Plasma endothelin‐1 is not increased in overweight/obese hypertensive African women

Abstract

Introduction. Endothelin‐1 (ET‐1) has been found to be higher in hypertensive African Americans and obese hypertensive Caucasians compared to normotensive controls with an enhanced ET‐1‐dependent vasoconstrictor tone. ET‐1 levels and the associations thereof with cardiovascular function in overweight/obese normotensive and hypertensive African women have not been investigated. It is therefore hypothesized that ET‐1 levels are elevated in overweight/obese hypertensive African women compared to overweight/obese and lean normotensive controls. Additionally, it is hypothesized that these elevated ET‐1 levels are associated with increased total peripheral resistance (TPR) and decreased arterial compliance (C_W). Materials and methods. A case–case control study was performed which included 98 African women. The subjects were divided into lean normotensive (lean NT), overweight/obese normotensive (OW/OB NT) and overweight/obese hypertensive (OW/OB HT). The Finometer apparatus was used to obtain a more elaborate cardiovascular profile and plasma immunoreactive ET‐1 levels were determined. Results. ET‐1 levels were similar for the three groups. Although a decrease in vascular function was observed in the OW/OB HT group, no correlations were obtained between ET‐1 and the cardiovascular profile, before and after adjusting for age. Conclusion. In African women, ET‐1 levels did not differ between lean and overweight/obese and normotensive and hypertensive subjects. The lack of significant associations between ET‐1 and decreased vascular function in the overweight/obese hypertensive group suggests that ET‐1 is not implicated in obesity‐related hypertension in African women.

• African women
• cardiovascular function
• endothelin‐1

Introduction

The prevalence of obesity among African women far exceeds those in Caucasian women in South Africa and is approaching that of African American women [1]. However, the prevalence of hypertension in these African women is not the highest compared to the other population groups from South Africa [2]. Survey results have shown that African women have the highest average body mass index (BMI), but the second lowest prevalence of hypertension in comparison with the other population groups, i.e. Caucasians, Indians and people of mixed origin [3].

The high prevalence of hypertension among African Americans prompted Ergul et al. [4] to determine whether endothelin‐1 (ET‐1) possibly plays a role in the etiology of hypertension in African Americans. It was found that ET‐1 levels are significantly elevated in hypertensive African American men and women compared to normotensive controls, implicating ET‐1 in the development or maintenance of hypertension in this population group. Recently, Campia et al. [5] also found that African Americans have an enhanced ET‐1‐dependent vasoconstrictor tone compared to Caucasians. Additionally, Parrinello et al. [6] added obesity in their study on Caucasians and determined that ET‐1 levels are significantly elevated in obese hypertensives compared to obese and lean normotensives. Moreover, ET‐1 levels were also higher in obese normotensives compared to lean normotensives [6] and Cardillo et al. [7] determined that obesity is also associated with an enhanced ET‐1‐dependent vasoconstrictor activity. Thus, in summary, ET‐1 levels are elevated in hypertensive obese Caucasians compared to normotensive obese controls [6] with an enhanced ET‐1‐dependent vasoconstrictor tone [7], and elevated in hypertensive African Americans compared to normotensive controls [4], with an enhanced ET‐1‐dependent vasoconstrictor tone [5].

Apart from the vasoconstrictive effects, ET‐1 stimulates vascular smooth muscle cell proliferation and hypertrophy, leading to decreased arterial distensibility and compliance [8]. ET‐1 levels and the influence thereof on cardiovascular function have not been determined in African women. It would thus seem relevant to determine whether circulating ET‐1 levels are also different in African women with different degrees of adiposity and blood pressure and to determine whether circulating levels are associated with cardiovascular function in these women.

It is therefore hypothesized that ET‐1 levels are elevated in overweight/obese hypertensive African women compared to overweight/obese and lean normotensive controls. Additionally, it is hypothesized that these elevated ET‐1 levels are associated with increased total peripheral resistance (TPR) and decreased arterial compliance (C_W).

Materials and methods

Participants

A case–case control survey was performed which included a sample of 102 urban African women volunteers working at a government institution in the Potchefstroom district of the North West Province, South Africa. A dietician, employed at the institution, recruited the subjects according to the initial study design. They had to be apparently healthy African women aged between 20 and 50 years. The dietician attempted to recruit only HIV‐negative subjects (according to their status as determined 3 months prior), but the negative status of all subjects cannot be guaranteed. Subjects were initially recruited based on their BMI as measured at the institution's medical station in lean (BMI: 18.5–24.9 kg/m²), overweight (BMI: 25–29.9 kg/m²) and obese (BMI: ⩾30 kg/m²). Exclusion criteria were pregnancy, lactation, diabetes mellitus and treatment for mental or neurological diseases. The mean age of this group was 31.3±8.6 years. Out of this total sample of 102 subjects, three groups were selected: (i) a group of 35 lean normotensive subjects; (ii) secondly, a group of 46 normotensive subjects, which consisted of a combination of overweight (n = 19) and obese (n = 27) subjects; (iii) a group of 17 newly diagnosed hypertensive subjects, also consisting of a combination of overweight (n = 6) and obese (n = 11) subjects. The subjects were regarded as hypertensive if their blood pressure was ⩾140 and/or ⩾90 mmHg [9]. A total of 98 subjects were included in the analysis of this article. The four subjects that were excluded were the only lean hypertensive subjects in the total subject group, rendering the group too small for statistical use.

All research subjects gave informed consent in writing. The Ethics Committee of the North‐West University approved the study.

Experimental procedure

During the course of the study, the subjects reported at a metabolic ward facility (consisting of 10 single bedrooms, a living room and kitchen) at 18.00 h. They were all introduced to the experimental set‐up, including the Finometer apparatus. The purpose of the introduction to the experimental set‐up was to minimize anticipation stress and the white coat effect. The subjects received a light meal at about 19.00 h, which excluded caffeine and alcohol, and went to sleep before 23.00 h. The purpose of the dietary restrictions was to enable a good night's sleep and a stable, resting blood pressure the next morning. Finometer recordings were obtained between 06.00 and 08.00 h the next morning, after the overnight rest and before breakfast. Subjects were not permitted to walk around or have anything to eat or drink (except water) until all recordings were completed.

The subject was awake and lying in the Fowler's position in a quiet single bedroom while the Finometer device was connected to the subject. The cuffs were of appropriate size. Blood pressure was recorded continuously for a period of at least 7 min. After a recording of at least 2 min, the Finometer performed a return‐to‐flow systolic calibration. This is an individual patient level adjustment, which calibrates the upper arm pressure of each specific subject with the finger pressure. Highest precision in blood pressure readings is obtained only after this calibration. From the 7‐min continuous blood pressure recording of the Finometer, the average systolic and diastolic blood pressures were determined from the last 2 min of the recording.

The Finometer device computed all cardiovascular variables online and stored the data in result files on a hard disk. The systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP), heart rate (HR), stroke volume (SV), cardiac output (CO), TPR, and C_W[10], [11] were computed and stored.

The vascular unloading technique of Peñáz together with the Physiocal criteria of Wesseling provided reliable, non‐invasive and continuous estimates of blood pressure, which are usable especially in comparative studies [12], [13]. Since the pressure waveform is available continuously, computations provide further information on the dynamics of the cardiovascular system, similar to intra‐arterial measurements [11], [14–18].

Body mass and height measurements were also taken. These anthropometric measurements were performed according to standard methods as described by Norton & Olds [19]. Maximum height was measured to the nearest 0.1 cm, with the head in the Frankfort plane, by means of a stadiometer (Invicta, IP 1465, UK). Body mass was measured to the nearest 0.1 kg by means of a calibrated electronic scale (Precision Health Scale, A&D Company, Japan). During these measurements, the subjects had to stand erect with the feet together and without volitionally contracting the gluteal muscles.

After the Finometer recordings were taken, fasting blood samples were drawn from the vena cephalica or medial cubital vein and plasma was prepared according to standard methods. ET‐1 was determined by means of a ¹²⁵I RIA kit (AEC Amersham (PTY) LTD, Cat No. RPA 545).

Statistical analysis

The computer software package Statistica v/6.0 was used for the analysis of the data. The analysis of covariance (ANCOVA) was used to show significant differences between groups while adjusting for age. The results were seen as statistically significant when the p‐value was ≤0.05, which means that each of the three comparisons were tested on a 0.05/3 = 0.017 level of significance. That is the same as to calculate 98.3% confidence intervals for each of the groups and to determine whether these intervals overlap [20]. Partial correlation coefficients were used to show associations between various variables. A power analysis (80% power) determined that the sample size would be adequate.

Results

Means, standard errors and confidence intervals of cardiovascular and anthropometric variables as well as ET‐1 levels are presented in Table I. Results were adjusted for age due to differences obtained (p≤0.07) between the three groups.

Table I. Means±standard error of cardiovascular and anthropometric variables as well as ET‐1 levels adjusted for age at a 98.3% confidence interval.

	Lean NT (n = 35)	OW/OB NT (n = 46)	OW/OB HT (n = 17)
Age (years)	28.3±1.4 [25.7–30.9]	31.3±1.2 [29.1–33.6]	37.4±2.0 [33.7–41.1]
SBP (mmHg)	123±2.3 [118–129]^a	124±1.9 [120–129]^b	156±1.9 [148–164]^ab
DBP (mmHg)	72±1.4 [69–76]^c	77±1.2 [74–79]^d	91±2.1 [86–96]^cd
HR (beats/min)	64.8±1.6 [60.8–68.7]	70.5±1.4 [67.1–73.8]	69.1±2.4 [63.3–74.9]
SV (ml)	80.0±2.4 [74.2–85.9]	87.7±2.1 [82.7–92.7]	89.1±3.5 [80.5–97.7]
CO (l/min)	5.13±0.2 [4.67–5.59]^ef	6.10±0.2 [5.71–6.49]^e	6.09±0.3 [5.42–6.77]^f
TPR (mmHg/ml/s)	1.16±0.04 [1.06–1.25]	0.98±0.04 [0.90–1.06]	1.23±0.06 [1.09–1.37]
CW (ml/mmHg)	1.78±0.04 [1.68–1.89]^g	2.00±0.04 [1.91–2.09]^gh	1.65±0.06 [1.49–1.80]^h
PP (mmHg)	51.2±1.8 [46.9–55.5]^i	47.8±1.5 [44.1–51.4]^j	64.7±2.6 [58.4–71.0]^ij
ET‐1 (fmol/ml)	47.4±7.0 [30.4–64.5]	47.4±6.0 [32.9–62.0]	47.9±10.3 [22.9–73.0]
BMI (kg/m^2)	22.3±0.7 [20.5–24.0]^kl	31.3±0.6 [29.8–32.8]^k	32.4±1.1 [29.9–35.0]^l
Waist circum. (cm)	96.4±1.6 [92.6–100]^mn	112±1.3 [109–116]^m	114±2.3 [109–120]^n

Lean NT, lean normotensive; OW/OB NT, overweight/obese normotensive; OW/OB HT, overweight/obese hypertensive; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; SV, stroke volume; CO, cardiac output; TPR, total peripheral resistance; C_W, Windkessel arterial compliance; PP, pulse pressure; ET‐1, endothelin‐1; BMI, body mass index.

Means with the same superscript: statistically significant (p≤0.05).

ET‐1 levels were similar for the lean normotensive (lean NT), overweight/obese normotensive (OW/OB NT) and overweight/obese hypertensive (OW/OB HT) groups. Since this result seems surprising, the total group was re‐divided into normotensive (NT) (n = 81) and hypertensive (HT) (n = 17), irrespective of obesity (Table II). By doing so, again no significant difference could be obtained.

Table II. Means±standard errors for endothelin‐1 (ET‐1) adjusted for age at a 95% confidence interval.

	Normotensive (n = 81)	Hypertensive (n = 17)
ET‐1 (fmol/ml)	47.4±4.5 [38.4–56.4]	47.9±10.2 [27.6–68.3]
	Lean (n = 35)	Overweight/obese (n = 63)
ET‐1 (fmol/ml)	47.4±7.0 [33.6–61.3]	47.6±5.2 [37.3–57.8]

Similarly, no significant difference was obtained when the total group was again re‐divided into lean (n = 35) and overweight/obese (OW/OB) (n = 63), irrespective of blood pressure (Table II).

Apart from the above‐mentioned lack of significant differences for ET‐1 levels, C_W was significantly lower (p≤0.05) and TPR higher (p≤0.08) in the OW/OB HT group compared to the OW/OB NT group. PP was significantly higher in the OW/OB HT group compared to the lean and OW/OB NT groups.

No correlations were obtained between ET‐1 and the cardiovascular profile of the total group, lean NT, OW/OB NT and OW/OB HT groups before (Table III) and after (Table IV) adjusting for age.

Table III. Partial correlation coefficients between endothelin‐1 (ET‐1) and cardiovascular variables.

	Total froup	Lean NT	OW/OB NT	OW/OB HT
	ET‐1	ET‐1	ET‐1	ET‐1
SBP	r = 0.02; p = 0.87	r = −0.23; p = 0.17	r = −0.20; p = 0.18	r = 0.31; p = 0.22
DBP	r = −0.08; p = 0.42	r = −0.19; p = 0.28	r = −0.18; p = 0.23	r = 0.09; p = 0.74
PP	r = 0.09; p = 0.36	r = −0.07; p = 0.67	r = −0.06; p = 0.70	r = 0.37; p = 0.15
TPR	r = −0.03; p = 0.75	r = −0.09; p = 0.62	r = −0.22; p = 0.14	r = 0.24; p = 0.34
CW	r = −0.01; p = 0.94	r = 0.03; p = 0.88	r = 0.10; p = 0.50	r = −0.20; p = 0.44

Lean NT, lean normotensive; OW/OB NT, overweight/obese normotensive; OW/OB HT, overweight/obese hypertensive; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; TPR, total peripheral resistance; C_W, Windkessel arterial compliance.

Table IV. Partial correlation coefficients between endothelin‐1 (ET‐1) and cardiovascular variables adjusted for age.

	Total group	Lean NT	OW/OB NT	OW/OB HT
	ET‐1	ET‐1	ET‐1	ET‐1
SBP	r = −0.13; p = 0.20	r = −0.18; p = 0.31	r = −0.24; p = 0.12	r = 0.28; p = 0.33
DBP	r = −0.15; p = 0.14	r = −0.19; p = 0.28	r = −0.22; p = 0.15	r = −0.08; p = 0.79
PP	r = −0.03; p = 0.75	r = −0.01; p = 0.95	r = −0.06; p = 0.68	r = 0.33; p = 0.24
TPR	r = −0.10; p = 0.36	r = −0.08; p = 0.66	r = −0.22; p = 0.15	r = 0.39; p = 0.17;
CW	r = 0.06; p = 0.57	r = −0.02; p = 0.90	r = 0.19; p = 0.21	r = −0.31; p = 0.28

Lean NT, lean normotensive; OW/OB NT, overweight/obese normotensive; OW/OB HT, overweight/obese hypertensive; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; TPR, total peripheral resistance; C_W, Windkessel arterial compliance.

Discussion

The results of this study show no differences for ET‐1 levels between the lean NT, OW/OB NT and OW/OB HT groups. This was unexpected, since Ergul et al. [4] demonstrated that ET‐1 levels are elevated in hypertensive African American women compared to normotensive controls and Parinello et al. [6] found that ET‐1 levels are also elevated in obese hypertensive Caucasians compared to obese normotensive controls. Additionally, ET‐1 levels were also higher in obese normotensives compared to lean normotensives [6], indicating that ET‐1 levels are elevated possibly as a consequence of human obesity and hypertension. One would thus expect ET‐1 levels to be invariably elevated in OW/OB HT hypertensive African women.

Since the results of this study seemed surprising, the total group was re‐divided into normotensive and hypertensive, irrespective of obesity and into lean and overweight/obese, irrespective of blood pressure. By doing so, in both cases, still no significant differences could be obtained, suggesting that circulating ET‐1 levels are not influenced by either increased adiposity or blood pressure in African women.

A decrease in vascular function in the OW/OB HT group was observed with TPR and PP being higher and C_W significantly lower in the OW/OB HT group compared to the OW/OB NT group. Increased TPR reflects vasoconstriction [21] whereas decreased C_W and increased PP reflects stiffened arteries [22], which have both, as mentioned, been found in previous studies to be influenced by circulating ET‐1 levels in Caucasian populations [23], [24]. Furthermore, the enhanced ET‐1‐dependent vasoconstrictor tone observed in obese hypertensive Caucasians [7] and hypertensive African Americans [5] would suggest that ET‐1 would be associated with the overweight/obese hypertensive African women's increased TPR and/or decreased C_W in this study. Yet no correlations were obtained between ET‐1 and these variables. Thus, ET‐1, as opposed to hypertensive African American women and obese hypertensive Caucasians, does not seem to play a significant role in obesity‐related hypertension in African women.

In conclusion, in African women, ET‐1 levels did not differ between lean and overweight/obese, and normotensive and hypertensive subjects. Along with the lack of significant associations between ET‐1 and decreased vascular function in the overweight/obese hypertensive group, it is suggested that ET‐1 is not implicated in obesity‐related hypertension in African women.

Limitations of the study were that the HIV‐negative status of the subjects could not be guaranteed as well as the inability to differentiate between the types of hypertension due to a limited number of hypertensive subjects.

Acknowledgements

The authors are grateful to Prof. HS Steyn of the Statistical Consultation Service at the North‐West University for statistical assistance. The authors are also grateful to those funding this project, namely the South African National Research Foundation (NRF GUN number 2054068), the Medical Research Council and the Research Focus Area 9.1 of the North‐West University (Potchefstroom Campus).