrimination.
Statistical analyses were performed using Stata 17.0 and R version 4.2.0 (R Foundation for Statistical Computing, Vienna, Austria). Two-sided P values < 0.05 were considered statistically significant.
Results
Baseline characteristics
Among the 3,044 participants included in the final analysis, the median (IQR) age was 57 (51, 63) years, and 1426 (46.8%) were male. The median (IQR) remnant cholesterol level was 18 (10, 29) mg/dl. According to remnant cholesterol categories, 2,355 (77.4%) participants had normal-range levels, 422 (13.9%) had mildly elevated levels, and 267 (8.8%) had significantly elevated levels. Participants with higher remnant cholesterol levels were more likely to be older, female, and have higher BMI, total cholesterol, triglycerides levels, baseline blood pressure, and a higher prevalence of diabetes (Table 1).
Table 1.
Patient demographics and baseline characteristics.
| Characteristic | Remnant cholesterol | P value | |||
|---|---|---|---|---|---|
| Overall (N = 3,044) |
< 31 (N = 2,355) |
31–46 (N = 422) |
≥ 46 (N = 267) |
||
| Age, years | 57 (51, 63) | 57 (51, 64) | 56 (49, 61) | 57 (51, 63) | 0.033 2 |
| Gender | 0.048 3 | ||||
| Male | 1,426 (46.8%) | 1,130 (48.0%) | 176 (41.7%) | 120 (44.9%) | |
| Female | 1,618 (53.2%) | 1,225 (52.0%) | 246 (58.3%) | 147 (55.1%) | |
| Marriage | 0.092 3 | ||||
| Married | 2,768 (90.9%) | 2,141 (90.9%) | 376 (89.1%) | 251 (94.0%) | |
| Other | 276 (9.1%) | 214 (9.1%) | 46 (10.9%) | 16 (6.0%) | |
| Residence | < 0.001 3 | ||||
| Rural | 2,155 (70.8%) | 1,707 (72.5%) | 283 (67.1%) | 165 (61.8%) | |
| Urban | 889 (29.2%) | 648 (27.5%) | 139 (32.9%) | 102 (38.2%) | |
| Education | 0.666 3 | ||||
| Primary and below | 2,134 (70.1%) | 1,660 (70.5%) | 292 (69.2%) | 182 (68.2%) | |
| Secondary and above | 910 (29.9%) | 695 (29.5%) | 130 (30.8%) | 85 (31.8%) | |
| CESD | 7 (4, 12) | 7 (4, 12) | 7 (3, 12) | 7 (4, 11) | 0.349 2 |
| BMI, kg/m2 | 22.5 (20.4, 24.7) | 22.1 (20.1, 24.2) | 23.9 (21.5, 26.1) | 24.1 (22.2, 26.7) | < 0.001 2 |
| Sleep duration at night, h/day | 6 (5, 8) | 6 (5, 8) | 6 (5, 8) | 6 (5, 8) | 0.747 2 |
| Former drinking | 0.826 3 | ||||
| None | 1,890 (62.1%) | 1,460 (62.0%) | 267 (63.3%) | 163 (61.0%) | |
| Yes | 1,154 (37.9%) | 895 (38.0%) | 155 (36.7%) | 104 (39.0%) | |
| Current drinking | 0.417 3 | ||||
| None | 2,022 (66.4%) | 1,559 (66.2%) | 291 (69.0%) | 172 (64.4%) | |
| Yes | 1,022 (33.6%) | 796 (33.8%) | 131 (31.0%) | 95 (35.6%) | |
| Former smoking | 0.556 3 | ||||
| None | 1,864 (61.2%) | 1,430 (60.7%) | 265 (62.8%) | 169 (63.3%) | |
| Yes | 1,180 (38.8%) | 925 (39.3%) | 157 (37.2%) | 98 (36.7%) | |
| Current smoking | 0.172 3 | ||||
| None | 2,101 (69.0%) | 1,609 (68.3%) | 295 (69.9%) | 197 (73.8%) | |
| Yes | 943 (31.0%) | 746 (31.7%) | 127 (30.1%) | 70 (26.2%) | |
| SBP at baseline, mmHg | 117 (109, 124) | 116 (108, 124) | 117 (110, 125) | 119 (110, 126) | < 0.001 2 |
| DBP at baseline, mmHg | 69 (63, 75) | 69 (63, 75) | 70 (65, 75) | 71 (64, 76) | 0.002 2 |
| SBP at follow-up, mmHg | 119 (110, 130) | 119 (109, 129) | 119 (110, 131) | 122 (113, 136) | < 0.001 2 |
| DBP at follow-up, mmHg | 71 (65, 77) | 71 (64, 77) | 72 (65, 78) | 73 (66, 81) | < 0.001 2 |
| Total cholesterol, mg/dl | 188 (165, 211) | 183 (162, 207) | 197 (174, 224) | 210 (182, 238) | < 0.001 2 |
| Triglyceride, mg/dl | 97 (70, 141) | 84 (65, 111) | 165 (134, 196) | 273 (220, 344) | < 0.001 2 |
| HDL-C, mg/dl | 51 (41, 61) | 54 (46, 64) | 42 (36, 51) | 36 (31, 42) | < 0.001 2 |
| LDL-C, mg/dl | 112 (92, 134) | 113 (93, 134) | 114 (96, 138) | 102 (76, 126) | < 0.001 2 |
| Remnant cholesterol, mg/dl | 18 (10, 29) | 14 (9, 21) | 37 (34, 41) | 61 (51, 80) | < 0.001 2 |
| Heart disease | 0.147 3 | ||||
| None | 2,787 (91.6%) | 2,165 (91.9%) | 376 (89.1%) | 246 (92.1%) | |
| Yes | 257 (8.4%) | 190 (8.1%) | 46 (10.9%) | 21 (7.9%) | |
| CKD | 0.166 3 | ||||
| None | 2,855 (93.8%) | 2,199 (93.4%) | 404 (95.7%) | 252 (94.4%) | |
| Yes | 189 (6.2%) | 156 (6.6%) | 18 (4.3%) | 15 (5.6%) | |
| Diabetes | < 0.001 3 | ||||
| None | 2,727 (89.6%) | 2,147 (91.2%) | 366 (86.7%) | 214 (80.1%) | |
| Yes | 317 (10.4%) | 208 (8.8%) | 56 (13.3%) | 53 (19.9%) | |
1 Data were presented as Median (IQR) and n (%).
2 Kruskal-Wallis rank sum test.
3 Pearson’s Chi-squared test.
HDL = high-density lipoprotein; LDL = low-density lipoprotein; CESD = Center for Epidemiologic Studies Depression; CKD = chronic kidney disease; SBP = systolic blood pressure; DBP = diastolic blood pressure; BMI = body mass index.
Association between remnant cholesterol and blood pressure levels at baseline
In the multiple linear regression analysis, after adjusting for potential confounders, participants with significantly elevated remnant cholesterol had a 3.97 mmHg (95% CI: 1.96, 5.98) higher systolic blood pressure and a 2.50 mmHg (95% CI: 1.20, 3.80) higher diastolic blood pressure compared with those in the normal range group. The associations remained significant after further adjustment for traditional cardiovascular risk factors (Model 4: systolic blood pressure β = 2.36 mmHg, 95% CI: 0.55, 4.17; diastolic blood pressure β = 1.66 mmHg, 95% CI: 0.47, 2.84) (Table 2). Restricted cubic spline analysis revealed a positive, approximately linear relationship between remnant cholesterol and blood pressure level (SBP: P for nonlinearity=0.088; DBP: P for nonlinearity=0.975) (Fig. 2).
Table 2.
Association of remnant cholesterol and blood pressure level at baseline *.
| Outcomes | Model 1 | Model 2 | Model 3 | Model 4 |
|---|---|---|---|---|
| β (95% CI), P-value | β (95% CI), P-value | β (95% CI), P-value | β (95% CI), P-value | |
| SBP | ||||
| Remnant Cholesterol | ||||
| <31, mg/dl | Ref. | Ref. | Ref. | Ref. |
| 31–46, mg/dl | 1.31 (95% CI: −0.34, 2.95), 0.120 | 1.66 (95% CI: 0.03, 3.28), 0.046 | 1.69 (95% CI: 0.06, 3.31), 0.042 | 0.52 (95% CI: −0.95, 2.00), 0.486 |
| ≥46, mg/dl | 3.97 (95% CI: 1.96, 5.98), < 0.001 | 4.15 (95% CI: 2.16, 6.14), < 0.001 | 4.13 (95% CI: 2.14, 6.12), < 0.001 | 2.36 (95% CI: 0.55, 4.17), 0.011 |
| DBP | ||||
| Remnant Cholesterol | ||||
| <31, mg/dl | Ref. | Ref. | Ref. | Ref. |
| 31–46, mg/dl | 1.08 (95% CI: 0.02, 2.14), 0.047 | 1.06 (95% CI: 0.01, 2.11), 0.048 | 1.07 (95% CI: 0.02, 2.12), 0.046 | 0.51 (95% CI: −0.46, 1.48), 0.30 |
| ≥46, mg/dl | 2.5 (95% CI: 1.20, 3.80), < 0.001 | 2.45 (95% CI: 1.17, 3.74), < 0.001 | 2.41 (95% CI: 1.13, 3.70), < 0.001 | 1.66 (95% CI: 0.47, 2.84), 0.006 |
SBP = systolic blood pressure; DBP = diastolic blood pressure; Ref. = Reference.
Model 1: unadjusted.
Model 2: adjusted for age, gender, marriage, residence, and education.
Model 3: adjusted for Model 2 + former drinking, current drinking, former smoking, current smoking, sleep duration.
Model 4: adjusted for Model 3 + body mass index, heart disease, diabetes, chronic kidney disease, depressive score, baseline systolic blood pressure, and diastolic blood pressure.
* Blood pressure measurements refer to values obtained during follow-up.
Fig. 2.
The restricted cubic spline for the relationships between remnant cholesterol and blood pressure. The β from the multivariate multiple linear regression models were adjusted for the variables of Model 4 in Table 2. The duck red lines indicate the adjusted β and the light red shade indicates the 95% CI. Blood pressure measurements refer to values obtained during follow-up.
Association between remnant cholesterol and incident hypertension events
During a median follow-up of 4.0 years, 839 participants developed hypertension, corresponding to an incidence rate of 76.5 cases per 1,000 person-years. Among the 839 incident hypertension cases, 204 cases were identified through self-reported physician diagnosis, and 635 cases were identified through blood pressure measurements meeting the diagnostic criteria (≥ 140/90 mmHg).
In the Cox proportional hazards analysis, compared with the remnant cholesterol normal-range group, the HR for incident hypertension was 1.12 (95% CI: 0.93, 1.36) for the mildly elevated group and 1.47 (95% CI: 1.19, 1.83) for the significantly elevated group in a crude model (Table 3). The associations between significantly remnant cholesterol elevated and incident hypertension remained robust after further adjustment for covariates (Model 4: HR = 1.28; 95% CI: 1.03, 1.59 for significantly elevated; Harrell’s C-index = 0.674) (Table 3).
Table 3.
Association of remnant cholesterol and new-onset hypertension.
| Outcomes | Case/sample | Incidence Rate, per 1000 Person-Years | 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 | |||
| Remnant Cholesterol | ||||||
| <31, mg/dl | 609/2,355 | 64.78 | Ref. | Ref. | Ref. | Ref. |
| 31–46, mg/dl | 132/422 | 78.10 | 1.12 (95% CI: 0.93, 1.36), 0.224 | 1.14 (95% CI: 0.95, 1.38), 0.167 | 1.14 (95% CI: 0.94, 1.38), 0.178 | 0.99 (95% CI: 0.82, 1.20), 0.948 |
| ≥46, mg/dl | 98/267 | 92.00 | 1.47 (95% CI: 1.19, 1.83), < 0.001 | 1.49 (95% CI: 1.21, 1.85), < 0.001 | 1.51 (95% CI: 1.22, 1.87), < 0.001 | 1.28 (95% CI: 1.03, 1.59), 0.026 |
HR = hazard risk; 95% CI = 95% confidence interval; Ref. = Reference.
Model 1: unadjusted.
Model 2: adjusted for age, gender, marriage, residence, and education.
Model 3: adjusted for Model 2 + former drinking, current drinking, former smoking, current smoking, sleep duration.
Model 4: adjusted for Model 3 + body mass index, heart disease, diabetes, chronic kidney disease, depressive score, baseline systolic blood pressure, and diastolic blood pressure.
Excluding participants with self-reported hypertension, the association between the two remained robust (HR = 1.33; 95% CI: 1.09, 1.62). We conducted an additional sensitivity analysis excluding individuals with baseline diabetes or obesity (BMI ≥ 28 kg/m²). The association between remnant cholesterol and incident hypertension remained statistically significant (Table S9).
In the multivariable Cox regression models including additional lipid marker separately, only the highest tertile of triglycerides showed a significant association with increased risk of hypertension (Table S5-S8). The Harrell’s C-indices for models with HDL-C, LDL-C, triglycerides, and total cholesterol were 0.672, 0.671, 0.675, and 0.672, respectively, indicating similar discrimination performance among these lipid markers.
For the association between significantly elevated remnant cholesterol (≥ 46 mg/dl) and incident hypertension, the E-value was 1.88 for the point estimate and 1.21 for the lower confidence interval bound. This suggests that an unmeasured confounder would need to be associated with both remnant cholesterol and hypertension by a risk ratio of at least 1.88 (or 1.21 for the lower confidence bound) to fully explain away the observed association.
Restricted cubic spline analysis revealed a positive, approximately linear relationship between remnant cholesterol levels and incident hypertension risk (P for nonlinearity=0.997). The risk of hypertension increased progressively with higher remnant cholesterol levels, with no apparent threshold effect (Fig. 3).
Fig. 3.
The restricted cubic spline for the relationships between remnant cholesterol and risk of new-onset hypertension. The HR from the multivariate Cox proportional risk models were adjusted for the variables of Model 4 in Table 3. The duck red lines indicate the adjusted hazard ratio, and the light red shade indicates 95% CI.
Subgroup analysis
The association between remnant cholesterol and incident hypertension was generally consistent across various subgroups defined by age (< 60 vs. ≥60 years), gender, diabetes, chronic kidney disease, and heart disease (all P for interaction >0.05) (Table 4). However, the association appeared to be slightly stronger among male participants (HR = 1.39; 95% CI: 1.02, 1.90) compared with females (HR = 1.29; 95% CI: 0.94, 1.77). Interestingly, a significant association between significantly elevated remnant cholesterol and new-onset hypertension was shown in the diabetes-free and heart disease-free populations (non-diabetes: HR = 1.32; 95% CI:1.03, 1.68; non-heart disease: HR = 1.32; 95% CI:1.03, 1.68).
Table 4.
Subgroup analysis for remnant cholesterol and incident hypertension.
| Subgroup | Case/total | < 31, mg/dl | 31–46, mg/dl | ≥ 46, mg/dl | P for interaction |
|---|---|---|---|---|---|
| Ref. | HR (95% CI), P-value | HR (95% CI), P-value | |||
| Age group | 0.673 | ||||
| < 60 years | 472/1,843 | Ref. | 0.93 (95% CI:0.72, 1.19), 0.549 | 1.26 (95% CI:0.95, 1.68), 0.112 | |
| ≥ 60 years | 367/1,201 | Ref. | 1.08 (95% CI:0.81, 1.44), 0.610 | 1.29 (95% CI:0.89, 1.85), 0.177 | |
| Gender | 0.770 | ||||
| Female | 417/1,618 | Ref. | 1.04 (95% CI:0.81, 1.33), 0.765 | 1.29 (95% CI:0.94, 1.77), 0.118 | |
| Male | 421/1,426 | Ref. | 0.95 (95% CI:0.70, 1.27), 0.709 | 1.39 (95% CI:1.02, 1.90), 0.036 | |
| Diabetes | 0.238 | ||||
| None | 722/2,727 | Ref. | 1.04 (95% CI:0.85, 1.28), 0.702 | 1.32 (95% CI:1.03, 1.68), 0.029 | |
| Yes | 117/317 | Ref. | 0.75 (95% CI:0.47, 1.20), 0.232 | 1.21 (95% CI:0.72, 2.05), 0.468 | |
| Heart disease | 0.829 | ||||
| None | 773/2,787 | Ref. | 0.99 (95% CI:0.82, 1.21), 0.959 | 1.32 (95% CI:1.04, 1.66), 0.020 | |
| Yes | 66/257 | Ref. | 1.05 (95% CI:0.55, 2.00), 0.884 | 0.79 (95% CI:0.34, 1.87), 0.595 | |
| Chronic kidney disease | 0.146 | ||||
| None | 801/2,855 | 0.98 (95% CI:0.81, 1.19), 0.825 | 1.25 (95% CI:0.99, 1.57), 0.057 | ||
| Yes | 38/189 | 0.82 (95% CI:0.18, 3.70), 0.792 | 1.91 (95% CI:0.60, 6.11), 0.275 | ||
| HR = hazard risk; 95% CI = 95% confidence interval; Ref. = Reference | |||||
Discussion
Principal findings
In this large prospective cohort study of middle-aged and older Chinese adults, we found that elevated remnant cholesterol levels were independently associated with an increased risk of incident hypertension. This association remained robust after adjusting traditional cardiovascular risk factors and showed a linear relationship.
While the observed associations were modest in magnitude, our sensitivity analyses suggest they are unlikely to be fully explained by unmeasured confounding. The E-value analysis indicates that an unmeasured confounder would need to be associated with both remnant cholesterol and hypertension by a risk ratio of at least 1.88 (or 1.21 for the confidence interval) to fully explain away the observed association. Given that we have already adjusted for major known risk factors for hypertension including age, gender, body mass index, diabetes, kidney disease, and baseline blood pressure, the existence of such a strong unmeasured confounder that could completely nullify our findings is less likely.
Comparison with previous studies
Our findings extend previous research on the relationship between remnant cholesterol and cardiovascular outcomes. While several studies have demonstrated associations between remnant cholesterol and atherosclerotic cardiovascular disease3,23, evidence linking remnant cholesterol directly to hypertension has been limited. The Copenhagen General Population Study reported that elevated remnant cholesterol was associated with increased blood pressure levels5, but did not examine incident hypertension. A recent analysis of the Framingham Offspring Study found that triglyceride-rich lipoproteins were associated with incident hypertension24, which aligns with our findings given that remnant cholesterol largely reflects triglyceride-rich lipoproteins. Our study provides novel evidence in an Asian population, where dietary patterns and cardiovascular risk profiles differ from Western populations.
Although we observed a significant association between elevated remnant cholesterol levels and incident hypertension, our observational study cannot determine whether this relationship is direct or mediated by other metabolic risk factors. Many individuals with high remnant cholesterol also have obesity, diabetes, or other conditions known to increase hypertension risk. While we adjusted for these confounders, residual confounding or mediation effects may persist. Additionally, due to the lack of intervention in this study, it remains unclear whether lowering remnant cholesterol would reduce hypertension risk. Thus, causality cannot be inferred, and further interventional studies are needed to clarify these relationships.
Potential mechanisms
Several mechanisms might explain the association between remnant cholesterol and hypertension development. First, remnant cholesterol particles can penetrate and accumulate in the arterial wall more easily than larger lipoproteins, leading to inflammation and endothelial dysfunction25. Second, elevated remnant cholesterol may promote oxidative stress and reduce nitric oxide bioavailability, contributing to arterial stiffness and vasoconstriction26. Third, remnant cholesterol-rich lipoproteins can activate the renin-angiotensin-aldosterone system, directly affecting blood pressure regulation27. Additionally, insulin resistance may serve as a common pathway linking remnant cholesterol metabolism and blood pressure regulation28.
Strengths and limitations
Our study has several strengths. First, it is a large prospective cohort study with standardized measurements and detailed information on potential confounders. Second, blood pressure was measured multiple times following standardized protocols, reducing measurement error. However, several limitations should be acknowledged. First, remnant cholesterol was calculated rather than directly measured, which might introduce measurement error. However, this calculation method has been validated and widely used in epidemiological studies29. Second, we lacked information on changes in remnant cholesterol levels during follow-up, which might have led to misclassification of exposure. Third, despite comprehensive adjustment for confounders, residual confounding cannot be completely ruled out. Fourth, our findings in middle-aged and older Chinese adults may not be generalizable to other populations. Fourth, our analysis treated blood pressure as a cross-sectional measurement and did not account for the longitudinal nature of blood pressure measurements or potential changes in remnant cholesterol levels over time. A more sophisticated approach using mixed-effects models or time-varying exposure analysis could potentially provide additional insights into the temporal relationship between remnant cholesterol and blood pressure changes. Future studies employing such methodologies with more frequent measurements of both blood pressure and remnant cholesterol would be valuable to better understand the dynamic relationship between these factors. Fifth, the lack of information on antihypertensive medication use in the CHARLS database is a limitation of our study. This might have led to underestimation of blood pressure levels in treated participants and potential selection bias. However, our use of both self-reported diagnosis and objective blood pressure measurements for hypertension ascertainment, which is consistent with previous studies using CHARLS data, helps minimize misclassification. Additionally, sensitivity analyses excluding self-reported cases showed similar results, suggesting the robustness of our findings despite this limitation. Sixth, we did not account for the use of lipid-lowering medications among participants. The inclusion of individuals taking such medications could have affected cholesterol levels and potentially influenced the observed associations between remnant cholesterol and incident hypertension in this study. Finally, Our study focused on the association rather than prediction. Thus, we did not perform further analyses such as c-statistics, net reclassification index, or integrated discrimination index.
Implications for clinical practice
Our findings have several important clinical implications. First, they suggest that remnant cholesterol measurement might improve hypertension risk assessment beyond traditional risk factors. Second, the observed association with even mildly elevated levels indicates that earlier intervention might be beneficial. Third, while our findings highlight an association between remnant cholesterol and incident hypertension, our observational design does not establish causality. Therefore, it is premature to conclude that targeting remnant cholesterol would prevent hypertension. Nevertheless, recent trials of PCSK9 inhibitors and other novel therapies targeting triglyceride-rich lipoproteins provide promising avenues for lipid management, which may warrant further investigation in the context of hypertension prevention30.
Conclusion
In conclusion, this longitudinal study provides evidence that elevated remnant cholesterol levels are independently associated with increased blood pressure levels and a higher risk of new-onset hypertension among middle-aged and older Chinese adults. The consistency of these findings across various subgroup analyses strengthens the robustness of our results. These findings suggest that remnant cholesterol might serve as a novel therapeutic target for hypertension prevention and highlight the importance of monitoring and managing remnant cholesterol levels in clinical practice.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We thank the China Center for Economic Research at Peking University for providing us with the data, and the CHARLS research and field team for collecting the data.
Author contributions
Linghui Cui: Conceptualization, Data curation, Formal analysis, Methodology, Writing-original draft. Chang Liu: Conceptualization, Writing-review & editing. Luzong Yang: Data curation. Jing Liang: Data curation. Hao Wang: Data curation. Chenyang Liu: Data curation. Fan Zhang: Conceptualization, Data curation, Formal analysis, Methodology, Writing-original draft. Min Cao: Funding acquisition, Supervision, Writing-review & editing. All authors read and approved the final manuscript. All authors contributed to this article.
Funding
This study was supported by the National Natural Science Foundation of China (82374397), the Fifth National Training Program for Clinical Excellence in Traditional Chinese Medicine [National TCM Human Resource Education Letter (2022) No. 1], and Shanghai University of Traditional Chinese Medicine High-Level Science and Innovation Program for Healthcare Integration (602075D).
Data availability
The data that support the findings of this study are available from CHARLS project site, subject to registration and application process. Further details can be found at https://charls.charlsdata.com/. The datasets generated and/or analysed during the current study are not publicly available and is subject to registration and application process but are available from the corresponding author on reasonable request.
Declarations
Conflict of interest
There are no conflicts of interest that are directly relevant to the content of this article.
Ethics statement
The studies involving human participants were reviewed and approved by the Biomedical Ethical Review Committee of Peking University (IRB00001052-11015). The participants provided their written informed consent to participate in this study. The privacy rights of participants were observed.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Co-first authors: Linghui Cui, Chang Liu and Luzong Yang.
Contributor Information
Fan Zhang, Email: fan_zhang1993@163.com.
Min Cao, Email: caomin_cm@163.com.
References
- 1.Worldwide trends in hypertension. Prevalence and progress in treatment and control from 1990 to 2019: a pooled analysis of 1201 population-representative studies with 104 million participants. Lancet (London England). 398, 957–980. 10.1016/s0140-6736(21)01330-1 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wang, Z. et al. Status of hypertension in china: results from the China hypertension survey, 2012–2015. Circulation137, 2344–2356. 10.1161/circulationaha.117.032380 (2018). [DOI] [PubMed] [Google Scholar]
- 3.Nordestgaard, B. G. Triglyceride-Rich lipoproteins and atherosclerotic cardiovascular disease: new insights from epidemiology, genetics, and biology. Circ. Res.118, 547–563. 10.1161/circresaha.115.306249 (2016). [DOI] [PubMed] [Google Scholar]
- 4.Varbo, A. & Nordestgaard, B. G. Remnant cholesterol and Triglyceride-Rich lipoproteins in atherosclerosis progression and cardiovascular disease. Arterioscler. Thromb. Vasc Biol.36, 2133–2135. 10.1161/atvbaha.116.308305 (2016). [DOI] [PubMed] [Google Scholar]
- 5.Jørgensen, A. B. et al. Genetically elevated non-fasting triglycerides and calculated remnant cholesterol as causal risk factors for myocardial infarction. Eur. Heart J.34, 1826–1833. 10.1093/eurheartj/ehs431 (2013). [DOI] [PubMed] [Google Scholar]
- 6.Guo, D. C. et al. Remnant cholesterol and risk of incident hypertension: a population-based prospective cohort study. Hypertens. Res.47, 1157–1166. 10.1038/s41440-023-01558-7 (2024). [DOI] [PubMed] [Google Scholar]
- 7.Yu, Z. et al. Remnant cholesterol and the risk of carotid plaque in hypertension: results from a community-based screening among old adults in hangzhou, China. Sci. Rep.14, 8407. 10.1038/s41598-024-58484-y (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Varbo, A. et al. Remnant cholesterol as a causal risk factor for ischemic heart disease. J. Am. Coll. Cardiol.61, 427–436. 10.1016/j.jacc.2012.08.1026 (2013). [DOI] [PubMed] [Google Scholar]
- 9.Jepsen, A. M. et al. Increased remnant cholesterol explains part of residual risk of All-Cause mortality in 5414 patients with ischemic heart disease. Clin. Chem.62, 593–604. 10.1373/clinchem.2015.253757 (2016). [DOI] [PubMed] [Google Scholar]
- 10.Vandenbroucke, J. P. et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. PLoS Med.4, e297. 10.1371/journal.pmed.0040297 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhao, Y., Hu, Y., Smith, J. P., Strauss, J. & Yang, G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int. J. Epidemiol.43, 61–68. 10.1093/ije/dys203 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Varbo, A. & Nordestgaard, B. G. Remnant lipoproteins. Curr. Opin. Lipidol.28, 300–307. 10.1097/mol.0000000000000429 (2017). [DOI] [PubMed] [Google Scholar]
- 13.Kinoshita, M. et al. : Japan Atherosclerosis Society (JAS) Guidelines for Prevention of Atherosclerotic Cardiovascular Diseases 2017. J Atheroscler Thromb 25 846–984, (2018). 10.5551/jat.GL2017. [DOI] [PMC free article] [PubMed]
- 14.Whelton, P. K. et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 138 (2018) e426-e483, (2017). 10.1161/cir.0000000000000597 [DOI] [PubMed]
- 15.Chen, J. et al. Association between platelet to high-density lipoprotein cholesterol ratio (PHR) and hypertension: evidence from NHANES 2005–2018. Lipids Health Dis.23, 346. 10.1186/s12944-024-02342-3 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wu, J. & Guo, J. Non-high-density lipoprotein cholesterol to high-density lipoprotein cholesterol ratio (NHHR) and hypertension in American adults: a NHANES cross-sectional study. Front. Physiol.15, 1398793. 10.3389/fphys.2024.1398793 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yang, S., Zhang, Y., Zhou, Z. & Duan, X. Association of triglyceride-Glucose index, triglyceride to High-Density lipoprotein cholesterol ratio, and related parameters with prehypertension and hypertension. J. Clin. Hypertens. (Greenwich). 10.1111/jch.14926 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes-2021. Diabetes care 44. S15-s33, (2021). 10.2337/dc21-S002 [DOI] [PubMed]
- 19.Levey, A. S. et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of diet in renal disease study group. Ann. Intern. Med.130, 461–470. 10.7326/0003-4819-130-6-199903160-00002 (1999). [DOI] [PubMed] [Google Scholar]
- 20.Andresen, E. M., Malmgren, J. A., Carter, W. B. & Patrick, D. L. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for epidemiologic studies depression Scale). Am. J. Prev. Med.10, 77–84 (1994). [PubMed] [Google Scholar]
- 21.Troyanskaya, O. et al. Missing value Estimation methods for DNA microarrays. Bioinformatics17, 520–525. 10.1093/bioinformatics/17.6.520 (2001). [DOI] [PubMed] [Google Scholar]
- 22.VanderWeele, T. J. & Ding, P. Sensitivity analysis in observational research: introducing the E-Value. Ann. Intern. Med.167, 268–274. 10.7326/m16-2607 (2017). [DOI] [PubMed] [Google Scholar]
- 23.Varbo, A., Benn, M. & Nordestgaard, B. G. Remnant cholesterol as a cause of ischemic heart disease: evidence, definition, measurement, atherogenicity, high risk patients, and present and future treatment. Pharmacol. Ther.141, 358–367. 10.1016/j.pharmthera.2013.11.008 (2014). [DOI] [PubMed] [Google Scholar]
- 24.Wang, J. et al. Lipoprotein subclass profiles in individuals with varying degrees of glucose tolerance: a population-based study of 9399 Finnish men. J. Intern. Med.272, 562–572. 10.1111/j.1365-2796.2012.02562.x (2012). [DOI] [PubMed] [Google Scholar]
- 25.Borén, J. & Williams, K. J. The central role of arterial retention of cholesterol-rich apolipoprotein-B-containing lipoproteins in the pathogenesis of atherosclerosis: a triumph of simplicity. Curr. Opin. Lipidol.27, 473–483. 10.1097/mol.0000000000000330 (2016). [DOI] [PubMed] [Google Scholar]
- 26.Bernelot Moens, S. J. et al. Remnant cholesterol elicits arterial wall inflammation and a multilevel cellular immune response in humans. Arterioscler. Thromb. Vasc Biol.37, 969–975. 10.1161/atvbaha.116.308834 (2017). [DOI] [PubMed] [Google Scholar]
- 27.Ference, B. A. et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European atherosclerosis society consensus panel. Eur. Heart J.38, 2459–2472. 10.1093/eurheartj/ehx144 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Packard, C. J., Boren, J. & Taskinen, M. R. Causes and consequences of hypertriglyceridemia. Front. Endocrinol. (Lausanne). 11, 252. 10.3389/fendo.2020.00252 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Langsted, A. & Nordestgaard, B. G. Nonfasting versus fasting lipid profile for cardiovascular risk prediction. Pathology51, 131–141. 10.1016/j.pathol.2018.09.062 (2019). [DOI] [PubMed] [Google Scholar]
- 30.Pradhan, A. D. et al. Rationale and design of the pemafibrate to reduce cardiovascular outcomes by reducing triglycerides in patients with diabetes (PROMINENT) study. Am. Heart J.206, 80–93. 10.1016/j.ahj.2018.09.011 (2018). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from CHARLS project site, subject to registration and application process. Further details can be found at https://charls.charlsdata.com/. The datasets generated and/or analysed during the current study are not publicly available and is subject to registration and application process but are available from the corresponding author on reasonable request.