Different Effects of Atherogenic Lipoproteins and Blood Pressure on Arterial Structure and Function: The Bogalusa Heart Study

Abstract

Differential impact of non–high‐density lipoprotein cholesterol (total cholesterol minus high‐density lipoprotein cholesterol) and blood pressure on arterial wall thickness and stiffness was examined in 900 black and white adults aged 24–43 years. Blacks compared with whites had greater values of pulse wave velocity (5.4 m/sec vs. 5.2 m/sec; p<0.01) and carotid artery intima–media thickness (0.83 mm vs. 0.80 mm; p<0.01). Non–high‐density lipoprotein cholesterol was significantly associated with carotid intima–media thickness (standardized regression coefficient [b]=0.21; p<0.01), but not with pulse wave velocity (b=0.03; p=0.37), after adjusting for race, sex, age, body mass index, insulin, glucose, and smoking. Systolic blood pressure was associated significantly stronger with pulse wave velocity (b=0.36; p<0.01) than with carotid intima–media thickness (b=0.15; p<0.01). No race difference in these relationships was found. The results of this study indicate that atherogenic lipoproteins and blood pressure may play different roles in the development of arterial wall stiffness and atherosclerosis.

Alterations of vascular structure and function have been increasingly recognized as strong predictors of subsequent cardiovascular (CV) morbidity and mortality. ¹ , ² , ³ During the aging process, major structural and functional changes in arteries occur in terms of increases in wall thickness and decreases in elasticity and compliance. ⁴ , ⁵ Furthermore, both vascular structure and function measures are associated with an adverse profile of CV disease risk factors and their clustering. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ Among the traditional risk factors, high levels of blood pressure and atherogenic lipoproteins are the major determinants of atherosclerosis. Adulthood carotid intima–media thickness (IMT) and arterial pulse wave velocity (PWV) are found to be related even to childhood low‐density lipoprotein (LDL) cholesterol and blood pressure. ¹⁴ , ¹⁵ , ¹⁶ Although there is overwhelming evidence for the influence of age and hypertension on arterial wall stiffening and thickening, results from recent studies are not consistent regarding the impact of LDL cholesterol on arterial wall thickness and stiffness. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³

The objective of this study is to simultaneously examine the differential impact of blood pressure and atherogenic, apoprotein B‐containing lipoproteins measured as non–high‐density lipoprotein (HDL) cholesterol (total cholesterol minus HDL cholesterol) on arterial wall thickness and stiffness in healthy, asymptomatic young adults from the Bogalusa Heart Study. ⁶

Study Cohort

The Bogalusa Heart Study is a community‐based epidemiologic study of the early natural history of CV disease in children and young adults from the semirural, biracial (65% white, 35% black) community of Bogalusa, LA. A total of 900 subjects (648 whites, 252 blacks) aged 24–43 years were examined during 2000 to 2002 for CV risk factors, and ultrasonography of the carotid artery and aortofemoral PWV were evaluated. Subjects on antihypertensive (n=66), cholesterol‐lowering (n=27), or insulin (n=7) medications were excluded from the analysis. All subjects gave informed consent. Study protocols were approved by the Institutional Review Board of the Tulane University Medical Center.

General Examinations

All examinations followed the same protocols; procedures for general examination were described elsewhere. ¹⁷ Subjects were instructed to fast for 12–14 hours before the screening. Compliance of fasting was determined by interview on the morning of examination. Height (±0.1 cm) and weight (± 0.1 kg) were measured twice. Body mass index (BMI) (weight in kilograms divided by the square of the height in meters) was used as a measure of obesity. Blood pressure levels were measured on the right arm of subjects in a relaxed, sitting position by two randomly assigned nurses (three replicates each). The first and fifth Korotkoff phases were used to determine systolic and diastolic blood pressures. Means of replicate readings were used for analyses. Self‐administered questionnaires were used to obtain information on tobacco use.

Serum Lipids, Insulin, and Glucose

Cholesterol levels in serum were determined by an enzymatic procedure ¹⁸ on the Abbott VP instrument (Abbott Laboratories, North Chicago, IL). Serum lipoprotein cholesterols were analyzed by a combination of heparin‐calcium precipitation and agar‐agarose gel electrophoresis procedures. ¹⁹ The procedures met the performance requirements of the lipid standardization program of the Centers for Disease Control and Prevention (CDC) in Atlanta, GA. The laboratory has been monitored for precision and accuracy of lipid measurements by the agency's surveillance program since 1973. Measurements on CDC‐assigned quality control samples showed no consistent bias over time within or between surveys. Intraclass correlation coefficients, a measure of reproducibility of the entire process from blood collection to data processing, between the blind duplicate values (n=103) were 0.98 for total cholesterol and 0.99 for HDL cholesterol. Non‐HDL cholesterol was calculated as total cholesterol minus HDL cholesterol.

Fasting plasma insulin was measured by a commercial radioimmunoassay kit (Padebas Pharmacia, Piscataway, NJ) and plasma glucose by an enzymatic method using the Beckman Instant Glucose Analyzer (Beckman Instruments, Palo Alto, CA). Intraclass correlation between the blind duplicate values was 0.98 for both insulin and glucose. Homeostasis model assessment of insulin resistance (HOMA‐IR) was calculated according to the formula: HOMA‐IR=fasting insulin (μU/mL) × fasting glucose (mmol/L)/22.5. This model is considered useful to assess insulin resistance in epidemiologic studies. ²⁰

Aorta‐Femoral PWV

Aorta‐femoral PWV was measured by a trained sonographer using an echocardiographic instrument (Power Vision Toshiba SSH‐380 Digital Ultrasound System, Toshiba America Medical Systems, Carrollton, TX). A nondirectional transcutaneous Doppler flow probe (Toshiba PSK25AT, 2.5 MHz) was positioned at the suprasternal notch and another probe (Toshiba PCK703AT, 7.5 MHz) at the left femoral artery in a supine position. After the collection of the waveform data, the distance between the aorta and femoral arteries was measured. Aorta‐femoral PWV was calculated by dividing the distance traveled by the time differential between the two waveforms. Results from three data collection runs for each participant were averaged.

Carotid Ultrasonography

Carotid ultrasound measurements were done on a Toshiba Ultrasound instrument (Power Vision Toshiba SSH‐380 Digital ultrasound system, Toshiba America Medical Systems, Carrollton, TX), using a 7.5‐MHz linear array transducer. Images were recorded on the right and left common carotid arteries, carotid bulb, and internal carotid artery according to previously developed protocols for the Atherosclerosis Risk in Communities (ARIC) study. ²¹ Images were recorded on super VHS videotapes and examined by certified readers from the Vascular Ultrasound Research Laboratory in Wake Forest, NC, using semiautomatic ultrasound imaging. The mean values of three left and three right far walls measurements were calculated for the three carotid artery sites.

Statistical Methods

All data analyses were performed using Statistical Analysis System version 9.1 (SAS Institute, Inc., Cary, NC). Race and sex differences in mean values of PWV, carotid IMT, and CV risk factors were tested by analysis of covariance models. The impact of non‐HDL cholesterol and systolic blood pressure on PWV and carotid IMT was examined by multivariate regression models, adjusting for age, sex, BMI, HOMA‐IR, smoking, and/or race. To compare the effect size of non‐HDL cholesterol and systolic blood pressure, standardized regression coefficients were reported. The mean values of PWV and carotid IMT were compared among subgroups with low and high non‐HDL cholesterol and systolic blood pressure. The subgroups were classified by being in the ≤40th and >60th percentiles of non‐HDL cholesterol and systolic blood pressure specific for race, sex, age, and BMI, with 20% of the sample in the middle excluded to make the two subgroups more separate. The relationship between PWV and carotid IMT was examined by Pearson correlation using sex‐, age‐, BMI‐, and/or race‐specific z scores. The difference in Pearson correlation and standardized regression coefficients between race groups and submodels was tested using Fisher's z‐transformation approach because the two statistics were identical when using a standardized scale (SD).

RESULTS

Table I shows mean levels of study variables by race and sex. BMI was significantly higher in black women than in white women; sex differences in BMI were not noted. White men showed significantly higher levels of glucose, insulin, and HOMA‐IR than white women. Significant race and sex differences were noted for blood pressure (black >white and men >women) and non‐HDL cholesterol (black <white and men >women). Race (black >white) and sex (men >women) differences were significant for carotid IMT. Race differences for PWV (black >white) were significant, but sex differences for PWV were not. In the combined sample of men and women, blacks had greater values of IMT (0.83 mm vs. 0.80 mm; p<0.01) and PWV (5.4 m/sec vs. 5.2 m/sec; p<0.01) compared with whites. Frequency of smokers (current smoking) did not significantly differ between races or sexes.

Table I.

Mean Levels and SD of Cardiovascular Risk Factors, Carotid Intima‐Media Thickness, and Arterial Stiffness by Race and Sex

	White				Black				p Value For Race
	Male (n=289)		Female (n=359)		Male (n=97)		Female (n=155)		Difference
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Male	Female
Age (yr)	36.6	4.4	36.2	4.4	36.4	4.5	35.1	4.8*	0.69	<0.01
BMI (kg/m2)	28.4	5.2	27.6	6.6	28.7	6.8	30.4	7.6	0.63	<0.01
Glucose (mg/dL)	87.0	20.6	82.1	14.4**	90.3	33.6	87.5	33.5	0.24	<0.01
Insulin (μU/mL)	12.2	8.8	10.9	7.2*	11.8	8.9	14.7	13.8	0.71	<0.01
HOMA‐IR	2.8	2.4	2.3	2.1*	2.8	2.5	3.3	3.3	0.93	<0.01
Systolic BP (mm Hg)	117.4	10.4	110.3	10.5**	126.6	15.6	117.6	15.7**	<0.01	<0.01
Diastolic BP (mm Hg)	79.8	7.5	74.6	7.8**	85.2	10.9	78.4	10.7**	<0.01	<0.01
Non‐HDLC (mg/dL)	153.9	40.1	141.1	35.7**	141.1	43.9	124.4	30.5**	0.02	<0.01
Carotid IMT (mm)	0.85	0.16	0.76	0.12**	0.88	0.17	0.79	0.15**	0.05	<0.01
PWV (m/sec)	5.24	0.79	5.16	0.81	5.54	0.97	5.28	1.15	<0.01	0.04
Smoker (%)	29.8		30.1		40.2		31.6		0.06	0.73
BMI=body mass index; HOMA‐IR=homeostasis model assessment of insulin resistance; BP=blood pressure; non‐HDLC=non–high‐density lipoprotein cholesterol; IMT=intima‐media thickness; PWV=aortofemoral pulse wave velocity; sex difference: *p<0.05; **p<0.01

Table II presents standardized regression coefficients using PWV and carotid IMT as dependent variables in separate regression models. The regression coefficient is the rate of change in PWV and carotid IMT (dependent variables) per unit change in non‐HDL cholesterol and systolic blood pressure (independent variables). The influences of non‐HDL cholesterol and systolic blood pressure on PWV and carotid IMT did not differ significantly between blacks and whites. Therefore, the impact of non‐HDL cholesterol and systolic blood pressure was examined for the total sample in multivariate regression models, adjusting for race, sex, age, BMI, HOMA‐IR, and smoking. Non‐HDL cholesterol was a significant, independent predictor of carotid IMT, but not of PWV. The magnitude of association of systolic blood pressure with PWV was significantly greater than that with carotid IMT.

Table II.

Standardized Regression Coefficients of Carotid Artery Intima‐Media Thickness (IMT) and Aorta‐Femoral Pulse Wave Velocity (PWV) on Cardiovascular Risk Variables

	White		Black		Total
	IMT	PWV	IMT	PWV	IMT	PWV
Race (blacks>whites)					0.10*	0.01
Sex (male>female)	−0.19*	0.08**	−0.16**	0.02	−0.18*	0.06
Age	0.26*	0.24*	0.26*	0.19*	0.26*	0.21*
BMI	0.05	−0.01	0.03	−0.12	0.04	−0.05
Smoking	0.14*	0.12*	0.02	0.06	0.11*	0.09*
HOMA‐IR	0.06	0.08**	−0.09	0.17**	0.01	0.11*
Systolic BP	0.15*	0.34*	0.15**	0.36*	0.15*	0.36*
Non‐HDLC	0.22*	0.01	0.13**	0.07	0.21*	0.03
BMI=body mass index; HOMA‐IR=homeostasis model assessment of insulin resistance; BP=blood pressure; non‐HDLC=non–high‐density lipoprotein cholesterol; *p<0.01; **p<0.05

Figure 1 illustrates the relationships of non‐HDL cholesterol and systolic blood pressure to PWV and carotid IMT in terms of standardized regression coefficients. In both cases, the slopes were significantly different from each other. Unlike PWV, carotid IMT increased significantly with increases in non‐HDL cholesterol. Although both PWV and carotid IMT increased significantly with higher systolic blood pressures, the slope of PWV was more than two times that of carotid IMT.

Figure 1.

Relationships of non‐high‐density lipoprotein (non‐HDL) cholesterol and systolic blood pressure (BP) to aortofemoral pulse wave velocity (PWV) and carotid intima‐media thickness (IMT) in the Bogalusa Heart Study; b=standardized regression coefficient

In Figure 2, mean values of PWV and carotid IMT in the subgroups with either high systolic blood pressure or high non‐HDL cholesterol were significantly greater than those in the low‐low group, except for PWV in the group with high non‐HDL cholesterol only. Subjects with both high systolic blood pressure and high non‐HDL cholesterol had the highest values of PWV and carotid IMT. Mean age was similar across the four subgroups because race‐, sex‐, age‐, and BMI‐specific percentiles were used for grouping.

Correlation between PWV and carotid IMT did not differ between blacks (r=0.13; p<0.05) and whites (r=0.13; p<0.01), adjusting for age, sex, and BMI. The impact of systolic blood pressure and non‐HDL cholesterol on the correlation between PWV and carotid IMT was examined separately for the total sample, adjusting for either systolic blood pressure or non‐HDL cholesterol along with race, age, sex, BMI, HOMA‐IR, and smoking. The correlation was 0.10 (p=0.004) when adjusting for non‐HDL cholesterol only and 0.06 (p=0.09) when adjusting for systolic blood pressure only.

DISCUSSION

It is well known that arterial structural and functional alterations are both related to recognized CV risk factors such as aging, hypertension, diabetes, and dyslipidemia. ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ Age and blood pressure have been shown to be consistently associated with increased arterial thickness and stiffness. ⁴ , ⁵ , ⁶ , ¹³ The impact of LDL cholesterol on carotid IMT has been established in both cross‐sectional and longitudinal studies. ⁶ , ¹² , ¹⁵ , ¹⁶ Increasing LDL cholesterol levels even in childhood are associated with changes in adulthood carotid IMT. ¹⁵ , ¹⁶ Although the association between lipoprotein cholesterol and arterial PWV has been observed in animal studies, ²² , ²³ reports on the relationship between cholesterol levels and arterial stiffness in humans are not consistent. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ HDL cholesterol was found to be inversely related to brachial‐ankle PWV in a large cohort of Japanese women. ¹³ Higher cholesterol levels resulted in stiffer carotid arteries, although the effect was less pronounced than on IMT. ²⁴ In patients with familial hypercholesterolemia, cholesterol‐lowering therapy for 1 year significantly decreased arterial wall stiffness and thickness. ¹² In contrast, other studies have found no association between cholesterol levels and arterial stiffness. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹

In the present study, we used non‐HDL cholesterol that includes cholesterol present in all apoprotein B‐containing lipoprotein particles. This is considered a better predictor of CV disease than other lipid fractions. ²⁵ The observed differential impact of non‐HDL cholesterol on aortic PWV and carotid IMT and the dependency of the relationship between arterial thickness and stiffness on blood pressure levels may partially explain the discrepancies in previous studies from populations with different blood pressure levels.

Studies focusing on the correlation between vascular thickness and stiffness have reported conflicting results. ⁹ , ¹⁰ , ¹¹ , ²⁶ , ²⁷ , ²⁸ In a large number of black and white subjects (n=10,920) aged 45–64 years from the ARIC study, a thicker common carotid artery wall was no stiffer than a thinner wall, except for the top 10% of subjects with the thickest artery. ²⁶ Arterial compliance measures are not related to carotid IMT in children with familial hypercholesterolemia. ¹⁰ On the other hand, other studies have shown significant relationships in this regard. ⁹ , ²⁸ In the present study, we found that the correlation between PWV and carotid IMT was weak, although significant, and was dependent on blood pressure levels. The finding that non‐HDL cholesterol affected arterial thickness rather than stiffness may help explain the weak or lack of correlation between arterial thickness and stiffness noted in this and other studies. ¹⁰ , ¹¹ , ²⁶ , ²⁷

The observed differential impact of non‐HDL cholesterol and blood pressure on arterial wall thickness and stiffness is consistent with the notion that changes in IMT, a measure of , are more strongly influenced by metabolic disorders than PWV. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ²⁹ Elevated PWV, a measure of sclerosis, on the other hand, results from aging and hypertensive medial elastin–collagen alterations. ³⁰ , ³¹ Further, the IMT at the early phase of atherosclerosis is related to atherogenic lipoproteins and cholesterol‐laden foam cells, which tend to decrease rather than increase PWV. ³² The hemodynamic alterations to increases in blood pressure provoke adaptive vascular remodeling and the attendant lower compliance and higher reactivity. This, in turn, augments the lipoprotein efflux across the arterial wall and the development of atherosclerosis in the presence of dyslipidemia. ³³ , ³⁴ Importantly, endothelial dysfunction caused by other CV risk factors is also an underlying factor linking adverse changes in arterial thickness and stiffness. ³⁵ , ³⁶ , ³⁷

Although the temporal relationship between arterial stiffness and elevated blood pressure has not been well established, it is generally considered that elevated blood pressure from childhood to young adulthood accelerates arterial stiffness and, in turn, arterial stiffness increases systolic blood pressure leading to isolated systolic hypertension and widening of pulse pressure in older individuals. ³ , ⁴ , ¹⁴ , ³⁵ , ³⁸ In this study, blood pressure, among the risk factors, remained the most important determinant of aortic PWV in both blacks and whites (Table II and Figure 1). Further, blacks showed significantly greater values of carotid IMT and aortic PWV than whites. It is also well known that blacks have higher blood pressure levels and stiffer arteries compared with whites. ³⁹ , ⁴⁰ Moreover, the regression line (slope) of age‐adjusted PWV on systolic blood pressure was also significantly greater in blacks than in whites. ⁴¹ However, we did not find a black‐white difference in the rate of change in PWV with increasing systolic blood pressure.

In summary, the correlation in young adults (24–43 years of age) between aortic PWV and carotid IMT was dependent on blood pressure rather than atherogenic lipoproteins measured as non‐HDL cholesterol. Non‐HDL cholesterol was significantly associated with carotid IMT but not with PWV, whereas systolic blood pressure was associated with both PWV and carotid IMT, albeit relatively stronger with the former. These results indicate that atherogenic lipoproteins and blood pressure may play different roles in the development of arterial wall stiffness and atherosclerosis. Noninvasive measures of both arterial wall thickness and stiffness in conjunction with CV risk factors can be helpful in risk assessment among high‐risk populations.

Disclosure: This study was supported by grants 0555168B from Southeast Affiliate of the American Heart Association, HD‐047247–02 from the National Institute of Child Health and Human Development, and AG‐16592 from the National Institute on Aging.