Introduction

Hypertension—or elevated blood pressure—is a serious medical condition that significantly increases the risks of heart, brain, kidney, and other diseases (WHO (World Health Organisation 2019). Several meta-analyses have shown a log-linear relationship between high blood pressure and an increased risk of cardiovascular disease which increases considerably with age (Tackling and Borhade 2019. According to the World Health Organization, 39% of adults aged 18 years and over were overweight in 2016, and 13% were obese (WHO 2018). Currently, it is established that obesity increases the risk of diabetes, hypertension, and dyslipidemia, which have an impact on cardiovascular diseases (CVD) mortality and morbidity (Visscher and Seidell 2001; Klein et al. 2007). The same reports were made by other authors who have shown that high blood pressure promotes the risk of various cardiovascular diseases (Flint et al. 2019; Rapsomaniki et al. 2014; Whelton et al. 2017), such as strokes, coronary heart disease, heart failure, fibrillation atrial (Angeli et al. 2014), and peripheral vascular diseases. HBP is considered the main risk factor for mortality in the world, and according to the WHO, it is responsible for 7.5 million deaths per year, which represents 12.8% of all deaths (WHO 2009). For 2025, researchers have estimated that 29.2% of the adult population will be hypertensive, or 1.56 billion people, for an increase of 60% in 25 years. So it seems that hypertension is a major public health problem all over the world (Kearney et al. 2005). Body mass index (BMI) has become the most widely used index to evaluate overweight and obesity. Some authors identify a new phenotype of metabolically healthy obese who lack the usual cardiometabolic risk factors of the obese subject (Disse 2014; Boirie et al. 2016). Others find that BMI is losing interest in older people. Beyond age 75, the increase in BMI is protective and no longer appears to be an independent risk factor for mortality (Stevens et al. 1998). However, this finding does not corroborate with other studies that have shown that increasing BMI in the elderly is not automatically a protective factor (Osher and Stern 2009). The metabolic disorders of obesity were then attributed to the android distribution of fat rather than to the overall corpulence (Kang et al. 2011; Oppert 2003; Vague 1956). Nonetheless, the link between android fat deposition and its association with cardiovascular risk has recently been disputed (Sari et al. 2019). A lot of attention worldwide use has also been made of the use of waist circumference (WC) for assessment and risk management of CVD because waist circumference is strongly correlated with abdominal fat (Klein et al. 2007). Recently, the WHtR has been proposed as the best screening tool than WC and BMI for cardiometabolic risk factors in adults. Thus, anthropometric measurements evaluating abdominal fat were proposed to better predict the comorbidities of obesity, such as hypertension (Lee et al. 2008). In fact, in the context of large-scale epidemiological studies and in the absence of possible measurement of blood pressure, this possible anthropometric measurement would be a simple, inexpensive, and effective asset to detect people at high risk of hypertension. In fact, in the context of large-scale epidemiological studies and in the absence of possible measurement of blood pressure, this possible anthropometric measurement would be a simple, inexpensive and effective asset to detect people at high risk of hypertension according to their BMI, WC and WHtR. It is in this perspective that we conducted this study which aimed to analyze the associations between different anthropometric measures and hypertension in an Algerian population composed of overweight, obese, and normal weight adults.

Methods and materials

This retrospective analytical study was conducted between 2013 and 2015 in a population of adults coming for consultation in different medical centers for health problems related to the metabolic profile in Tebessa, a city in eastern Algeria. A total of 900 subjects aged 18 to 92 years were invited to participate in the study, including 785 (529 women and 256 men) having complete data on hypertension, diabetes and dyslipidemia, sex, and age, and the three anthropometric indices (BMI, WC, WHtR) were included in the analyses. The surveyed subjects were selected among patients coming to consultation in different sanitary sewers of the city of Tebessa (Polyclinic Bachir Mentouri, Polyclinic Skanska, Laboratory of Biochemical Analyses (Dr. Brahmi), Hospital Establishment Bouguerra Boulaares Bekkaria, Tebessa). We included in this study any person aged 18 and over coming in consultation in different care centers and having agreed to participate in our study regardless of gender. Excluded were those who refused to participate, those who did not want to take a blood sample those who refused to make anthropometric measurements (weight, height, and waist circumference), and pregnant women. We also excluded anyone who was unable to understand and answer our questions (very sick subjects, etc.). A standardized questionnaire was used to collect information on age, sex, physical activity, personal and family history, tobacco consumption, medical history, socioeconomic level, and education level. The subjects surveyed were subdivided into five age groups by sex in order to study the prevalence of hypertension according to these two unchangeable variables.

Measurements of height and weight were carried out by students who were trained in precise measurement techniques. Height and weight were measured in lightweight clothing without shoes using a standardized WHO protocol. The weight was measured on a scale “TIAN SHAN-2003A” measuring its capacity of 180 kg (396 Ib), its accuracy is 100 g for a kg, and each subject must remove the jacket and shoes, stand upright and straight on the scale. The height of the subjects (without shoes) was measured to the nearest 0.1 cm with a portable stadiometer. BMI (kg/m2) was calculated by dividing the weight by height squared. The WC was measured with a flexible measuring tape at a level midway between the lowest coastal margin and the iliac crest at the end of a normal exhalation. WHtR was calculated as WC (in cm) divided by height (in cm). We referred to the World Health Organization (WHO) criteria for defining the weight status of patients (WHO 2000).

Two consecutive readings of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were taken to obtain a mean value that was used in this study. A rest of 10 min was granted to the subjects surveyed before the blood pressure was taken. The blood pressure was measured by an automatic blood pressure monitor “OMRON M3.” Diagnosis of arterial hypertension was retained if the systolic and diastolic arterial pressure was higher than or equal to 140/90 mmHg or if the patient was already known to be hypertensive or under treatment with antihypertensive medication (Mancia et al. 2013). A blood sample was taken for the assay of biochemical parameters, total cholesterol, triglycerides, LDL cholesterol, HDL cholesterol, and blood glucose.

The data were captured and analyzed using the SPSS software version 25. In the descriptive study, we calculated simple frequencies and relative frequencies for qualitative variables and means and standard deviations for quantitative variables. In the analytical study, the ANOVA was used to compare the averages of the three quantitative variables with a normal distribution. The chi-square test was used for the qualitative variables for the comparison between the percentages. The link between two quantitative variables was studied by the Pearson correlation coefficient. Receiver operating characteristic (ROC) curves were established for anthropometric parameters (weight, BMI, TT, and RTT) to distinguish between hypertensive and non-hypertensive patients. We used the cutoff value of the variable that had the best “sensitivity-specificity” pair. This value corresponded to the point of the ROC curve closest to the point having the coordinates (x = 0, y = 1). All the subjects interviewed agreed to participate in this study. Those who did not agree were dropped from the study.

All the analyses were carried out in laboratories approved by the Algerian state by respecting the laws of the official newspaper of the Algerian Republic No. 35 of August 15, 1990. The study was approved by the Regional Ethics Committee for Algerian state. A verbal consent was obtained from each participant. All methods were applied in accordance with relevant guidelines and regulations. In all statistical tests, the cutoff of significance was set at 0.05.

Results

The final study sample consisted of 785 participants (529 (67.4%) women and 256 (32.6%) men with a median age of 47.17 ± 16.36 years (46.4 ± 16 years in women and 48.8 ± 17 years in men)). No significant difference was observed for the mean age by sex. The prevalence of hypertension was 49.7% (women 47.26%, men 55.08%, P = 0.04).The prevalence increased significantly by age in both males and females (Fig. 1), from 24.83% among those aged between 18 and 29 years (men 34.09%; women 20.95%) up to 78.75% among those aged 70 and over (men 81.08%; women 76.74%). Men had significantly higher mean SBP and DBP than women (systolic 131.2 versus 126.9 mmHg, P = 0.009; diastolic 78.0 versus 75.2 mmHg, P = 0.008). No significant difference was observed between hypertensive men and women concerning the mean values of SBP and DSP. The comparison of hypertensive subjects with non-hypertensive subjects revealed statistically significant differences in the clinical and anthropometric indices between the two groups. The anthropometric data for hypertensive vs. non-hypertensive were the following: BMI, 29.08 ± 5.54 (kg/m2) vs. 27.71 ± 5.5 (kg/m2); WC, 100.3 ± 16.9 cm vs. 96.8 ± 16.1 cm; and WHtR, 0.60 ± 0.10 vs. 0.58 ± 0.10. The clinical data were the following: SBP (mmgH), 145.0 ± 17.4 vs. 111.76 ± 8.99; DBP (mmgH), 83.9 ± 12.3 vs. 68.31 ± 9.46; TC (g/l), 1.88 ± 0.48 vs. 1.79 ± 0.46; and TG (g/l) 1.39 ± 0.61 vs. 1.29 ± 0.57 (Table 1).

Fig. 1
figure 1

Prevalence of hypertension by age and sex

Full size image
Table 1 Anthropometric and clinical characteristics of hypertensive and non-hypertensive subjects
Full size table

A positive correlation was observed between BMI and SBP, BMI, DBP, and waist circumference (r = 0.11, P = 0.03). Mean age was 52.4 years and 42 years, respectively, among hypertensive and non-hypertensive (P < 0.0001).

Descriptive information of normal weight, overweight, and obese is shown in Table 2.Obese had a higher mean weight, WC, WHtR, SBP, DBP, FG, TC, and TG than overweight and normal weight. The results are respectively in normal weight group, overweight group, and obese group: weight (kg), 62.49 ± 7.57 vs. 76.42 ± 8.23 vs. 92.36 ± 13.24; WC (cm), 85.82 ± 11.28 vs. 97.33 ± 11.68 vs. 110.69 ± 15.4; WHtR, 0.51 ± 0.06 vs. 0.58 ± 0.07 vs. 0.67 ± 0.09; SBP (mmHg), 124.68 ± 21.83 vs. 128.21 ± 21.65 vs. 131.55 ± 20.97; DBP (mmHg), 73.37 ± 13.95 vs. 76.78 ± 13.72 vs. 77.86 ± 12.44; FG (g/l), 1.08 ± 0.5 vs. 1.11 ± 0.43 vs. 1.24 ± 0.58; TC (g/l), 1.7 ± 0.42 vs. 1.81 ± 0.45 vs. 1.97 ± 0.48; and TG (g/l), 1.18 ± 0.55 vs. 1.35 ± 0.52 vs. 1.47 ± 0.65. The prevalence of hypertension and diabetes was higher in obese than overweight and normal weight. The hypertension data were the following: 56.69% vs. 49.80% vs. 41.94%. For diabetes data were the following: 40.82% vs. 34.18 vs. 25%. The prevalence of obesity was higher in women than in men (40.45% vs. 27.34%, P = 0.001) (Fig. 2). The prevalence of hypertension increased significantly with BMI from 41.94% among normal weight to 49.8% among overweight to 56.69% among obese (P = 0.003). When the BMI increases, the prevalence of hypertension increases significantly. We conclude that the prevalence of HBP is significantly higher in obese people than in overweight and normal weight (Fig. 3). The findings remained the same after adjusting for sex. However, hypertensive women are significantly more obese than men (0.026). Our results showed that obesity increased the risk of hypertension by a factor of 1.54 (95% CI [1.15, 2.06], P = 0.004).

Table 2 Anthropometric and clinical characteristics of the population studied according to the body mass index
Full size table
Fig. 2
figure 2

Prevalence of obesity and overweight by gender of subjects surveyed

Full size image
Fig. 3
figure 3

Prevalence of obesity and overweight among hypertensive and non-hypertensive subjects surveyed

Full size image

The Pearson correlation for the association between anthropometric indicators and each of SBP and DBP is presented in Table 3. For total population, BMI was positively correlated with SBP and DBP, respectively. However, WC and WHtR were significantly correlated with SBP. No significant correlation was found between WC, WHtR and DBP. According to the BMI group, Pearson’s correlation data analysis showed that there was no relationship between SBP and anthropometric parameters (BMI, WC, and WHtR). Only DBP was negatively associated with WHtR in overweight group (r = − 0.151, P = 0.015 (Table 3).Given the continuous and quantitative characteristics of weight, BMI, WC, and WHtR, we realized ROC curves for each anthropometric parameter in order to evaluate its association with arterial hypertension. All these parameters had areas under the curve between 0.409 and 0.618. Their contributions were thus low (Table 4). The cutoff value of anthropometric WHtR parameters associated with the risk of hypertension was higher among women than men regardless of the BMI group considered. Table 5 illustrates the cutoff values, sensitivities, and specificities of the different anthropometric indices. For normal BMI group, weight cut points were 61.2 kg, 59.5 kg for women (54% sensitivity, 45% specificity) and 66.05 kg for men (45% sensitivity, 44% specificity); BMI cut points were 22 kg/m2, 22 kg/m2 for women (54% sensitivity, 54% specificity) and 22 kg/m2 for men (55% sensitivity, 53% specificity); WC cut points were 85.5 cm, 85.5 cm for women (62% sensitivity, 62% specificity) and 84.5 cm for men (50% sensitivity, 46% specificity); and WHtR cut points were 0.52, 0.52 for women (62% sensitivity, 61% specificity) and 0.50 for men (55% sensitivity, 59% specificity).

Table 3 Pearson’s correlation coefficients between anthropometric parameters and blood pressure in normal BMI, overweight, and obese
Full size table
Table 4 Areas under the ROC curve of different anthropometric parameters distinguishing hypertensive patients from normotensive patients according to their BMI
Full size table
Table 5 Cutoff, sensitivities, and specificities of anthropometric indices in the prediction of high blood pressure
Full size table

For overweight group, weight cut points were 76.8 kg, 73.6 kg for women (52% sensitivity, 50% specificity) and 80.5 kg for men (61% sensitivity, 56% specificity); BMI cut points were 27 kg/m2, 27 kg/m2 for women (52% sensitivity, 52% specificity) and 27 kg/m2 for men (56% sensitivity, 52% specificity); WC cut points were 99.5 cm, 95.5 cm for women (46% sensitivity, 50% specificity) and 99.5 cm for men (60% sensitivity, 56% specificity); and WHtR cut points were 0.59, 0.61 for women (47% sensitivity, 52% specificity) and 0.58 for men (54% sensitivity, 52% specificity).

For obese group, weight cut points were 90 kg, 90 kg for women (48% sensitivity, 57% specificity) and 95 kg for men (50% sensitivity, 58% specificity); BMI cut points were 33 kg/m2, 33 kg/m2 for women (52% sensitivity, 55% specificity) and 33 kg/m2 for men (47% sensitivity, 55% specificity); WC cut points were 110 cm, 110 cm for women (48% sensitivity, 49% specificity) and 110 cm for men (39% sensitivity, 46% specificity); and WHtR cut points were 0.68, 0.69for women (52% sensitivity, 51% specificity) and 0.64for men (45% sensitivity, 46% specificity).

Discussion

Hypertension is one of the main health problems in the world. A better understanding of anthropometric risk factors and their correlations was the priority. Given the results mentioned in this study, our work aimed to determine the relationship between blood pressure and anthropometric indices (using WC, WHtR, and BMI as indicators) in adult Algerian subjects. In our study as in other studies, the risk of hypertension increased with age (Tazi et al. 2009; Dua et al. 2014). Blood pressure increases with age. Indeed, aging promotes the loss of elasticity of the arteries. The prevalence of hypertension in our study was higher than that of other studies 49.7% vs. 44.8% (Ho Anh Hien et al. 2018). However, time differences in blood pressure (BP) visits, BP measurement methods, sample size, age of subjects examined, ethnicity, socioeconomic status, and differences in geographic areas between studies make it difficult to compare the results. Nevertheless, several studies have suggested that blood pressure is a common health problem. In this study, the prevalence of hypertension was significantly higher among men as compared with women as in other studies (Everett and Zajacova 2015, Choi et al. 2017, Ho Anh Hien et al. 2018).

The study demonstrated that among Algerian adults, the presence of abdominal obesity is a risk factor for hypertension. The positive association observed in our study between waist circumference and hypertension has been reported in other studies (Bose et al. 2003; Huang et al. 2007). Abdominal obesity has been recognized as a risk factor for CVD. Many health risks have been linked to abdominal obesity including diabetes, hypertension, cardiovascular disease, arthritis, respiratory disease, breast cancer, ovarian dysfunction, menstrual irregularities, and bad social image (Onuoha et al. 2016). Abdominal obesity, assessed by the waist measurement (WC), has been shown to better predict the risk of these long-term cardiometabolic complications. The waist circumference of overweight and obese categories were significantly (P < 0.0001) higher than the normal weight category. The two measures together, WC and BMI, are even better predictors of these diseases if they are combined. In reality, it is well shown today that it is the excess of fat mass and in particular its perivisceral or abdominal localization which is involved in this predisposition to cardiometabolic complications (Pataky et al. 2009). The relationship between obesity and arterial hypertension has been documented in the literature (Cambien 1982; Gebel 2011; Praso et al. 2012). This study confirmed the relationship between obesity and arterial hypertension since its prevalence is significantly higher in obese than overweight and normal weight. The same findings were reported by Longo-Mbenza et al. (2007). On the other hand, our study has shown that DBP and SBP were positively correlated to the BMI. Results of a study conducted in Serbia by Pantelinac (2007) showed that systolic and diastolic blood pressures are significantly associated with BMI. According to the results of a study conducted in Morocco, Tazi et al. (2009) showed that the risk increased with increased body mass index, waist circumference, and hypercholesterolaemia. In Tunisia, researchers have shown in 5802 subjects that obesity is an important risk factor for the development of hypertension (Boujnah et al. 2018). Some studies suggest that the association between obesity and blood pressure could be explained by a mechanism related to the variation in the amount of abdominal fat estimated by a single clinical parameter waist circumference. Among the obese in this study, 87.32% had abdominal obesity. Our results also showed that WC and WHtR were significantly correlated with SBP. In this study, the prevalence of obesity is significantly higher in women than in men, but the prevalence of arterial hypertension is higher in obese men than in obese women. Several studies have reported a high prevalence of female obesity in relation to men in both urban and rural settings (Otang-Mbeng et al. 2017; Rheeder et al. 2017). The WC cutoff points corresponding to BMI values of 25 kg/m2 were 99.5 cm in men and 95.5 cm in women. When the BMI is greater than or equal to 30 kg/m2, the cut points were 110 cm in both men and women. In this study, a positive correlation was found between BMI and WC, r = 0.716, P < 0.0001. Taking into consideration gender, we found a linear and higher correlation between BMI and WC in men than in women. Another study carried out in Tunisia showed WC cutoff for a BMI equal to 25 kg/m2 lower than those found in this study for both men and women 85 cm and 79 cm, respectively (Bouguerra et al. 2007).

Second, we concluded that all the anthropometric parameters studied were equal in terms of HTA identification in all three groups. Threshold values of anthropometric parameters to distinguish between hypertensive patients and non-hypertensive patients, regardless of their BMI, were not very sensitive or very specific. As a result, in emerging countries such as Algeria, anthropometry could not be an interesting public health asset in the detection of hypertension in people with normal BMI as in the overweight and/or obese. The same findings were shown by Sebai et al. (2018) in patients with type 2 diabetes. Other studies have shown that WHtR has performed better than BMI and WC for the association with hypertension and diabetes (Cai et al. 2013). Third, the BMI cutoff associated with the presence of hypertension in normal BMI, overweight, and obese subjects was equal among men and women in each BMI group. No significant association was observed. In this study, we have more women than men, and this is a limitation of this work. The high number of women is due to their significant presence in the places where the survey is conducted. This may explain some results obtained in this investigation.

Conclusions

In this study, blood pressure is significantly linked to obesity, suggesting the role of weight loss in the treatment of hypertension in overweight and/or obese people. In addition, unlike the data in the literature, the discriminating power of anthropometry in predicting HTA is limited or absent regardless of the value of BMI. Its interest in epidemiological studies in developing countries would therefore be modest. Studies at the national level are necessary to support the validity of our results and to identify specific references to the Algerian population while taking into account the overweight.