Predictors and Reference Values of Pulse Wave Velocity in Prepubertal Angolan Children

Abstract

Carotid‐femoral pulse wave velocity (PWV) has been used as the gold standard method to estimate arterial stiffness. However, its use in clinical practice is still limited because reference values for specific groups, such as black children, remain unknown. The authors aimed to investigate predictors and to propose preliminary reference values of PWV in this population. Prepubertal schoolchildren (N=157; mean age, 9.36±1.41 year) from Luanda (Angola) with normal blood pressure values and without obesity were included. Mean PWV was 5.73±0.68 m/s, with no difference between the sexes. Univariate regression analysis showed a significant (P<.05) positive correlation between PWV and height, age, body weight, lean body weight, and blood pressure. In multivariate analysis, however, only height remained an independent predictor of PWV [PWV=0.018×height (cm)+3.230]. Curves of PWV percentiles as a function of height are proposed, thus identifying normal PWV in black children.

The role of arterial elasticity for the physiological functioning of the cardiovascular system has been investigated for a long time. The decrease in arterial elasticity (stiffness increase) impairs the role of large arteries in the maintenance of cardiovascular homeostasis and contributes to the development of highly prevalent cardiovascular diseases in adults, such as arterial hypertension1 In 1922, Bramwell and colleagues2 reported that the pulse wave travels at 4 m/s to 10 m/s in human arteries, depending mainly on the elastic condition of the arterial wall. Arterial stiffness, the inverse of elasticity, depends on a variety of factors. However, age is a key determinant of its value, leading to the concept that arterial stiffness can be considered a proxy of arterial age and vascular health. Stiffer arteries are associated with higher blood pressure (BP) levels and thus has been increasingly used as an independent predictor of adverse cardiovascular events in both hypertensive and normotensive individuals.2, 3

The carotid‐femoral pulse wave velocity (PWV) has been largely used to estimate the stiffness of the large arteries (mainly aorta), and robust epidemiological studies have indicated age, sex, BP, and uric acid as its main determinants4, 5 Some studies have also indicated body fat as an independent predictor of PWV6 an association not found in other studies7, 8 Beyond these factors, ethnicity also seems to influence arterial stiffness. Accordingly, higher PWV values have been observed in different populations of African ancestry9, 10, 11 even after adjusting for several covariates, including BP.9 In adults and children with uremia and end‐stage kidney diseases, changes in calcium and phosphate homeostasis also increase PWV12, 13, 14

Despite the increasing importance of arterial stiffness in the evaluation of cardiovascular health, the clinical use of this parameter in children is still limited because of the lack of reference values. Most of the studies that aimed to obtain such values were performed in Caucasian children15 In a study to establish PWV reference values in children and adolescents, age, height, and mean BP were identified as independent predictors of this variable15 Reference values of PWV in African adults were obtained in a single study5 and no values for African children were determined until now. Given that adults with African ancestry show higher PWV values9, 10 and considering the increasing importance of chronic cardiovascular diseases to the epidemiological context of African countries experiencing rapid epidemiological transition, we decided to determine the factors associated with PWV values in a cohort of prepubertal Angolan school children. Considering the interplay between birth weight and BP in adults16, 17, 18 we also included this variable in the present analysis. Finally, we determined preliminary reference values of this variable in normotensive and nonobese children included in the present cohort.

Methods

Sample Selection

A cross‐sectional, observational, descriptive study was conducted in a sample of children enrolled in a public school of the first cycle of primary education in the urban area of Luanda, the Angolan capital. All children were black. A total of 214 children aged between 7 and 12 years were evaluated from June 2012 to November 2013. We removed from this analysis the children not classified as having Tanner stage I criteria19, 20 and/or who had completed 12 years between recruitment and examination (n=16). An additional 41 were removed because of high BP levels during examination (systolic or diastolic pressure ≥90 percentile for sex, age, and height) or obesity. Therefore, the present analyses comprise data collected from 157 apparently healthy prepubertal schoolchildren. The project was approved by the Independent Committee of Ethics and Research of the Faculty of Medicine at Agostinho Neto University following the guidelines for research involving human beings in accordance with the Declaration of Helsinki. Parents and/or guardians were asked to read and sign a consent form before data collection.

Examinations

All data were collected during a single visit in the morning at the Department of Physiology, Faculty of Medicine, Agostinho Neto University on a prescheduled day. A fasting blood sample was collected by venipuncture and processed in the same day in the Department of Biochemistry to determine glucose, urea, uric acid, creatinine, total cholesterol, high‐density lipoprotein cholesterol, and triglyceride values. All dosages were performed with BioSystems reagents (Barcelona, Spain). Sociodemographic data, physical activity, and dietary habits were obtained through interview with the child in the presence of the mother or guardian.

Anthropometric measurements were performed by trained technicians. Body weight was obtained using a digital electronic scale (SECA, Mod 763, Hamburg, Germany) with 0.1 kg precision while fasting and after voiding with the children barefoot in only underclothes. Height was measured using a fixed stadiometer with 0.5 cm precision. The child was placed on the central part of the scale platform with heels together, head and buttocks resting against the stadiometer, and eyes looking toward the Frankfurt horizontal plane. Body mass index (BMI) was calculated by dividing body weight by height squared (kg/m²). The thickness of four skinfolds (triceps, suprailiac, subscapular, and abdominal) was measured with a manual caliper with a precision of 1 mm. Body composition was measured with a tetrapolar bioimpedance analyzer (Maltron BioScan, Model 916; Maltron International Ltd, Essex, England), to obtain lean mass, percentage of body fat, and fat mass. The later was used as an index of overall obesity. Waist and hip were measured with an inextensible plastic tape with 0.5 cm precision following standard techniques21

Hemodynamic Parameters

BP was measured with an automatic sphygmomanometer (OMRON, model HEM‐742 IntelliSense, Nanjing, China). The cuff was selected according to the size of the arm circumference following manufacturer recommendations. In each child, three consecutive BP measurements were obtained in the left arm in the sitting position at 2‐minute intervals after a rest period of 5 to 10 minutes. The forearm was supported on a flat surface at nearly 120° with the arm. Systolic BP (SBP), diastolic BP (DBP), and resting heart rate were calculated as the arithmetic mean of the last two readings. Pulse pressure (PP) was calculated as the difference between the SBP and DBP, and mean arterial pressure (MAP) was calculated by the formula [SBP+(2×DBP)]/3. BP was considered normal if either the SBP or DBP was below the 90th percentile according to sex, age, and height. If SBP or DBP was above the 95th percentile, the child was included in the category of having “high BP.” If SBP or DBP was intermediary between the extremes (>P90<P95), the child was classified as having “borderline BP.”22

Pulse Wave Velocity

Carotid‐femoral PWV was measured according to the European Expert Consensus on Arterial Stiffness. The distance between the suprasternal notch to the right femoral pulse in the inguinal region was measured with an inelastic tape without correction for the abdomen arc. PWV was estimated as a function of the distance between the carotid to femoral pulses and the delay between the carotid and the femoral pulse waves. Measurements were performed in a quiet room with stable temperature in the supine position after 5 to 10 minutes of rest using a noninvasive automatic device (Complior SP, Artech Medical, Pantin, France). All examinations were performed by a single experienced observer blinded to other clinical characteristics of the children, except for BP, which was measured just before PWV. Intraobserver reproducibility of PWV values (r>0.9) was based on 30 examines performed in children with similar characteristics. Technical features of this device were described in detail in a validation study.23 Two measurements were obtained in each child and when the difference between the values was higher than 0.5 m/s, a third measurement was performed. The arithmetic mean of the two lowest values was used.

Reference Values of PWV

To generate reference values of the carotid‐femoral PWV, we excluded from the present analysis all children with high systolic or diastolic BP (higher than the 90th percentile predicted for sex, age, and height) or those with obesity (BMI ≥95th percentile). After exclusions, 157 children remained in our databank (around 80% of the initial prepubertal group).

Statistical Analysis

Data analysis was performed using SPSS software, version 20.0 (SPSS Inc, Chicago, IL). The normality of the data was examined using the Kolmogorov‐Smirnov test. Continuous variables are presented as mean±standard deviations (SDs) and categorical variables as proportions. Comparisons of two means were performed by Student t test for independent samples and proportions were compared by chi‐square test. Univariate linear regression was performed to evaluate the correlation between PWV and other variables. To determine independent predictors of PWV, a multiple linear regression analysis was performed including in the model all variables with significant (P<.05) partial Pearson correlation coefficients (r). Quantile regression (R version 3.2.2, quantreg package) was used to generate estimates of the percentiles (from 10th to 90th) of PWV by age or height. The growth charts were constructed using a specific package for R software (quantregGrowth, version 0.3‐1). Statistical significance was set at P<.05.

Results

Characteristic of Sample

The general characteristics of the sample (N=157; boys=39%; mean age=9.36±1.41 years) before and after sex stratification are shown in Table 1. Birth weight, waist to hip ratio, and PP were higher in boys, while heart rate was higher in girls. All other parameters were unrelated to sex, including PWV, BP, and biochemical parameters (see Table S1 for further details).

Table 1.

Anthropometric and Clinical Characteristics of Prepubertal Children

Parameters	Boys (n=61)	Girls (n=96)	All	P Value
No.	61	96	157
Birth weight, kg	3.29±0.58	3.01±0.60	3.17±0.60	.046
Age, y	9.46±0.99	9.29±1.07	9.36±1.41	.326
Height, cm	136.9±8.1	138.0±9.5	137.6±8.9	.440
Weight, kg	30.9±6.3	31.2±6.9	31.1±6.7	.767
BMI, kg/m2	16.4±1.9	16.3±2.1	16.3±2.1	.676
Lean mass, kg	25.7±4.0	25.3±4.9	25.5±4.6	.632
Waist/hip ratio	0.83±0.04	0.81±0.04	0.81±0.04	.009
Glycemia, mg/dL	87.7±16.1	86.9±13.5	87.3±14.5	.754
Triglycerides, mg/dL	64.6±30.2	63.1±29.2	63.7±29.5	.751
LDL‐C, mg/dL	103.0±33.1	103.9±34.8	103.6±34.0	.868
HDL‐C, mg/dL	56.8±13.3	59.9±11.4	58.7±12.3	.130
SBP, mm Hg	102±7	102±7	102±7	.640
DBP, mm Hg	61±6	62±5	62±6	.095
MBP, mm Hg	74±6	75±5	75±5	.348
PP, mm Hg	42±6	39±6	40±6	.036
HR, beats per min	80±10	84±10	83±10	.011
PWV, m/s	5.76±0.74	5.72±0.64	5.73±0.68	.760

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL‐C, high‐density lipoprotein cholesterol; HR, heart rate; LDL‐C, low‐density lipoprotein cholesterol; MBP, mean arterial pressure; PP, pulse pressure; PWV, pulse wave velocity; SBP, systolic blood pressure. Data are presented as mean±standard deviation.

To verify the possible influence of birth weight in hemodynamic and PWV values, the children were classified according to birth weight (Table 2). Except for birth weight itself, all other parameters were similar. However, a tendency (P<.10) to higher values of PP was observed in the lowest birth weight group (<2500 g), which also showed a tendency (P<.10) to lower body weight and DBP.

Table 2.

Values of Anthropometric and Hemodynamic Parameters According to Birth Weight

Parameters	Birth Weight, kg
	<2.500	2.500–2.999	3.000–3.500	>3.500	P Value
No.	15	36	68	38
Age, y	9.5±1.1	9.4±0.9	9.4±1.1	9.2±1.0	.833
Birth weight, kg	2.05±0.46a	2.78±0.11a	3.24±0.17a	3.87±0.43a	<.001
Weight, kg	28.3±3.8	29.6±5.4	31.7±7.0	32.5±7.6	.082
Height, cm	135.9±7.1	136.0±8.2	138.5±9.1	138.1±9.9	.690
BMI, kg/m2	15.4±1.29	15.9±1.98	16.4±1.93	16.9±2.43	.249
SBP, mm Hg	102±8	102±7	102±7	102±7	.991
DBP, mm Hg	59±7	62±6	62±4	62±6	.059
MBP, mm Hg	73±6	76±6	75±4	75±6	.293
PP, mm Hg	43±8	39±4	40±6	40±6	.095
HR, beats per min	78±10	83±10	83±9	83±10	.153
PWV, m/s	5.84±0.59	5.63±0.55	5.78±0.78	5.72±0.63	.692

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HR, heart rate; MBP, mean blood pressure; PP, pulse pressure; PWV, pulse wave velocity; SBP, systolic blood pressure. Data are presented as mean±standard deviation. ^aDifference between groups.

Anthropometric and hemodynamic characteristics increased along the age quartiles, as shown in Table 3. Heart rate, as expected, decreased in the same direction. Birth weight was similar in all quartiles. Interestingly, PWV remained stable along the ages while BP increased along age quartiles.

Table 3.

Anthropometric and Hemodynamic Characteristics of the Study Subgroup

Age (Quartiles)
	First Quartile (<8 y)	Second Quartile (8.5 to 9 y)	Third Quartile (9.5 to 10 y)	Fourth Quartile (10.5 to 11 y)	P for Trend
Age, y	7.9±0.29a	9.0±0.0a	10.0±0.0a	11.0±0.0a	<.0001
Weight, kg	26.9±4.57	28.9±3.64	32.8±5.88c	38.5±8.50b	<.0001
Height, cm	129.9±8.12a	135.2±5.34a	140.8±6.53a	147.7±8.09a	<.0001
BMI, kg/m2	16.0±1.65	15.8±1.61	16.5±1.96	17.5±2.92c	.002
Lean mass, kg	22.3±3.27	23.9±2.69	26.9±4.10c	30.6±4.87b	<.0001
SBP, mm Hg	99.8±6.71	100.5±6.43	103.1±5.77	105.1±7.64c	.004
DBP, mm Hg	60.2±6.28	61.2±5.49	61.6±4.27	63.9±5.91	.08
MBP, mm Hg	73.4±5.79	74.3±5.25	75.5±4.08	77.6±5.63c	.012
HR, beats per min	85.8±9.1	83.5±9.2	80.0±9.5	80.1±10.5	.033
PP, mm Hg	39.5±5.91	39.2±5.33	41.5±5.46	41.3±7.06	.190
PWV, m/s	5.59±0.91	5.71±0.61	5.76±0.58	5.92±0.58	.306

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HR, heart rate; MBP, mean blood pressure; PP, pulse pressure; PWV, pulse wave velocity; SBP, systolic blood pressure. Data are presented as mean±standard deviation. ^aDifference between all groups. ^bDifference in the three groups. ^cDifference between the first and second quartiles.

Reference PWV Values

Univariate analysis of PWV with anthropometric, hemodynamic, and biochemical variables is shown in Table 5. PWV was positively and significantly associated with age (r=0.158; P=.024), weight (r=0.173; P=.015), height (r=0.240; P=.001), lean body mass (r=0.177; P=.013), SBP (r=0.153; P=.028), DBP (r=0.135; P=.046), and MAP (r=159; P=.023). After including all these variables in PWV modeling in a multivariate analysis, the only independent predictor of PWV was height (β=0.018; standard error β=0.006; P=.002; R ²=0.058). Therefore, according to our data, the best estimate of PWV of prepubertal black African children can be given by the linear relationship:

$$\text{PWV}=0.018\times \text{height}(\text{cm})+3.230$$

Table 5.

Correlation Matrix Between PWV and Anthropometric, Biochemical, and Hemodynamic Parameters in Prepubescent Healthy Children

	PWV	Age	Birth Weight	Weight	Height	BMI	Lean Mass	SBP	DBP	MBP	PP	Uric Acid	Total Cholesterol
PWV	–
Age	0.158a	–
Birth weight	–0.011	–0.053	–
Weight	0.173a	0.567b	0.143a	–
Height	0.240c	0.654b	0.083	0.803b	–
BMI	0.056	0.254c	0.132a	0.803b	0.309b	–
Lean mass	0.177a	0.603b	0.132a	0.961b	0.838b	0.721b	–
SBP	0.153a	0.276c	–0.016	0.364b	0.270b	0.332b	0.360b	–
DBP	0.135a	0.187a	0.097a	0.367b	0.322b	0.253c	0.310b	0.565b	–
MBP	0.159a	0.247c	0.061	0.411b	0.339b	0.318b	0.369b	0.819b	0.936b	–
PP	0.049	0.143a	–0.11a	0.075	0.008	0.146a	0.124	0.624b	–0.292b	0.063	–
Uric acid	0.106	0.101	0.089	0.047	0.055	0.010	0.031	0.153a	0.106	0.139a	0.077	–
Total cholesterol	0.061	0.031	0.005	–0.074	0.012	–0.122	–0.102	0.032	0.066	0.060	–0.025	0.143a	–

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; MAP, mean blood pressure; PP, pulse pressure; PWV, pulse wave velocity; SBP, systolic blood pressure.

^a P<.05. ^b P<.001. ^c P<.01.

Because PWV was unrelated to sex in prepubertal children, the percentile distribution of PWV values was determined according to age and height, as shown in Table 4. Distribution of percentiles by age and height is shown in Figure 1. Data show that the interquartile coefficients of PWV decrease with increasing age and height. Figure 2 shows the correlation between PWV and height with respective 95% confidence intervals (CIs), indicating that the points above the upper limit show high PWV values.

Table 4.

Values of Percentiles of PWV Based on Age and Height

Age, y	Patients, No.	Mean±SD	Percentile
			5th	10th	25th	50th	75th	90th	95th
<8	34	5.6±0.91	3.8	4.3	5.1	5.6	6.4	6.6	7.2
8.0–9.0	56	5.7±0.61	4.6	4.9	5.4	5.6	6.1	6.6	6.8
9.1–10.0	41	5.8±0.59	5.0	5.0	5.3	5.6	6.1	6.8	6.9
10.1–11.0	26	5.9±0.63	5.0	5.2	5.5	5.8	6.2	6.9	7.3
Height, cm
<126	14	5.1±0.99	3.0	3.6	4.4	5.3	5.7	6.5
126–135	55	5.8±0.66	4.9	5.0	5.3	5.6	6.3	6.7	6.9
136–145	55	5.7±0.54	4.9	5.1	5.3	5.6	6.0	6.4	6.8
>145	33	5.9±0.63	4.8	5.2	5.5	5.8	6.3	6.9	7.2
All	157	5.74±0.68	4.6	5.0	5.3	5.6	5.2	6.6	6.9

Abbreviations: PWV, pulse wave velocity; SD, standard deviation.

Figure 1.

Pulse wave velocity percentile curves according to age (A) and height (B).

Figure 2.

Correlation between pulse wave velocity and height (y=0.0182x+3.2298; r=0.24, P=.001).

Discussion

We observed that in prepubertal and healthy black schoolchildren in Angola, height is the best predictor to generate reference PWV values. Moreover, we showed that in this group, birth weight does not show any significant influence on BP and PWV.

The carotid‐femoral PWV is the most tested noninvasive method to measure arterial stiffness because of its simplicity, accuracy, and reproducibility23, 24 Artery stiffness of the large arteries is mostly dependent on the content and organization of elastic fibers of the vascular wall25 A stiffer arterial tree leads to increased cardiac afterload and energy expenditure in the systolic work25 Despite being an independent predictor of cardiovascular morbidity and mortality, the use of PWV in clinical settings depends on the existence of reference values for different groups of patients. Studies to determine reference values of PWV and the identification of its main determinants have mostly been performed in adults5, 10, 26, 27, 28 Some studies on this subject have also conducted in children and adolescents15, 29, 30 However, most investigations studied Caucasians or black populations living out of Africa. To the best of our knowledge, this is the first study aimed to define reference values of PWV in prepubertal and healthy children in Africa. Many countries in this continent, such as Angola, are now experiencing a rapid epidemiological transition with a great increase in morbidity and mortality caused by chronic diseases. In this context, identification of particular characteristics of cardiovascular risk factors in Africans is important to support local health policies. Studies in this area should not only encompass adults but also children because prevention of chronic diseases in adulthood must begin in childhood.

In prepubertal children not exhibiting high BP or obesity, height was the only independent predictor of PWV according to our data.

Race is a well‐defined variable that influences PWV in adults, with many studies showing higher PWV values in blacks compared with whites9, 11, 31, 32 The overall mean PWV found in our study (5.74±0.68; 95% CI, 4.93–6.55) was higher than that found in previous studies conducted in Caucasian children, such as those of Reusz and colleagues (95% CI, 2.809–5.902), Fischer and colleagues (95% CI, 3.4–5.4), and Hidvégi and colleagues (5.5±0.65 m/s), although different methods have been used to measure PWV. Comparisons among these studies, however, should be done cautiously because different methods to measure PWV were used. In addition, contrary to expectations, the Caucasian children had higher BP values, ruling out the possibility that the higher PWV velocity found in African children was secondary to this factor. Studies in adults have shown that blacks had higher PWV values than whites, even after correcting for BP and age9, 11 The elastic properties of the great vessels play an important role in determining the propagation of the PWV influenced by the arterial wall components, such as collagen, elastin, and vascular smooth muscle cells. It is thought that changes in the vessel wall structure characterized by increased deposition of collagen and elastin degradation layer begins in childhood16 In addition, studies show that the composition of the black arteries contains more collagen. These observed changes in the structure and function of arteries with increasing age are similar to those observed in hypertensive patients, with the difference being at an early age16, 33 It can be speculated that the differences observed in this study between PWV and those performed in Caucasian children is caused by the difference in composition of the arterial wall.

Age is seen as the strongest predictor of PWV in adults and adolescents5, 15, 28, 34 In our study, height, not age, was the best predictor of PWV. It is likely that this finding depends on the narrow age range included in our study. Additionally, in this age range, age and height are two collinear variables (r=0.673; P<.0001). In our multivariate analysis, the higher R ² value was found when height and not age was included in the model. However, when age was forced in the model (instead of height), a very small decrease of R ²=0.052 was observed. However, different results were found by Reusz and colleagues15 in a sample of Caucasian children and adolescents. In our study, the difference between the fifth and 95th percentiles was relatively small, ranging from 0.1 m/s to 0.5 m/s.

Corroborating with findings from other studies, our results show similarity in the growth process in both sexes related to height15, 29, 30, 34 Height becomes sex dependent as puberty develops34 Height was also identified as one predictor of PWV in prepubertal children, a finding also identified by Reusz and colleagues15

In a study of 219 healthy children and adolescents (7–18 years; mean age, 11.3 years), Lurbe and colleagues35 reported increased arterial stiffness in girls. Similarly, in another study evaluating and comparing arterial stiffness between prepubertal and postpubertal children, Ahimastos and colleagues34 found that prepubertal girls had higher PWV in central and peripheral arteries than boys, while in the postpubertal group, the finding was inverse, but only in the peripheral arteries. Presumably, sexual hormones may affect artery stiffness differently since this parameter is influenced by several factors, including endothelium function, proliferation, and organization of the elastic tissue in the vessel's wall, the intravascular distensible forces generated by blood flow.

Mean BP has also been identified as an independent predictor of PWV in children and adolescents15, 29 Therefore, crude reference values of PWV should be derived in groups with normotensive BP values, a procedure adopted in our study. According to our data (Table 3), PWV remained unchanged with age, but BP increased as age increased, a finding also observed in other studies15, 29, 30 These results raise the hypotheses that during the aging process, at least in infancy, a BP increase precedes the increase in arterial stiffness. In adults, the increase in arterial stiffness has been seen as a possible cause of the age‐dependent BP increase. Therefore, the short‐term BP fluctuations can have an impact on vascular properties, independent of the absolute BP levels. It is believed, however, that a chronic increase in BP may accelerate central arterial stiffening, thereby contributing to rapid structural and functional alterations in the walls of these arteries, and thus may create a vicious cycle between arterial stiffness and BP.4, 25, 34, 36

Obesity and central adiposity are also considered by some studies as factors that affect several characteristics of the arterial wall leading to increasing PWV6, 8, 37 BMI was associated with increased PWV in a study of children aged 9 and 10 years,38 which was not found in our study. Other indexes of obesity also were not associated to PWV in our study. Our results corroborate with previous studies in adults7 and in children and adolescents8 It is likely that the PWV increase with obesity would depend on the effect of fat accumulation on BP. Since we removed children with obesity from our analysis, this association was lost in our sample.

Lurbe and colleagues35 found a positive relationship between PWV and PP with birth weight 35in children and adolescents. This association was not found by Batista and colleagues38 who assessed these variables in children aged 9 and 10 years. We did not observe a significant association between birth weight and PWV, but the correlation with PP was negative and significant, confirming the hypothesis that children with low birth weight have greater PP, possibly because of an earlier return of the pulse wave reflection.

Study Limitations

One limitation of our study was the small sample size, mainly when analyses in subgroups were necessary. In addition, all data were obtained from a convenience sample with limited representation in the general population of African children. This sample was chosen because the school was in the neighborhood of the Faculty of Medicine, facilitating the contact and presence of mothers during examinations. However, the chosen school is similar to other public schools attending children of the general population of Luanda. In addition, we did not include data in our analysis of the family history of cardiovascular diseases, food intake habits, and the socioeconomic status of the family. However, it is unlikely that substantial changes in our results would be obtained if a larger sample of children with similar characteristics were tested. However, more robust studies are necessary to define reference values extensive to the spectrum of the population. Moreover, the cross‐sectional design of our study did not establish causality when examining the relationship between several covariates contributing to PWV values.

Conclusions

Our study identified height as the main predictor of PWV in prepubertal Angolan black children. Reference values for this group may be used in further studies investigating the association of higher PWV values with other cardiovascular risk factors in this population. Our data did not show any association between birth weight and BP and PWV in this age population.

Author Contributions

ABTS, MCBM, and JGM designed the study. ABTS explained the study to the direction of the school as well as parents and/or guardians and students at the schools and institutes. ABTS, DPC, and PM participated in the data collection and managed the equipment in the work space. ABTS wrote the article and was responsible for data management, interpretation of the results, and conducting of the literature review. MPB assisted with the data analysis. JGM coordinated the study and participated in all stages of the writing. All authors read and approved the final manuscript.

Supporting information

Table S1. Anthropometric and Clinical Characteristics of Prepubertal Children With High Pressure and Obesity.

J Clin Hypertens (Greenwich). 2016;18:725–732. DOI: 10.1111/jch.12739. © 2015 Wiley Periodicals, Inc.