Remnant cholesterol associates with hypertension beyond low-density lipoprotein cholesterol among the general US adult population

Abstract

Background: Previous findings have indicated that elevated low-density lipoprotein cholesterol (LDL-C) and remnant cholesterol (RC) are associated with hypertension. We aim to explore whether higher RC levels may be associated with hypertension beyond LDL-C in the general US adult population.

Methods: This study included 10,842 adults from the National Health and Nutrition Examination Survey (NHANES) 1999-2018. Weighted multivariable logistic regression models were used to estimate the odds ratios (ORs) of hypertension for LDL-C and RC. We also performed analyses examining the association between hypertension and LDL-C vs. RC concordant/discordant groups.

Results: A total of 4,963 (41.54%, weighted) individuals had hypertension. The weighted median levels were LDL-C: 118mg/dL, RC: 20mg/dL. At lower LDL-C clinical cut-point, the proportion of discordantly high RC dramatically increased. After multivariable adjustment, log RC was associated with higher prevalence of hypertension [OR 2.54, 95% confidence interval (CI) 2.17-2.99]. Participants with the highest tertile of RC were more likely to have hypertension (OR 2.18; 95% CI 1.89-2.52) compared with those with the lowest tertile of RC. This association remained marked after including body mass index (BMI), LDL-C, high-density lipoprotein cholesterol (HDL-C) or triglycerides. The association between LDL-C and hypertension was absent after adjusting for BMI, RC or triglycerides. Compared with low LDL-C/low RC group, the discordant low LDL-C/high RC group was associated with hypertension (OR 2.04; 95% CI 1.72-2.42), whereas the high LDL-C/low RC group was not, regardless of BMI, HDL-C or triglycerides. Similar results were observed when examining discordance among different clinical cut-points, except for the cut-point of LDL-C 70 mg/dL and RC 13 mg/dL. To better understand the association, we performed an additional analysis, which showed that among participants with apolipoprotein B < median (92mg/dL), those with discordant RC ≥ median (20mg/dL) had significantly higher odds of having hypertension (OR 1.73; 95% CI 1.38-2.17).

Conclusion: RC was associated with hypertension beyond LDL-C in the general US adult population. This association went beyond increased triglycerides levels, and lipoproteins other than apoB may be involved.

Keywords: apolipoprotein B; dyslipidemia; hypertension; low-density lipoprotein cholesterol; remnant cholesterol.

Figure 1

Proportions of concordance/discordance among individuals with LDL-C below clinical cut-points in the NHANES 1999 to 2018. Discordantly high RC definition: LDL-C < 70mg/dl, RC > 13mg/dl; LDL-C < 100mg/dl, RC > 17mg/dl; LDL-C < 130mg/dl, RC > 23mg/dl. LDL-C cut-points (70, 100, and 130 mg/dL) were based on established guideline recommendations. The equivalent population percentile corresponding to these LDL-C values was used to determine the respective RC cut-points. Percentages were weighted. LDL-C, low-density lipoprotein cholesterol; RC, remnant cholesterol; TG, triglycerides.

Figure 2

Associations (log odds ratios, 95%CIs) of RC and LDL-C concentrations with hypertension using a restricted cubic spline regression model in the NHANES 1999 to 2018. (A) Association between RC concentrations and hypertension. (B) Association between LDL-C concentrations and hypertension. Results were adjusted for age (continuous), sex (male/female), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, other), educational level (less than high school, high school or equivalent, college or above), family income-poverty ratio (≤1.0, 1.1-3.0, >3.0), smoking status (never smoker, former smoker, current smoker), alcohol drinking (non-drinker, low to moderate drinker, heavy drinker), chronic kidney disease (yes or no), diabetes mellitus (yes or no), coronary heart disease (yes or no), eGFR (continuous), FBG (continuous), and HbA1c (continuous). All estimates accounted for complex survey design. Restricted cubic spline regression model was conducted with 3 knots. Shadows represent the 95% CIs for the spline model (with respective medians as reference). CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; RC, remnant cholesterol.

Figure 3

Adjusted ORs (95%CIs) of hypertension according to different concordant/discordant groups by LDL-C clinical cut-points (70, 100, and 130 mg/dL) and percentile-equivalents for RC in the NHANES 1999 to 2018. Model 1: adjusted for age (continuous), sex (male/female), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, other), educational level (less than high school, high school or equivalent, college or above), family income-poverty ratio (≤1.0, 1.1-3.0, >3.0), smoking status (never smoker, former smoker, current smoker), and alcohol drinking (non-drinker, low to moderate drinker, heavy drinker). Model 2: further adjusted (from Model 1) for chronic kidney disease (yes or no), diabetes mellitus (yes or no), and coronary heart disease (yes or no). Model 3: further adjusted (from Model 2) for eGFR (continuous), FBG (continuous), and HbA1c (continuous). Squares represent odds ratios and solid black lines indicate 95% confidence intervals. All estimates accounted for complex survey design. OR, odds ratio; CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; RC, remnant cholesterol.

Figure 4

Associations (log odds ratios, 95%CIs) of different concordant/discordant groups with hypertension according to age quintiles. Low LDL-C and low RC was used as the reference group. Results were adjusted for age (continuous), sex (male/female), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, other), educational level (less than high school, high school or equivalent, college or above), family income-poverty ratio (≤1.0, 1.1-3.0, >3.0), smoking status (never smoker, former smoker, current smoker), alcohol drinking (non-drinker, low to moderate drinker, heavy drinker), chronic kidney disease (yes or no), diabetes mellitus (yes or no), coronary heart disease (yes or no), eGFR (continuous), FBG (continuous), and HbA1c (continuous). Circles represent log odds ratios and vertical lines indicate 95% confidence intervals. All estimates accounted for complex survey design. All estimates accounted for complex survey design. CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; RC, remnant cholesterol.