Blood pressure is an important indicator of cardiovascular and metabolic health. Children and adolescents with high blood pressure levels are highly likely of becoming hypertensive adults. Therefore, early diagnosis and treatment can prevent long-term adverse cardiovascular events (1). Recently, a study has shown that high blood pressure (HBP) at younger ages (6 to 12 years old) may be associated with damage targeting organs such as the heart, brain, kidneys, retina, and blood vessels (2).

Epidemiological studies have shown differences between countries in the prevalence of HBP such as Africa 12.7%, China 7.7%, India 7.6%, United States 13.6%, and Brazil 14.3% (3–7). The southern region of Brazil is one of the regions with the highest prevalence of BPH in adolescents (7). Data on the prevalence of overweight in the same population is worrying, such as Mediterranean region 25.0% (2 to 13 years old), Atlantic region 19.3% (2 to 13 years old), in China, overweight is 13.2% and obesity 9.3% (adolescents aged 12–18 years), the United States, overweight prevalence is 38.7% (adolescents aged 12–15 years) and 41.5% (adolescents aged 16–19 years), and in Brazil, overweight prevalence is 17.5% and obesity is 11.7% (8–11). The southern region of Brazil has the highest prevalence of overweight adolescents in the country (18.2%) (12).

In this sense, studies have pointed out that excess weight (overweight and obesity) in the pediatric population is associated with higher blood pressure values (3, 4, 6, 13). Therefore, when body mass index (BMI) reaches the 85th percentile (overweight) in adolescents over 10 years of age, the risk of developing arterial hypertension in adulthood increases (14, 15). Furthermore, during childhood, excess weight can cause endothelial damage, consequently atherosclerosis and less arterial stiffness (16–18).

Obesity is already considered the cardiovascular risk factor with the greatest association with HBP (19). Previous studies have identified that anthropometric indicators such as BMI, waist circumference (WC), waist-to-height ratio (WHtR), waist-to-hip ratio (WHR), body adiposity index (BAI), conicity index (C), skinfolds (SF), and adiposity body shape index (ABSI) are effective in diagnosing body fat and are associated with HBP in children and adolescents (6 to 19 years old) (20–28).

However, different cutoff points have been used to classify anthropometric indicators and there is no standardization regarding HBP screening in adolescents. This is evident when verifying the different methodological procedures used such as the various protocols for assessing and measuring blood pressure, different instruments used, high range of age groups analyzed, intervals between measurements, number of incongruent measurements, and lack of consensus regarding the best cutoff point for each anthropometric indicator for HBP screening in children and adolescents (5 to 19 years old) (23, 28–36). In addition, some studies do not show diagnostic accuracy measures such as positive predictive values (PPV), negative predictive values (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-), and area under the curve (AUC), which could provide more accurate information about cutoff points to identify the target condition or the individual's health (37).

Considering the relationship between overweight and obesity with HBP in adolescents, the use of anthropometric measures of body adiposity to identify a possible association with HBP in the pediatric population may be an effective and applicable strategy for health professionals to identify risk factors for cardiovascular diseases (13). In addition, our investigation proposes to estimate a greater number of measures of diagnostic accuracy to determine cutoff points of anthropometric indicators of body adiposity for HBP screening. This is important because the literature will have more detailed information on the cutoff points and the reader will be able to choose the anthropometric indicator that best suits the population of interest. Therefore, this study aimed to propose cutoff points for anthropometric indicators for HBP screening in adolescents and to identify, among these indicators, those more accurately for boys and girls.

A total of 634 adolescents aged 14 to 19 years completed all assessments of the present study. Most adolescents were male (62.5%). Significant differences were observed between sexes, with higher BM, height, WHR, ABSI, C Index, WC, and SBP values for male adolescents and higher FP, BAI, triceps, subscapularis, suprailiac, calf SFs, and sum of SFs and DBP values for female adolescents (Table 1).

Yellow/brown/indigenous/black skin color students were more frequent in males (48.2%), when compared to female adolescents (38.6%). Male adolescents had higher prevalence of HBP (29.8%) when compared to female adolescents (11.3%) (p-value = 0.00) (Table 2).

For male adolescents, anthropometric indicators BM, BMI, FP, WHtR, WC, and HC showed sufficient AUC (95%CI of AUC > 0.60), ranging from 0.65 to 0.67, with 95%CI ranging from 0.60 to 0.72. Likewise, LR+ values were >1, demonstrating that individuals are more likely of having the disease with a positive result, and with LR- values lower than 50, demonstrating that an individual with a negative result is 50% less likely of having the disease, for example, an adolescent with BM >64.8 kg is 1.73 times more likely of having HBP than an adolescent with BM <64.8 kg and an adolescent with BM <64.8 kg is 47% more likely of having HBP (Table 3).

Table 3 presents PPV and NPV, with an emphasis on NPV, demonstrating that the use of cutoff points of anthropometric indicators mentioned above presents an 80% chance of not having the disease when the test is negative, that is, if BM <64.8 kg, the adolescent is 83% more likely of not having HBP (Table 3). Anthropometric indicators such as BAI, triceps skinfold, subscapularis skinfold, supra iliac skinfold, calf skinfold, and sum of skinfolds, despite having AUC >0.60, showed 95%CI <0.60. In addition, these anthropometric indicators showed LR- ≤ 50%, demonstrating that an individual, even with a negative result, has a more than 50% chance of having the disease (Table 3).

Regarding cutoff points of anthropometric indicators with better accuracy for HBP screening, BM had cutoff point of >64.8 kg, BMI >21.76 kg/m², FP >15.75%, WHtR >0.41, WC >73.00 cm, and HC >92.25 cm. Furthermore, both anthropometric indicators showed sensitivity >70%, indicating that for every 100 adolescents evaluated, 70 have HBP. However, the same cannot be said about specificity, which presented values <60%, indicating that for every 100 adolescents evaluated, <60 are diagnosed as healthy, and the others are diagnosed with HBP (Table 3).

For female adolescents, BM and height presented AUC >0.60, however, with 95%CI <0.60. The cutoff points for anthropometric indicators with the best accuracy to identify HBP in female adolescents were BM >54.10 kg and height >158.50 cm (Table 4). Both anthropometric indicators showed sensitivity >70%, demonstrating that for every 100 adolescents evaluated, 70 have HBP; however, with specificity <43%, indicating that for every 100 adolescents evaluated, only 43 are diagnosed healthy, and the other 57 have HBP even though they are below the cutoff point (Table 4).

In addition, LR+ values were >1, demonstrating that individuals are more likely of having the disease with the positive result; however, 95%CI presented by BM was between 0 and 1, indicating that the chances of disease are equal between individuals with and without the disease and with LR- value ≥50, demonstrating that an individual with a negative result is 50% more likely of having the disease. Unlike height, which presented LR+ >1 and LR– values close to zero, demonstrating that individuals with the negative result are less likely of having the disease, for example, an adolescent with a height >158.50 cm is 1.54 times more likely of having HBP that an adolescent with height <158.50 cm and, in the same way, an individual with height <158.50 cm with 26% chance of having HBP. Both anthropometric indicators showed high NPV, demonstrating that the use of cutoff points showed a probability of more than 90% of not having the disease when the test is negative, that is, an adolescent with a height <158.50 cm is 90% more likely of not having HBP (Table 4).

The other anthropometric indicators (BMI, FP, WHtR, WHR, BAI, ABSI, C index, WC, HC, triceps skinfold, subscapularis skinfold, supra iliac skinfold, calf skinfold, and sum of skinfolds), which did not show sufficient AUC (95%CI AUC <0.60), also showed LR+ values close to 1 or below 1, demonstrating that individuals who presented a positive test result would have the same chances of having the disease when compared to those who presented a negative result (Table 4).

In the present study, the predictive capacity and proposition of cutoff points of anthropometric indicators to identify HBP in male and female adolescents were investigated. For males, indicators BM, BMI, FP, WHtR, WC, and HC presented sufficient AUC (i.e., 95%CI AUC > 0.60). For females, no anthropometric indicator had discriminatory power for HBP screening. This study is not intended to replace the clinical diagnosis, but to allow the identification, in a simple way, of adolescents who are more likely of having HBP in the initial screening and who would need further care and follow-up.

High values in anthropometric indicators of body adiposity are associated with HBP in adolescents (1, 2, 13, 21–23), and this association is justified by the high concentration of fatty acids in subjects with high body fat. This condition (i.e., high concentration of fatty acids) causes insulin resistance and may induce the renal system to retain more sodium. This retention increases activation of the sympathetic nervous system, which results in increased activity of the renin-angiotensin system and increases blood pressure (62).

This study proposed to update cutoff points of anthropometric indicators for HBP screening in Brazilian adolescents and added more anthropometric indicators for the debate with literature, such as FP, ABSI, triceps, subscapular, suprailiac, calf SFs, and sum of SFs, which were not present in studies published with Brazilian adolescents, in which the use of BMI, WHtR, and WC prevailed (21, 23, 63, 64). Furthermore, previous Brazilian studies have shown sensitivity and specificity values (21, 23, 63, 64), and in two studies (23, 63), the 95%CI sensitivity of anthropometric indicators was below 0.60, and in one study (23), LR+ and LR– values were not presented as 95%CI. AUC allows evaluating the performance on the test, in this case, the performance of anthropometric indicators for HBP screening in adolescents. If the AUC is less than 0.60, the accuracy of the diagnostic test is considered poor, random (37), and may or may not identify HBP in adolescents.

In the present study, for males, anthropometric indicators BM, BMI, FP, WHtR, WC, and HC presented AUC (95%CI >0.60), demonstrating sufficient accuracy for HBP screening (37). The BMI cutoff point for HBP screening in males in the present study was >21.76 kg/m², which is lower than cutoff points in other studies (21, 23, 65, 66). If compared to the BMI cutoff proposed by the World Health Organization (67) to identify obesity in adolescents [14 years (21.8 kg/m²), 15 years (22.7 kg/m²), 16 years (23.5 kg/m²), 17 years (24.3 kg/m²), 18 years (24.9 kg/m²), and 19 years (24.9 kg/m²)], the cutoff point of the present study was lower, which may have occurred due to the following factors: data from the World Health Organization are longitudinal population-based with the participation of several countries and stratified by age, while the present study has a cross-sectional design, carried out in a certain geographic region of Brazil, without stratification by age, and using the average age of adolescents. Although the cutoff point is considered low compared to other studies, this was one of the indicators that showed good accuracy for HBP screening in male adolescents in the present study, since it had an LR+ value close to two, demonstrating that an individual with a positive result is more likely of having the disease, LR– of 0.50, that is, an individual with a negative result is 50% more likely of having the disease and NPV of 82.50 (95%CI: 77.80–86.40).

Regarding the cutoff point for WHtR (>0.41) identified in this study for HBP screening in males, another study presented the same cutoff point of 0.41 (68). The cutoff point of 0.41 for WHtR is below cutoff points in other studies (21, 23, 63, 65, 66, 69–71). Since 1995 (72), the year of the creation of anthropometric indicator WHtR, a universal cutoff point (0.50) has been established to identify overweight individuals from those with normal weight and to be considered a risk factor for cardiovascular diseases, among them, hypertension, which can be used for both sexes and different age and ethnic groups (72, 73). However, this cutoff point of 0.50 was not created from AUC, therefore without diagnostic tests. Furthermore, WC measurement was performed on the umbilical line and not on the smallest portion of the trunk, which could therefore generate differences with the standardization used in the present study (72). In addition, different protocols for measuring WC were used by studies (21, 65, 66), which may interfere with the WHtR results and, consequently, with cutoff point values proposed by these studies.

The present study also identified other anthropometric measures and indicators with good accuracy for HBP screening in male adolescents, namely BM, FP, and HC. BM in the present study presented a sensitivity value of 72.88 (95%CI: 63.90–80.70), LR+ close to two, LR– <0.50, and NPV of 83.40 (95%CI: 78.60–87.30), demonstrating good accuracy for HBP screening in male adolescents, with a cutoff point of >64.80 kg. The other studies in the literature only used BM measurement to calculate other anthropometric indicators. In this sense, this study presents this anthropometric measurement, which is easy to be performed by any professional, as a strategy to be used for the initial HBP screening.

In the present study, HC was analyzed in isolation and showed good accuracy for HBP screening in male adolescents, when high, with a cutoff point of >92.25 cm. In addition, it was used in conjunction with WC and height to calculate other anthropometric indicators, such as WHR and BAI, but without showing good AUC accuracy to identify HBP in male and female adolescents. In other studies, HC was not analyzed in isolation, only in conjunction with other anthropometric measures to calculate BAI (23) and WHR (69) indicators. The study that used BAI showed good accuracy for HBP screening in male and female adolescents and the study that used WHR, carried out only with female adolescents, showed good accuracy for HBP screening, and both studies adopted AUC (95%CI >0.50), different from the present study that adopted AUC (95%CI >0.60) (23, 69).

In the present study, FP, calculated using two anthropometric measurements (height and WC), presented an AUC of 0.65 (95%CI: 0.60–0.70) and a cutoff point of >15.75%. A study carried out with Turkish adolescents aged 11 to 17 years also verified the accuracy of FP for HBP screening in adolescents; however, it was not clear how FP was calculated and, unlike the present study, AUC >0.60 was not identified for this indicator (65). FP calculated by height and HC has not yet been validated for Brazilian adolescents (74); however, this index showed better agreement with dual-energy x-ray absorptiometry (DXA) of 0.83 and 0.86 for female and male adolescents, respectively, compared to BMI and tri-ponderal mass index (TMI), which use is recommended in health care services and the school environment (47). In addition, it presented 3.45 (95%CI: 3.26–3.63) and 3.35 (95%CI: 3.22–3.49) as mean square errors for female and male adolescents aged 15 to 19 years, respectively (47). Furthermore, FP calculated through skinfolds (triceps and subscapularis) (75) is valid for up to the age of 18 years and still depends on constants for ethnicity, which is unnecessary for the calculation of FP with the equation used in the present study that differentiates only between the sex of adolescents and presents a direct estimate of FP (47).

For female adolescents, BM and height presented AUC >0.60; however, with 95%CI <0.60, demonstrating low accuracy to identify HBP in this population. In other studies (21, 23, 63, 64, 69, 70, 76), AUC >0.60 was identified for some anthropometric indicators, such as BMI, HC, WHtR, BAI, and C Index, demonstrating good accuracy for HBP screening in female adolescents. Of these, two studies presented more diagnostic accuracy measures (LR+ and LR–) (23), PPV and NPV (69). The present study did not present anthropometric indicators with good accuracy to identify HBP in female adolescents, which can be explained by the low prevalence of HBP (11.3%), which was below other studies carried out with Brazilian adolescents (7, 77).

A debate in the literature that deserves to be highlighted is the recommended AUC values for diagnostic tests and/or screening. As the present study aimed to propose cutoff points for anthropometric indicators to perform HBP screening among Brazilian adolescents, we chose to use a reference for classifying AUC values >0.60 as sufficient to identify the disease risk (37). There are other references for satisfactory AUC values, such as > 0.70 (78, 79). As there are still debates in the literature with no conclusions on which recommendations to use, it was decided to use AUC values >0.60 aiming at an early screening of the risk of HBP.

This study had strengths and limitations. As a strong point, the fact that 16 anthropometric indicators were investigated and calculated through eight anthropometric measurements is highlighted, enabling verifying the accuracy of the HBP screening in adolescents aged 14 to 19 years. In addition, this study presented measures of diagnostic accuracy not present in other studies, providing additional information to the literature. Finally, this research calculated the technical error of measurement among anthropometrists and used a blood pressure monitor validated for adolescents (80), which demonstrates care during data collection and analysis.

As limitation, the size of the sample collected was smaller than the calculated one, which limits inferences of the study. In addition, despite the measurement of sexual maturation, it was only used to characterize adolescents and did not enter as adjustments in anthropometric indicators. Another limitation of this research was the cross-sectional design, making it impossible to establish a cause-and-effect relationship.

Based on the results, it could be concluded that anthropometric indicators BM, BMI, FP, WHtR, WC, and HC have good accuracy for HBP screening in male adolescents. However, for female adolescents, anthropometric indicators did not present good diagnostic characteristics for HBP screening. Thus, it is suggested that, in a school environment, the use of anthropometric measurements is a simple, low-cost, and easy-to-apply method, which can be performed by the Physical Education teacher, being an effective strategy to contribute to obtaining information about adolescents' health.

This study is not intended to replace the clinical diagnosis of HBP through cutoff points found, but to demonstrate that male individuals with high BM, BMI, FP, WHtR, WC, and HC values should be referred for further clinical tests to verify the presence or absence of the disease.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by Ethics and Research Committee, Federal University of Santa Catarina. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

DS contributed to the study construction, data interpretation, writing, manuscript supervision and review, and obtained approval from the Research Ethics Committee of the Institution to carry out the study. AG, LM, and AU contributed to data interpretation and manuscript editing. LB contributed to data analysis and interpretation, writing, manuscript editing and review, and acquired the authors' authorization for manuscript submission. All authors contributed to the article and approved the submitted version.

This research was financed by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES/PROEX (Brazil) - n° 23038.007266 - by the Graduate Program in Physical Education/UFSC.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.