Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Abstract

Arterial stiffness (AS) and obesity are recognized as important risk factors of cardiovascular disease (CVD). The purpose of this study was to investigate the relationship between AS and obesity. AS was defined as high augmentation index (AIx) and low elasticity (C1, large artery elasticity; C2, small artery elasticity) in participants enrolled in the Multi-Ethnic Study of Atherosclerosis at baseline. We compared AIx, C1, and C2 by body mass index (BMI) (< 25, 25–29.9, 30–39.9, ⩾ 40 kg/m²) and waist–hip ratio (WHR) (< 0.85, 0.85–0.99, ⩾ 1). The obesity–AS association was tested across 10-year age intervals. Among 6177 participants (62 ± 10 years old, 52% female), a significant inverse relationship was observed between obesity and AS. After adjustments for CVD risk factors, participants with a BMI > 40 kg/m² had 5.4% lower AIx (mean difference [Δ] = −0.82%; 95% CI: –1.10, –0.53), 15.4% higher C1 (Δ = 1.66 mL/mmHg ×10; 95% CI: 1.00, 2.33), and 40.2% higher C2 (Δ = 1.49 mL/mmHg ×100; 95% CI: 1.15, 1.83) compared to those with a BMI < 25 kg/m² (all p for trend < 0.001). Participants with a WHR ⩾ 1 had 5.6% higher C1 (∆ = 0.92 mL/mmHg ×10; 95% CI: 0.47, 1.37) compared to those with a WHR < 0.85. The WHR had a significant interaction with age on AIx and C2, but not with BMI; the inverse relationships of the WHR with AIx and C2 were observed only in participants < 55 years between the normal (WHR < 0.85) and the overweight (0.85 ⩽ WHR < 0.99) groups. Different associations of WHR and BMI with arterial stiffness among older adults should be further investigated.

Keywords

obesity population health vascular medicine arterial stiffness

Introduction

Arterial stiffness is recognized as an important subclinical marker of cardiovascular disease (CVD).¹ While carotid–femoral pulse wave velocity (PWV) is considered the gold standard measure of arterial stiffness,² high augmentation index (AIx) and low elasticity (C1, large artery elasticity; C2, small artery elasticity) derived from radial artery pulse wave analysis have been used as surrogate measures of arterial stiffness. Studies have reported that high AIx was an independent marker for premature coronary artery disease³ and severe short- and long-term cardiovascular (CV) events in patients undergoing percutaneous coronary interventions.⁴ Previous studies reported that C1 and C2 are indices for elasticities of large and small arteries, respectively.⁵ While low C1 showed significant prediction for heart failure in unadjusted models, low C2 was independently associated with endothelial dysfunction,⁶ reduced kidney function,⁷ future coronary heart disease (CHD), stroke, and heart failure.⁵ C2 has also been proposed as a surrogate marker of endothelial function.⁸ These noninvasive measures of arterial stiffness have made it more feasible to identify CVD risk associated with arterial stiffness.

Obesity is a common and costly epidemiologic challenge; 39.8% of US adults were obese in 2015–2016,⁹ and the estimated annual medical cost of obesity was $147 billion in the US in 2008.¹⁰ The association of obesity with increased risk of CVD has already been documented.¹¹ Although underlying mechanisms remain unclear, studies have demonstrated that obesity, measured by body mass index (BMI) and elevated amounts of abdominal fat, is positively related to arterial stiffness measured by PWV.^12–15 Conversely, several studies have reported the so-called ‘obesity paradox’, in which obesity measured by BMI was associated with lower risk of all-cause mortality¹⁶ and better outcomes after percutaneous coronary intervention.¹⁷ An observational study has also suggested that vascular function and PWV were superior in severely obese persons compared to those who are obese, measured by BMI.¹⁸

One explanation for this apparently ‘paradoxical’ relationship is that BMI indicates lean body mass rather than adiposity, which is a stronger predictor of CV morbidity and mortality.¹⁹ A previous study reported that the obesity paradox is linked to the use of BMI, but not with waist–hip ratio (WHR).²⁰ WHR is known to be a superior measure of central adiposity to waist circumference or BMI in predicting arterial stiffness.²¹ Different types of obesity such as lean body mass or central adiposity may exert divergent effects on arterial stiffness. The objective of this study is to examine relationships between different measures of obesity (BMI and WHR) and arterial stiffness. We hypothesized that individuals with higher BMI are more likely to have lower arterial stiffness, while those with higher WHR are more likely to have higher arterial stiffness.

Methods

Study sample

The aims and design of the Multi-Ethnic Study of Atherosclerosis (MESA) have been previously published. In brief, the MESA is a multi-center cohort study of white, Chinese American, African American, and Hispanic adults living in six US cities (Winston-Salem, New York, Baltimore, Minneapolis, Chicago, and Los Angeles). The MESA cohort at baseline (MESA Exam 1) included 6814 men and women of 45–84 years of age without CVD who were recruited between 2000 and 2002. All participants provided written informed consent for study participation. The study was approved by the institutional review boards (IRBs) of all MESA field centers. This secondary analysis was approved by the University of Virginia IRB. We excluded participants who did not have C1 or C2 values (n = 478) and who had missing values of smoking (n = 14), heart rate (n = 36), low-density lipoprotein (n = 100), blood glucose (n = 4), and any other covariates (n = 6) from the final sample (n = 6177).

Assessment of obesity

Obesity was determined by BMI and WHR measured at the Exam 1 of the MESA Main Study. BMI (kg/m²) was categorized into normal (< 25), overweight (25–29.9), obesity (30–39.9), and extreme obesity (⩾ 40). WHR was categorized into three groups (< 0.85, 0.85–0.99, ⩾ 1). Previously, the thresholds 0.80 and 0.85 have been suggested for defining overweight and obesity for women and 0.90 and 1.00 for men.²² However, given the distribution of WHR in our cohort with about 75% of women being categorized as obese when 0.85 was used as the threshold, we used 0.85 and 1.00 for both men and women to define overweight and obesity in this study.

Assessment of arterial stiffness

The HDI/PulseWave CR-2000 tonometry device (Hypertension Diagnostics, Eagan, MN, USA) was used to acquire and analyze pulse waveforms from the radial arterial site on all MESA participants. Details on these measurements in MESA have been described elsewhere.²³ AIx (%) was calculated as the ratio of the secondary systolic peak to the first systolic peak multiplied by 100.²⁴ C1 and C2 are compliance of the proximal large and distal small arteries used to indicate how elastic these arteries are.⁵ They were derived by a computer-based, third-order, four-element modified Windkessel model²⁵ that analyzes the diastolic pulse contour by extracting the decay component of the waveforms and correcting for systemic vascular resistance (SVR). SVR was estimated as mean arterial blood pressure divided by cardiac output, and cardiac output was estimated from ejection time taken from the pulse waveform, heart rate, age, height, and weight.⁵

Assessment of risk factors

Other risk factors that can potentially confound the association between obesity and arterial stiffness included demographic characteristics (age, sex, race/ethnicity), lifestyle factors (smoking, physical activity), medication use for hypertension, hyperlipidemia, and diabetes, circulatory factors (systolic blood pressure, heart rate), and metabolic factors (fasting glucose, high-density lipoprotein, low-density lipoprotein, and triglyceride). Smoking status was classified as never, former, or current smoking. For physical activity levels, total intentional exercise as metabolic equivalent minutes (MET-min) per week was measured by using the MESA Typical Week Physical Activity Survey.²⁶

Statistical analysis

The relationships between obesity and arterial stiffness were examined using concurrent measures of BMI, WHR, AIx, C1, and C2. We used bivariate and multivariable analyses to examine the relationships. For multivariable analyses, we used ordinary least squares regression to model arterial stiffness (AIx, C1, and C2, dependent variables) as a function of obesity (BMI or WHR), adjusting for age, sex, race/ethnicity, smoking, physical activity, antidiabetic medications, lipid-lowering medications, antihypertensive medications, systolic blood pressure, heart rate, fasting glucose, high-density lipoprotein, low-density lipoprotein, and triglyceride.

To test whether the relationship between obesity and arterial stiffness may vary by age categories (45–54, 55–64, 65–74, and 75–84 years old), sex, and race, we conducted analysis of covariance (ANCOVA) that included interactions between obesity (BMI and WHR) and these demographic factors for all measures of arterial stiffness (AIx, C1, and C2). Least squares means were obtained for each subgroup and were tested for pairwise differences between adjacent weight categories. We used the Bonferroni correction for multiple comparison. Continuous variables were expressed as mean (SD) unless specified otherwise.

Since C1 and C2 have SVR as the denominator in the formula, in sensitivity analyses, we used C1 × SVR and C2 × SVR as additional outcomes to separate the information obtained from the pulse waveform from physical measures (SVR). All analyses were performed using Stata SE, Release 15 (StataCorp, College Station, TX, USA) and SPSS Statistics, Version 24 (IBM, Armonk, NY, USA).

Results

After excluding missing observations, the study sample included 6177 participants. All participants included in the sample had C1 and C2 values, but AIx was not able to be computed for 362 of them. Thus, all analyses involving AIx were based on 5815 participants. The sample was 62 ± 10 years old at the time of MESA Exam 1, 48% were male, 38% were white, 12% Chinese American, 27% African American, and 23% Hispanic. The average AIx was 14.88 ± 2.2%, C1 was 13.36 ± 5.6 mL/mmHg × 10, and C2 was 4.48 ± 2.8 mL/mmHg × 100 (Table 1).

Table 1.

Characteristics of the study participants: Multi-Ethnic Study of Atherosclerosis, 2000–2002.

Characteristic	Value
Total patients, n (row %)	6177 (100.0%)
AIx (%), mean (SD)	14.88 (2.2)
C1 (mL/mmHg ×10), mean (SD)	13.36 (5.6)
C2 (mL/mmHg ×100), mean (SD)	4.48 (2.8)
SVR (dynes/sec/cm⁵), mean (SD)	1648.09 (398.11)
C1 × SVR (seconds), mean (SD)	1.57 (0.58)
C2 × SVR (seconds), mean (SD)	0.51 (0.26)
Age, mean (SD)	62.07 (10.2)
Male sex	2941 (47.6%)
Race/ethnicity
White	2342 (37.9%)
Chinese American	756 (12.2%)
African American	1682 (27.2%)
Hispanic	1397 (22.6%)
Education
⩽ High school	2237 (36.21%)
Some college	1759 (28.48%)
College	1063 (17.21%)
Postgraduate	1118 (18.10%)
Cigarette smoking
Never	3140 (50.8%)
Former	2256 (36.5%)
Current	781 (12.6%)
Antidiabetic medications	601 (9.7%)
Lipid-lowering medications	980 (15.9%)
Antihypertensive medications	2036 (32.9%)
SBP (mmHg), mean (SD)	126.43 (21.4)
DBP (mmHg), mean (SD)	71.96 (10.23)
Heart rate (beats/min), mean (SD)	63.03 (9.6)
LDL cholesterol (mg/dL), mean (SD)	117.23 (31.4)
HDL cholesterol (mg/dL), mean (SD)	51.02 (14.6)
Triglycerides (mg/dL), mean (SD)	126.62 (65.8)
Total intentional exercise (MET-min/week), mean (SD)	25.67 (38.3)
Fasting glucose (mg/dL), mean (SD)	97.14 (29.8)
Study site
Wake Forest	1024 (16.58%)
Columbia	1004 (16.25%)
Johns Hopkins	825 (13.36%)
Minnesota	972 (15.74%)
Northwestern	1101 (17.82%)
UCLA	1251 (20.25%)

Data are presented as n (%) unless otherwise noted. AIx, augmentation index; C1, large artery elasticity; C2, small artery elasticity; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MET, metabolic equivalent; SBP, systolic blood pressure; SVR, systemic vascular resistance.

C1 × SVR and C2 × SVR were converted to be in units of seconds by multiplying them by 0.00000750064.

Table 2 shows a bivariate analysis of participant characteristics according to BMI groups. The majority of participants were classified in the Normal (28.8%), Overweight (39.3%), or Obese (28.3%) groups, with only 3.6% classified as the Very Obese group. The unadjusted average AIx linearly decreased from 15.06% to 14.37% from the Normal group to the Very Obese group (p < 0.001). C1 was highest in the Overweight group (13.61 mL/mmHg × 10) and lowest in the Very Obese group (12.92 mL/mmHg × 10). C2 increased linearly from 4.22 to 5.21 mL/mmHg × 100 from the Normal to the Very Obese groups (p < 0.001).

Table 2.

Study participant characteristics in function of BMI categories (n = 6177): Multi-Ethnic Study of Atherosclerosis, 2000–2002.

Variable	Normal	Overweight	Obese	Very Obese	p-value
Total patients, n (row %)	1780 (28.82%)	2426 (39.27%)	1747 (28.28%)	224 (3.63%)
AIx (%), mean (SD)	15.06 (2.28)	14.90 (2.18)	14.74 (2.11)	14.37 (2.11)	< 0.001
C1 (mL/mmHg ×10), mean (SD)	13.03 (5.53)	13.61 (5.85)	13.41 (5.43)	12.92 (4.82)	0.006
C2 (mL/mmHg ×100), mean (SD)	4.22 (2.84)	4.44 (2.82)	4.70 (2.82)	5.21 (2.49)	< 0.001
SVR (dynes/sec/cm⁵), mean (SD)	1769.23 (458.78)	1669.01 (375.17)	1533.21 (319.96)	1354.77 (257.77)	< 0.001
C1 × SVR (seconds), mean (SD)	1.63 (0.60)	1.62 (0.61)	1.47 (0.51)	1.27 (0.42)	< 0.001
C2 × SVR (seconds), mean (SD)	0.51 (0.28)	0.51 (0.26)	0.51 (0.25)	0.50 (0.19)	0.868
Age at Exam 1, mean (SD)	62.56 (10.62)	62.70 (10.26)	61.24 (9.75)	57.86 (8.97)	< 0.001
Male sex	799 (44.89%)	1326 (54.66%)	773 (44.25%)	43 (19.20%)	< 0.001
Race/ethnicity
White	754 (42.36%)	945 (38.95%)	587 (33.60%)	56 (25.00%)	< 0.001
Chinese American	492 (27.64%)	234 (9.65%)	29 (1.66%)	1 (0.45%)
African American	297 (16.69%)	627 (25.85%)	652 (37.32%)	106 (47.32%)
Hispanic	237 (13.31%)	620 (25.56%)	479 (27.42%)	61 (27.23%)
Education
⩽ High school	563 (31.63%)	913 (37.63%)	676 (38.69%)	85 (37.95%)	< 0.001
Some college	474 (26.63%)	643 (26.50%)	551 (31.54%)	91 (40.63%)
College graduation	360 (20.22%)	409 (16.86%)	270 (15.46%)	24 (10.71%)
Postgraduate	383 (21.52%)	461 (19.00%)	250 (14.31%)	24 (10.71%)
Cigarette smoking
Never	984 (55.28%)	1 189 (49.01%)	856 (49.00%)	111 (49.55%)	< 0.001
Former	568 (31.91%)	918 (37.84%)	680 (38.92%)	90 (40.18%)
Current	228 (12.81%)	319 (13.15%)	211 (12.08%)	23 (10.27%)
Taking medications for diabetes	94 (5.28%)	206 (8.49%)	256 (14.65%)	45 (20.09%)	< 0.001
Taking medications for cholesterol	193 (10.84%)	426 (17.56%)	324 (18.55%)	37 (16.52%)	< 0.001
Taking medications for hypertension	381 (21.40%)	785 (32.36%)	760 (43.50%)	110 (49.11%)	< 0.001
SBP (mmHg), mean (SD)	121.40 (21.99)	127.16 (20.77)	129.82 (20.37)	132.19 (22.46)	< 0.001
Heart rate (beats/min), mean (SD)	62.08 (9.31)	62.27 (9.32)	64.46 (9.83)	67.69 (10.40)	< 0.001
LDL cholesterol (mg/dL), mean (SD)	114.90 (30.93)	118.68 (31.72)	117.53 (31.14)	117.59 (33.40)	0.002
HDL cholesterol (mg/dL), mean (SD)	56.70 (16.70)	49.54 (13.23)	47.66 (12.90)	48.11 (11.13)	< 0.001
Triglycerides (mg/dL), mean (SD)	110.13 (59.16)	129.84 (65.83)	138.11 (68.99)	133.19 (64.79)	< 0.001
Total intentional exercise (MET-min/week), mean (SD)	27.67 (39.11)	27.87 (41.17)	22.18 (34.49)	13.24 (20.57)	< 0.001
Fasting glucose (mg/dL), mean (SD)	91.40 (26.59)	96.00 (26.35)	102.97 (34.61)	109.43 (37.40)	< 0.001
Study site
Wake Forest	240 (13.48%)	411 (16.94%)	330 (18.89%)	43 (19.20%)	< 0.001
Columbia	253 (14.21%)	409 (16.86%)	297 (17.00%)	45 (20.09%)
Johns Hopkins	173 (9.72%)	327 (13.48%)	282 (16.14%)	43 (19.20%)
Minnesota	185 (10.39%)	385 (15.87%)	360 (20.61%)	42 (18.75%)
Northwestern	463 (26.01%)	402 (16.57%)	215 (12.31%)	21 (9.38%)
UCLA	466 (26.18%)	492 (20.28%)	263 (15.05%)	30 (13.39%)

AIx, augmentation index; BMI, body mass index; C1, large artery elasticity; C2, small artery elasticity; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MET, metabolic equivalent; SBP, systolic blood pressure; SVR, systemic vascular resistance.

C1 × SVR and C2 × SVR were converted to be in units of seconds by multiplying them by 0.00000750064.

Table 3 shows a bivariate analysis according to WHR categories. The majority of the participants (66%) fell into the Overweight group (WHR 0.85–0.99), while the number of participants in the Normal (WHR < 0.85) and Obese (⩾ 1.00) groups were similar (16.2%, 17.7%, respectively). The average AIx tended to decrease linearly from 15.36% to 14.54% from the Normal to Obese groups (p < 0.001). C1 was the highest in the Overweight group but it tended to increase across all three groups (p = 0.006). The average C2 showed a nonlinear trend across the WHR groups, but the difference was not statistically significant (p = 0.188).

Table 3.

Study participant characteristics in function of WHR categories (n = 6177): Multi-Ethnic Study of Atherosclerosis, 2000–2002.

Variable	< 0.85	0.85–0.99	⩾ 1.00	p-value
Total patients, n (row %)	1002 (16.22%)	4079 (66.04%)	1096 (17.74%)
AIx (%), mean (SD)	15.36 (2.36)	14.85 (2.17)	14.54 (2.06)	< 0.001
C1 (mL/mmHg ×10), mean (SD)	12.89 (5.14)	13.51 (5.60)	13.25 (6.03)	0.006
C2 (mL/mmHg ×100), mean (SD)	4.40 (2.59)	4.53 (2.91)	4.38 (2.70)	0.188
SVR (dynes/sec/cm⁵), mean (SD)	1638.46 (437.66)	1650.06 (390.38)	1649.52 (388.81)	0.704
C1 × SVR (seconds), mean (SD)	1.50 (0.53)	1.59 (0.58)	1.56 (0.63)	< 0.001
C2 × SVR (seconds), mean (SD)	0.50 (0.25)	0.51 (0.27)	0.50(0.25)	0.098
Age at Exam 1, mean (SD)	58.55 (9.93)	62.19 (10.23)	64.84 (9.57)	< 0.001
Male sex	90 (8.98%)	2136 (52.37%)	715 (65.24%)	< 0.001
Race/ethnicity
White	474 (47.31%)	1468 (35.99%)	400 (36.50%)	< 0.001
Chinese American	111 (11.08%)	571 (14.00%)	74 (6.75%)
African American	303 (30.24%)	1137 (27.87%)	242 (22.08%)
Hispanic	114 (11.38%)	903 (22.14%)	380 (34.67%)
Education
⩽ High school	257 (25.65%)	1 470 (36.04%)	510 (46.53%)	< 0.001
Some college	298 (29.74%)	1161 (28.46%)	300 (27.37%)
College graduation	228 (22.75%)	702 (17.21%)	133 (12.14%)
Postgraduate	219 (21.86%)	746 (18.29%)	153 (13.96%)
Cigarette smoking
Never	557 (55.59%)	2087 (51.16%)	496 (45.26%)	< 0.001
Former	313 (31.24%)	1482 (36.33%)	461 (42.06%)
Current	132 (13.17%)	510 (12.50%)	139 (12.68%)
Taking medications for diabetes	27 (2.69%)	369 (9.05%)	205 (18.70%)	< 0.001
Taking medications for cholesterol	96 (9.58%)	660 (16.18%)	224 (20.44%)	< 0.001
Taking medications for hypertension	210 (20.96%)	1350 (33.10%)	476 (43.43%)	< 0.001
SBP (mmHg), mean (SD)	120.12 (22.17)	126.84 (21.05)	130.70 (20.47)	< 0.001
Heart rate (beats/min), mean (SD)	62.52 (8.75)	62.66 (9.57)	64.88 (10.21)	< 0.001
LDL cholesterol (mg/dL), mean (SD)	114.70 (30.47)	118.30 (31.46)	115.53 (31.97)	< 0.001
HDL cholesterol (mg/dL), mean (SD)	61.72 (16.00)	49.88 (13.72)	45.48 (11.52)	< 0.001
Triglycerides (mg/dL), mean (SD)	97.61 (49.87)	128.30 (65.77)	146.90 (69.55)	< 0.001
Total intentional exercise (MET-min/week), mean (SD)	27.43 (39.64)	26.21 (38.17)	22.07 (37.55)	0.002
Fasting glucose (mg/dL), mean (SD)	86.35 (18.50)	97.08 (27.67)	107.19 (40.64)	< 0.001
Study site
Wake Forest	186 (18.56%)	698 (17.11%)	140 (12.77%)	< 0.001
Columbia	141 (14.07%)	661 (16.20%)	202 (18.43%)
Johns Hopkins	179 (17.86%)	558 (13.68%)	88 (8.03%)
Minnesota	116 (11.58%)	613 (15.03%)	243 (22.17%)
Northwestern	256 (25.55%)	673 (16.50%)	172 (15.69%)
UCLA	124 (12.38%)	876 (21.48%)	251 (22.90%)

AIx, augmentation index; C1, large artery elasticity; C2, small artery elasticity; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MET, metabolic equivalent; SBP, systolic blood pressure; SVR, systemic vascular resistance; WHR, waist–hip ratio.

C1 × SVR and C2 × SVR were converted to be in units of seconds by multiplying them by 0.00000750064.

Multivariable analyses that controlled for all potential confounders are summarized in Tables 4 and 5 for BMI and WHR, respectively, with full models shown in online supplemental Tables A and B. Compared to the Normal group, the Overweight, Obese, and Very Obese groups showed lower average AIx, where the mean differences were –0.19% (95% CI, –0.32, –0.06), –0.38% (95% CI, –0.53, –0.23), and –0.82% (95% CI, –1.10 –0.53), respectively. This represents a 1.2%, 2.4%, and 5.4% decrease in AIx from the Normal group (p for trend < 0.001).

Table 4.

Adjusted mean differences in arterial stiffness in Grade 1–3 Obesity groups compared to the Normal weight (n = 5814)^a: Multi-Ethnic Study of Atherosclerosis, 2000–2002.

Measures of arterial stiffness	BMI categories (vs Normal)						p for trend
	Overweight		Obese		Very Obese
	Mean difference (95% CI)		Mean difference (95% CI)		Mean difference (95% CI)
AIx (%)	−0.189	(−0.317, −0.061)	−0.378	(−0.527, −0.230)	−0.816	(−1.102, −0.531)	< 0.001
C1 (mL/mmHg ×10)	0.777	(0.483, 1.071)	1.221	(0.880, 1.561)	1.662	(1.000, 2.325)	< 0.001
C2 (mL/mmHg ×100)	0.308	(0.157, 0.459)	0.741	(0.566, 0.916)	1.494	(1.153, 1.834)	< 0.001

Adjusted mean differences were estimated using ordinary least squares regression. The models were adjusted for age, sex, race/ethnicity, smoking, physical activity, antidiabetic medications, lipid-lowering medications, antihypertensive medications, systolic blood pressure, heart rate, fasting glucose, high-density lipoprotein, low-density lipoprotein, and triglyceride.

The full model is shown in Table A in the online supplemental material.

AIx, augmentation index; BMI, body mass index. C1, large artery elasticity; C2, small artery elasticity.

Table 5.

Adjusted mean differences in arterial stiffness in WHR groups (n = 6177)^a: Multi-Ethnic Study of Atherosclerosis, 2000–2002.

Measures of arterial stiffness	WHR groups (vs < 0.85)				p for trend
	0.85–0.99		⩾ 1.00
	Mean difference (95% CI)		Mean difference (95% CI)
AIx (%)	−0.106	(−0.257, 0.044)	−0.179	(−0.374, 0.016)	0.073
C1 (mL/mmHg ×10)	0.531	(0.185, 0.877)	0.920	(0.472, 1.368)	< 0.001
C2 (mL/mmHg ×100)	0.169	(−0.010, 0.348)	0.230	(−0.001, 0.462)	0.061

The full model is shown in Table B in the online supplemental material.

AIx, augmentation index; C1, large artery elasticity; C2, small artery elasticity; WHR, waist–hip ratio.

C1 and C2 also showed significant graded relationships with BMI. Compared to the normal group, C1 increased by an average of 0.78 mL/mmHg × 10 (95% CI: 0.48, 1.07), 1.22 (95% CI: 0.88, 1.56), and 1.66 (95% CI: 1.00, 2.33) in the Overweight, Obese, and Very Obese groups, which represents a 6%, 11%, and 15% increase over the Normal group (p for trend < 0.001). C2 showed the strongest graded relationship with BMI among all measures of arterial stiffness. Compared to the Normal group, the mean difference was 0.31 mL/mmHg × 100 (95% CI: 0.16, 0.46), 0.74 (95% CI: 0.57, 0.92), and 1.49 (95% CI: 1.15, 1.83) for Overweight, Obese, and Very Obese groups. These represented a 9%, 21%, and 40% increase over the Normal group (p for trend < 0.001).

As seen in Table 5, the multivariable analysis for WHR showed similar graded relationships with arterial stiffness. The mean difference in C1 between Normal and Overweight groups was 0.53 mL/mmHg × 10 (95% CI: 0.19, 0.88) and between Normal and Obese groups was 0.92 (95% CI: 0.47, 1.37). These represented a 3.4% and 5.6% increase over the Normal group (p for trend < 0.001).

We further investigated potential heterogeneous obesity–arterial stiffness relationships in participant subgroups by age, sex, and race. Although there were significant interaction effects between BMI and sex on C1 and C2, the trend of lower arterial stiffness associated with higher BMI remained unchanged. For example, the inverse relationship between BMI and arterial stiffness was stronger in men than in women. A significant interaction between WHR and race on C1 was observed; the significant inverse relationship between WHR and C1 was observed in white, black, and Chinese American groups but not in the Hispanic group. Further, we found significant divergent relationships between WHR and AIx and between WHR and C2 by age groups, as shown in Figure 1. The inverse relationships between WHR and AIx and between WHR and C2 were only significant among those 45–54 years of age. No such interaction by age was found between BMI and any of the arterial stiffness measures.

Figure 1.

Arterial stiffness by WHR and age categories^a: Multi-Ethnic Study of Atherosclerosis, 2000–2002. (A) AIx (%); (B) C2 (mL/mmHg ×100).

The sensitivity analyses showed that the patterns of the associations of C2 and C2 × SVR with WHR were similar in that both C2 and C2 × SVR were increasing for participants aged < 65 and decreasing for those aged 65 or older. Their adjusted least squares means are plotted in the online supplemental Figure A. Full results of the sensitivity analyses are presented in online supplemental Tables D and E.

Discussion

In this ethnically diverse community cohort, we found that higher obesity, as measured by both BMI and WHR, is significantly associated with lower arterial stiffness, as defined by high AIx and low C1 and C2, independent of common CVD risk factors. In particular, BMI had significant graded relationships with all three measures of arterial stiffness. The strongest relationship was observed between BMI and C2 with the Obese and Very Obese groups having values 21% and 40% higher than the Normal group. As for WHR, there was significant effect modification by age group in the association between WHR-based obesity and arterial stiffness in that such an inverse association exists only among persons < 55 years of age between normal (WHR < 0.85) and Overweight (0.85 ⩽ WHR < 0.99) groups.

Our results contrast with several previous studies showing obesity as a risk factor for increased arterial stiffness. A recent meta-analysis reported a significantly higher arterial stiffness measured by carotid–femoral PWV, brachial–ankle PWV, and AIx in healthy obese and overweight subjects.²⁷ This positive relationship between BMI and arterial stiffness occurs early in life, as seen in children²⁸ and young adults,²⁹ in hypertensive³⁰ and nonhypertensive individuals.³¹ Central or visceral adiposity measured by WHR was also significantly associated with increased arterial stiffness in individuals with African and Asian ancestry³² and in the elderly.³³ Further, one study of more than 1500 participants reported that WHR was independently related to increased arterial stiffness measured by carotid–femoral PWV in men.³⁴

On the other hand, there are studies showing inverse relationships between obesity and arterial stiffness. Higher BMI was associated with lower arterial stiffness measured by PWV among Africans.³⁵ Beyond reduced arterial stiffness, an increasing number of studies are reporting favorable CV and other health outcomes in obese individuals with existing CVD, further supporting the obesity paradox. For example, higher BMI was also associated with better progress following coronary artery bypass grafting,³⁶ reduced risk of CV events and complications after percutaneous coronary interventions,³⁷ lower mortality among patients with CHD,³⁸ reduced risk of major amputation and death in a large cohort of nonelderly men with diabetes,³⁹ better vascular function measured by flow-mediated dilation and carotid intima–media thickness in healthy subjects.¹⁸

It is possible that BMI as the measure of obesity may have contributed to inconsistencies in the reported relationships between obesity and arterial stiffness. BMI may not be a good measure of adiposity, since it does not differentiate between excess fat, muscle, or bone mass.⁴⁰ One study that included a subset of the data reported here demonstrated that BMI and total abdominal muscle tissue were associated with lower arterial stiffness (i.e. higher C2), whereas visceral adipose tissue was associated with higher arterial stiffness (i.e. lower C2), showing visceral fat may be a better marker than BMI to estimate detrimental effects of excess adiposity on arterial stiffness.⁴¹ Muscle mass seems to have an inverse relationship with arterial stiffness. Reduced muscle mass and increased visceral fat were independently associated with increased arterial stiffness in postmenopausal women with type II diabetes.⁴² Further, higher PWV was associated with muscle mass decline⁴³ and low thigh muscle mass among middle-aged to elderly men⁴⁴ but not women.³³

However, these inconsistent results may not be entirely attributed to inaccurate measurement of adiposity. Recent studies show that arterial stiffness is paradoxically low not only in people with higher BMI but also in people with higher body fat accumulation. In one study, all the measures of adiposity including BMI, WHR, body fat percentage, and body roundness index were inversely associated with arterial stiffness measured by brachial–ankle PWV and cardio–ankle vascular index (CAVI).⁴⁵ Another study reported that WHR was inversely associated with carotid–femoral PWV in an adjusted model.³⁴ Our study findings are consistent with these previous studies and showed the inverse associations between different measures of adiposity, including BMI and WHR, and arterial stiffness measured by AIx, C1, and C2, independent of demographic factors and comprehensive CV risk factors. To better understand our findings, we took age into consideration and explored the age interaction of BMI and WHR on arterial stiffness.

It is noteworthy that the inverse association of WHR on arterial stiffness is present only among people aged less than 55 years old, while there is no age interaction with BMI on arterial stiffness. This is consistent with a previous study reporting an inverse relationship between body fat percentage and aortic stiffness in a younger age group (vs a positive relationship in an older age group). Similar to our finding, no such divergent interaction with age was found with BMI.⁴⁶ Among the elderly, the inverse association of adiposity (WHR) on arterial stiffness noted in younger (or nonelderly) individuals may be counteracted by other factors, such as decreased insulin sensitivity. Insulin resistance, which is common with aging⁴⁷ and central obesity,⁴⁸ is associated with arterial stiffness.⁴⁹ Actions of glucocorticoids, which can be considered hormones of aging,⁵⁰ may also play a role in the inverse relationship between WHR and arterial stiffness that disappears in the elderly. Glucocorticoids cause central fat deposition and peripheral fat depletion, and are known to be associated with adiposity-derived inflammatory factors.^51,52 There is strong evidence on the association between subclinical inflammation and arterial stiffness in the general population.⁵³ On the other hand, the BMI measure includes weight of muscle mass, which is associated with less arterial stiffness.⁴³ Our finding of the inverse relationship between BMI and arterial stiffness that remained in old age groups may reflect that elderly individuals with more muscle mass have less arterial stiffness than their less muscular counterparts.

Strengths and limitations

The strength of the present study includes a large sample size, racial/ethnic diversity, the use of reliable measures of arterial stiffness, and adjustment for comprehensive CV risk factors. Several limitations must also be noted. The cross-sectional design limits the ability to establish a temporal order in the relationship between obesity and arterial stiffness. The possibility of residual confounding exists, despite adjusting for several potential confounders. For example, a study based on the same cohort and other studies have shown that sleep apnea is associated with arterial stiffness^54,55 but this information was not available as sleep study was not performed at the time of Exam 1. The limitation regarding the measures of arterial stiffness is important to note. The gold standard measure of arterial stiffness is carotid–femoral PWV, but this measure was not available at Exam 1. Survival bias is also one of the limitations although the risk of cardiovascular events prior to age 45 is fairly low. Further, the choice of 0.85 and 1.00 to define WHR categories may make our findings less generalizable according to the conventional thresholds. Finally, our results should be interpreted cautiously because the computation of C1 and C2 involves SVR, which is an estimate from physical measures including height and weight. The unexpected association of obesity with C1 and C2 may have been affected by SVR.

Conclusion

This study demonstrated that the subjects with higher BMI and WHR were more likely to have less arterial stiffness defined as low AIx and high C1 and C2, independent of common CVD risk factors, alluding to the obesity paradox. Our findings illustrate the complexity of the relationship between body habitus and arterial stiffness. Further research should investigate underlying mechanisms of divergent relationships of BMI and WHR on arterial stiffness among the elderly. Considering the prevalence of obesity and the prognostic value of arterial stiffness in the future development of CVD, explicating the elusive association between obesity and arterial stiffness will have significant implications for the early detection and prevention of CVD.

Supplemental Material

Supplementary_Figure_A – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Figure_A for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Supplemental Material

Supplementary_Table_A – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Table_A for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Supplemental Material

Supplementary_Table_B – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Table_B for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Supplemental Material

Supplementary_Table_C – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Table_C for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Supplemental Material

Supplementary_Table_D – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Table_D for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Supplemental Material

Supplementary_Table_E – Supplemental material for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA)

Supplemental material, Supplementary_Table_E for Association of obesity with arterial stiffness: The Multi-Ethnic Study of Atherosclerosis (MESA) by Jeongok G. Logan, Hyojung Kang, Soyoun Kim, Daniel Duprez, Younghoon Kwon, David R. Jacobs, Nketi Forbang, Jennifer Mason Lobo and Min-Woong Sohn in Vascular Medicine

Footnotes

Acknowledgements

The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at .

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this research was supported by contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS).

ORCID iD

Min-Woong Sohn

Supplementary material

The supplementary material is available online with the article.

References

Niiranen

Kalesan

Hamburg

, et al. Relative contributions of arterial stiffness and hypertension to cardiovascular disease: The Framingham Heart Study. J Am Heart Assoc 2016; 5: e004271.

Van Bortel

Laurent

Boutouyrie

, et al. Expert consensus document on the measurement of aortic stiffness in daily practice using carotid-femoral pulse wave velocity. J Hypertens 2012; 30: 445–448.

Weber

Auer

O’Rourke

, et al. Arterial stiffness, wave reflections, and the risk of coronary artery disease. Circulation 2004; 109: 184–189.

Weber

Auer

O’Rourke

, et al. Increased arterial wave reflections predict severe cardiovascular events in patients undergoing percutaneous coronary interventions. Eur Heart J 2005; 26: 2657–2663.

Duprez

Jacobs

Lutsey

, et al. Association of small artery elasticity with incident cardiovascular disease in older adults: The Multi-Ethnic Study of Atherosclerosis. Am J Epidemiol 2011; 174: 528–536.

Tao

Jin

Y-F

Yang

, et al. Reduced arterial elasticity is associated with endothelial dysfunction in persons of advancing age: Comparative study of noninvasive pulse wave analysis and laser Doppler blood flow measurement. Am J Hypertens 2004; 17: 654–659.

Peralta

Katz

Madero

, et al. The differential association of kidney dysfunction with small and large arterial elasticity: The Multiethnic Study of Atherosclerosis. Am J Epidemiol 2009; 169: 740–748.

Gilani

Kaiser

Bratteli

, et al. Role of nitric oxide deficiency and its detection as a risk factor in pre-hypertension. J Am Soc Hypertens 2007; 1: 45–55.

Hales

Carroll

Fryar

, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief 2017; (288): 1–8.

10.

Finkelstein

Trogdon

Cohen

, et al. Annual medical spending attributable to obesity: Payer- and service-specific estimates. Health Aff (Millwood) 2009; 28(5): w822–831.

11.

Poirier

Giles

Bray

, et al. Obesity and cardiovascular disease: Pathophysiology, evaluation, and effect of weight loss: An update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism. Circulation 2006; 113: 898–918.

12.

Kim

Kyung

Park

, et al. Normal-weight obesity is associated with increased risk of subclinical atherosclerosis. Cardiovasc Diabetol 2015; 14: 58.

13.

Strasser

Arvandi

Pasha

, et al. Abdominal obesity is associated with arterial stiffness in middle-aged adults. Nutr Metab Cardiovasc Dis 2015; 25: 495–502.

14.

Corrigan

3rd Kelli

Dhindsa

, et al. Changes in truncal obesity and fat distribution predict arterial health. J Clin Lipidol 2017; 11: 1354–1360.e3.

15.

Arkin

Alsdorf

Bigornia

, et al. Relation of cumulative weight burden to vascular endothelial dysfunction in obesity. Am J Cardiol 2008; 101: 98–101.

16.

Azhari

Ismail

Zuhdi

ASM

, et al. Association between body mass index and outcomes after percutaneous coronary intervention in multiethnic South East Asian population: A retrospective analysis of the Malaysian National Cardiovascular Disease Database–Percutaneous Coronary Intervention (NCVD-PCI) registry. BMJ Open 2017; 7: e017794.

17.

Patel

Elsaid

Shenoy

, et al. Obesity paradox in patients undergoing coronary intervention: A review. World J Cardiol 2017; 9: 731–736.

18.

Biasucci

Graziani

Rizzello

, et al. Paradoxical preservation of vascular function in severe obesity. Am J Med 2010; 123: 727–734.

19.

Lalazar

. Central obesity: Redefining normal BMI. Sci Transl Med 2015; 7: 316ec209.

20.

Antonopoulos

Tousoulis

. The molecular mechanisms of obesity paradox. Cardiovasc Res 2017; 113: 1074–1086.

21.

Lee

Hwang

Hong

, et al. Waist-to-hip ratio is better at predicting subclinical atherosclerosis than body mass index and waist circumference in postmenopausal women. Maturitas 2015; 80: 323–328.

22.

Nishida

Kumanyika

. Body fat distribution and noncommunicable diseases in populations: Overview of the 2008 WHO Expert Consultation on Waist Circumference and Waist-Hip Ratio. Eur J Clin Nutr 2010; 64: 2–5.

23.

Duprez

Jacobs

Lutsey

, et al. Race/ethnic and sex differences in large and small artery elasticity—Results of the Multi-Ethnic Study of Atherosclerosis (MESA). Ethn Dis 2009; 19: 243–250.

24.

Al-Naamani

Chirinos

Zamani

, et al. Association of systemic arterial properties with right ventricular morphology: The Multi-Ethnic Study of Atherosclerosis (MESA)-Right Ventricle Study. J Am Heart Assoc 2016; 5: e004162.

25.

Duprez

De Buyzere

MML

De Bruyne

, et al. Small and large artery elasticity indices in peripheral arterial occlusive disease (PAOD). Vasc Med 2001; 6: 211–214.

26.

Aaron

Tandri

Barr

, et al. Physical activity and right ventricular structure and function. Am J Respir Crit Care Med 2011; 183: 396–404.

27.

Wang

Liu

. Overweightness, obesity and arterial stiffness in healthy subjects: A systematic review and meta-analysis of literature studies. Postgrad Med 2017; 129: 224–230.

28.

Haapala

Lankhorst

de Groot

, et al. The associations of cardiorespiratory fitness, adiposity and sports participation with arterial stiffness in youth with chronic diseases or physical disabilities. Eur J Prev Cardiol 2017; 24: 1102–1111.

29.

Fernberg

Fernström

Hurtig-Wennlöf

. Arterial stiffness is associated to cardiorespiratory fitness and body mass index in young Swedish adults: The Lifestyle, Biomarkers, and Atherosclerosis study. Eur J Prev Cardiol 2017; 24: 1809–1818.

30.

Deepika

Vijayakumar

. Impact of body mass index on arterial stiffness in young prehypertensives: A cross sectional study. J Res Health Sci 2017; 18: e00402.

31.

Son

W-M

Kim

D-Y

Kim

Y-S

, et al. Effect of obesity on blood pressure and arterial stiffness in middle-aged Korean women. Osong Public Health Res Perspect 2017; 8: 369–372.

32.

Diemer

Brewster

Haan

, et al. Body composition measures and cardiovascular risk in high-risk ethnic groups. Clin Nutr 2019; 38: 450–456.

33.

Kohara

Ochi

Tabara

, et al. Arterial stiffness in sarcopenic visceral obesity in the elderly: J-SHIPP study. Int J Cardiol 2012; 158: 146–148.

34.

Van den Munckhof

ICL

Holewijn

de Graaf

, et al. Sex differences in fat distribution influence the association between BMI and arterial stiffness. J Hypertens 2017; 35: 1219–1225.

35.

Huisman

Schutte

Venter

, et al. Low BMI is inversely associated with arterial stiffness in Africans. Br J Nutr 2015; 113: 1621–1627.

36.

Reeves

Ascione

Chamberlain

, et al. Effect of body mass index on early outcomes in patients undergoing coronary artery bypass surgery. J Am Coll Cardiol 2003; 42: 668–676.

37.

Gurm

Brennan

Booth

, et al. Impact of body mass index on outcome after percutaneous coronary intervention (the obesity paradox). Am J Cardiol 2002; 90: 42–45.

38.

De Schutter

Kachur

Lavie

, et al. The impact of inflammation on the obesity paradox in coronary heart disease. Int J Obes (Lond) 2016; 40: 1730–1735.

39.

Sohn

M-W

Budiman-Mak

, et al. Obesity paradox in amputation risk among nonelderly diabetic men. Obesity (Silver Spring) 2012; 20: 460–462.

40.

Daniels

. The use of BMI in the clinical setting. Pediatrics 2009; 124(Suppl 1): S35–S41.

41.

Forbang

Allison

, et al. Associations of body composition measures and C2, a marker for small artery elasticity: The MESA. Obesity (Silver Spring) 2015; 23: 2294–2298.

42.

Tanaka

Kanazawa

Sugimoto

. Reduced muscle mass and accumulation of visceral fat are independently associated with increased arterial stiffness in postmenopausal women with type 2 diabetes mellitus. Diabetes Res Clin Pract 2016; 122: 141–147.

43.

Abbatecola

Chiodini

Gallo

, et al. Pulse wave velocity is associated with muscle mass decline: Health ABC study. Age (Dordr) 2012; 34: 469–478.

44.

Ochi

Kohara

Tabara

, et al. Arterial stiffness is associated with low thigh muscle mass in middle-aged to elderly men. Atherosclerosis 2010; 212: 327–332.

45.

Gomez-Sanchez

Garcia-Ortiz

Patino-Alonso

, et al. Adiposity measures and arterial stiffness in primary care: The MARK prospective observational study. BMJ Open 2017; 7: e016422.

46.

Corden

Keenan

de Marvao

ASM

, et al. Body fat is associated with reduced aortic stiffness until middle age. Hypertension 2013; 61: 1322–1327.

47.

Scheen

. Diabetes mellitus in the elderly: Insulin resistance and/or impaired insulin secretion? Diabetes Metab 2005; 31(Spec No. 2): 5S27–5S34.

48.

Westphal

. Obesity, abdominal obesity, and insulin resistance. Clin Cornerstone 2008; 9: 23–29; discussion 30–31.

49.

Urbina

Gao

Khoury

, et al. Insulin resistance and arterial stiffness in healthy adolescents and young adults. Diabetologia 2012; 55: 625–631.

50.

Goudochnikov

. The role of glucocorticoids in aging and age-related pharmacotherapy. Adv Gerontol 2011; 24: 48–53.

51.

Lee

M-J

Pramyothin

Karastergiou

, et al. Deconstructing the roles of glucocorticoids in adipose tissue biology and the development of central obesity. Biochim Biophys Acta 2014; 1842: 473–481.

52.

Tanaka

Kusunoki

, et al. Resistin is associated with the inflammation process in patients with systemic autoimmune diseases undergoing glucocorticoid therapy: Comparison with leptin and adiponectin. Mod Rheumatol 2013; 23: 8–18.

53.

Jain

Khera

Corrales-Medina

, et al. Inflammation and arterial stiffness in humans. Atherosclerosis 2014; 237: 381–390.

54.

Kwon

Jacobs

Lutsey

, et al. Sleep disordered breathing and ECG R-wave to radial artery pulse delay, The Multi-Ethnic Study of Atherosclerosis. Sleep Med 2018; 48: 172–179.

55.

Phillips

Butlin

Wong

, et al. Is obstructive sleep apnoea causally related to arterial stiffness? A critical review of the experimental evidence. Sleep Med Rev 2013; 17: 7–18.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB