Arterial Stiffness in Overweight and Obesity: Association with Sex, Age, and Blood Pressure

The FATCOR study was performed from 2009 to 2017 in Bergen, Norway as a collaboration between a general practitioner center specialized in the management of obese patients and the Department of Heart Disease, Haukeland University Hospital. Women and men aged 30–65 years with a BMI >27.0 kg/m² were recruited after advertisement in a local newspaper and screened by a general practitioner [13]. Exclusion criteria were history of myocardial infarction, gastrointestinal disorders, severe psychiatric illness, or the inability to communicate in Norwegian language. A total number of 620 participants were recruited. For the present analysis, participants who withdrew consent (n = 2), or with incomplete recording of cf-PWV (n = 24), pulse wave analysis (n = 26), and 24-hours (24-h) systolic BP (n = 55) were excluded, leaving 323 women and 225 men eligible for analysis.

The FATCOR study was approved by the Regional Ethics Committee (approval number 17173) and complied with the Declaration of Helsinki. All participants signed written informed consent.

Self-reported information about general health, including use of medication, was collected in a standardized questionnaire. The information was quality assured against medical hospital records by a study nurse. BMI was calculated as body weight in kilograms/body height in meters squared (kg/m²). Obesity was defined as BMI ≥30 kg/m². Waist and hip circumferences were measured as recommended by the WHO using a nonflexible measuring tape [14]. Waist–hip ratio was calculated as waist circumference/hip circumference, and waist-height ratio as waist circumference/body height, all measured in centimeters. Tetrapolar bioelectrical impedance analysis (Tanita-TBF-300A, Tanita Corporation of America, Arlington Heights, USA) was used to analyze body composition. Blood glucose and lipids were measured in fasting, venous blood samples. Diabetes mellitus was considered present if history of diabetes mellitus, fasting blood glucose ≥7 mmol, 2-hours plasma glucose ≥11.1 mmol/l during oral glucose tolerance test (OGTT), or glycated hemoglobin A1C (HbA_1c) ≥6.5% [15]. Estimated glomerular filtration rate (eGFR) was calculated using the equation from the Chronic Kidney Disease Epidemiology Collaboration [16]. Smoking was defined as current smoking. High-sensitive C-reactive protein (hs-CRP) was measured in serum by Matrix-Assisted Laser Desorption/Ionization Time-Of-Flight mass spectrometry.

BP was measured by a study nurse at the general practitioner’s office. Attended BP and heart rate were measured three times with 1 min intervals after 5 min initial rest in the sitting position, using a regularly calibrated digital automatic Omron M4 sphygmomanometer (Omron Healthcare Co. Ltd., Hoofdorp, The Netherlands) and an appropriate cuff size on the non-dominant arm [17]. Clinical BP and heart rate were calculated as the average of the last 2 measurements. 24-h ambulatory BP was recorded using a Diasys Integra II apparatus (Novacor, Cedex, France). BP was recorded every 20 minutes during daytime and every 30 minutes during nighttime with an appropriately sized cuff on the non-dominant arm. Participants were instructed to avoid hard exercise and to relax their arm while the apparatus was measuring, but otherwise engage in normal activities [17]. The 24-h BP recording was repeated if <70% of measurements were valid. Hypertension was considered present if the participant reported use of antihypertensive drugs or had elevated 24-h BP (average 24-h systolic BP ≥130 mmHg or average 24-h diastolic BP ≥80 mmHg) [18].

Arterial stiffness was assessed by AP, AIx, and cf-PWV by using applanation tonometry (SphygmoCor, AtCor Medical, Sydney, West Ryde, Australia) under standardized laboratory conditions in accordance with guidelines [4]. Carotid pulse wave analysis was used to estimate AP and AIx. Pulse pressure waveforms were acquired transcutaneously from the right common carotid and femoral arteries with simultaneous recordings of the electrocardiograms to synchronize times for carotid and femoral pulse waves. Cf-PWV was calculated as the distance in meters between the two recording sites divided by the transit time between the recording sites in seconds [4].

Statistical analyses were performed using the IBM SPSS version 28 (IBM, Armonk, New York, USA). Continuous variables are presented as means ± standard deviations, and categorical variables are presented as numbers and percentages. hs-CRP was not normally distributed in the cohort, and therefore presented as median and interquartile range in the total cohort and in groups of women and men, and log transformed before included in statistical analyses. In group comparison, the Student’s t-test was used for continuous variables and the Chi-square test was used for categorical variables. Factors associated with AP, AIx, and cf-PWV were identified in univariable linear regression in the total study cohort and in sex-specific analyses, and significant covariates were included in multivariable models. The multivariable models on AP and AIx were adjusted for sex, age, BMI, 24-h systolic BP, heart rate, and smoking. Multivariable models on cf-PWV were additionally adjusted for diabetes mellitus. In secondary models, BMI was replaced by waist–hip ratio, fat mass percentage, waist circumference, waist-height ratio, and obesity as a categorical variable. Results were reported as standardized β-coefficients and p-values. In all analyses, a p-value<0.05 was considered statistically significant.

Clinical characteristics and data on arterial stiffness are presented in Table 1. In this middle-aged population, women had higher mean AP and AIx compared to men, while men had higher mean cf-PWV than women (all p < 0.001) (Table 1) (Fig. 1). Women also had higher BMI, fat mass percentage, hip circumference, waist-height ratio, and heart rate than men (all p < 0.05). Men had higher waist circumference, waist-hip ratio, 24-h systolic BP and 24-h diastolic BP values, and higher prevalence of hypertension (all p < 0.001). Prevalence of obesity, diabetes mellitus, smoking, and mean eGFR did not differ by sex (all p>0.05).

Results from univariable analyses are presented in Table 2. In multivariable analyses, women had higher AP (β = 0.41) and AIx (β = 0.48) compared to men, independent of age, BMI, 24-h systolic BP, heart rate, and smoking (all p < 0.05) (Table 2). Men had higher cf-PWV (β = 0.16) compared to women, independent of age, BMI, 24-h systolic BP, heart rate, diabetes mellitus, and smoking (all p < 0.001).

Results from univariable analysis are presented in Table 3. In multivariable analysis in women, higher AP was associated with higher age and 24-h systolic BP, and lower heart rate (all p < 0.001) (Table 3). Higher AIx was associated with higher age and 24-h systolic BP, and lower BMI and heart rate (all p < 0.05). In the same model, when BMI was replaced by obesity as a dichotomous variable, obesity was not significantly associated with AIx in women (data not shown). Higher cf-PWV was associated with higher age and 24-h systolic BP (all p < 0.001). Measures of adiposity, including BMI, obesity, waist-hip ratio, fat mass percentage, waist circumference, and waist-height ratio were not significantly associated with cf-PWV in women (all p > 0.05) (Table 4).

In multivariable analysis in men, higher AP was associated with higher age and 24-h systolic BP (both p < 0.001), while higher AIx was associated with higher age only (p < 0.001) (Table 3). Higher cf-PWV was associated with higher age, 24-h systolic BP, heart rate, and with non-smoking habit (all p < 0.05). Among measures of adiposity, higher cf-PWV was associated with higher waist–hip ratio in men (p < 0.05) (Table 4) (Fig. 2), while no significant associations with BMI, obesity, fat mass percentage, waist circumference, or waist height-ratio were found (Table 4).

Our finding that AP and AIx were higher in women, and cf-PWV was higher in men, are in line with previous large population-based studies in healthy, normotensive, non-obese subjects [2, 3, 19]. Prior to the current study, sex-specific information regarding arterial stiffness in obesity was scarce. In a small study of 133 morbidly obese patients, cf-PWV was higher in men than women, and increased in parallel with obesity severity in women, but not in men [20]. However, this study included a high number of patients with diabetes mellitus and coronary artery disease. Current European guidelines for assessment of arterial stiffness recommend use of cf-PWV and consider cf-PWV >10 m/s as a common threshold indicating increased arterial stiffness in both sexes [17]. Taken together, these results suggest that sex-specific thresholds for diagnosis of increased arterial stiffness may be considered when BMI is increased. This should be further tested in prospective outcome studies.

While cf-PWV is dependent on aortic wall stiffness and lumen diameter, AP and AIx are both measures of the peripheral wave reflection and express central pressure wave characteristics in the aorta [21, 22]. In specific, AIx is the proportion of central pulse pressure that is attributable to late systolic increase in pressure caused by overlap between the forward and reflected pressure wave [22]. It has been suggested that AIx is a better measure for arterial stiffness than cf-PWV in individuals under 50 years of age [1]. Still, few previous studies on arterial stiffness in obesity have included assessment of arterial stiffness by measures of peripheral wave reflection like AP and AIx. The Copenhagen City Heart Study found that age-group specific high AIx was significantly related to all-cause mortality in men but not in women, even after adjusting for CV risk factors such as BP [23]. In the present study, higher age and 24-h systolic BP were the most important factors associated with higher AP in both sexes. Higher levels of 24-h systolic BP were strongly associated with AIx only in women, suggesting important sex-differences in BP related pulsative vascular load. This is in line with previous findings that young women are more prone to develop hypertension and associated complications in obesity compared to their male counterparts [24]. Furthermore, sex differences in pulsative vascular load may partly explain why hypertension is a stronger risk factor for heart disease in women compared to men [25].

Interestingly, lower BMI was associated with higher AIx in women in the current study. However, when adjusting for BMI as a dichotomous variable, obesity was not associated with AIx in women, nor in men. In contrast, a study in normotensive non-obese healthy men and women found that BMI, waist circumference, and waist-hip ratio were inversely associated with aortic AIx in women [26], after adjusting for confounders.

Body fat distribution differ by sex [9]. Visceral adipose tissue is on both relative and absolute terms higher in men than in women, while the amount of subcutaneous adipose tissue is relatively higher in women [9]. Measures of obesity have been associated with arterial stiffness in both women and men in population-based studies [27, 28]. In a large Italian population-based study, an association between adiposity and cf-PWV was found in both sexes [27]. A Czech population-based study found that measures of abdominal adiposity such as waist circumference, waist-hip ratio, and waist-height ratio correlated better with cf-PWV than general adiposity measures such as BMI, even after adjusting for age, sex, mean arterial pressure, hypertension, diabetes mellitus, and dyslipidemia [29]. In a small Brazilian study in obese adults without diabetes mellitus, lower American Heart Association CV health life’s essential 8 score (based on health behaviours and metabolic risk factors) was associated with higher AP, but not with cf-PWV [30]. However, lower score was associated with high-for-age cf-PWV, reflecting early vascular aging. Among overweight and obese participants in the current study, waist-hip ratio was the only measure of adiposity that was significantly associated with cf-PWV, but only in men. Taken together, these results may suggest that in a population with increased BMI, the level of obesity has less impact on arterial stiffness.

Traditionally, office BP measurement has been used to identify hypertension, and most studies on arterial stiffness have reported office BP [1, 20]. The use of 24-h BP is advantageous in obesity because the prevalence of masked hypertension is greater than in the general population [31]. Furthermore, 24-h BP has a stronger association with CV organ damage compared to office BP [32]. Therefore, 24-h systolic BP was measured in the present study. As demonstrated, higher age and 24-h systolic BP were the main covariables of higher AP and cf-PWV both in women and men.

Higher heart rate was associated with higher cf-PWV in men. In contrast, among women, higher heart rate was associated with both lower AP and AIx. Earlier studies have demonstrated the same negative association between heart rate and AIx [33]. The lower AIx found with increasing heart rate may be explained by a reduction in ejection duration, causing a shift of the reflected pulse wave into diastole [33].