Anthropometric indicators associated with high blood pressure in children living in urban and rural areas

Objective: to evaluate anthropometric and demographic indicators associated with high blood pressure in children aged 6 to 10 years in urban and rural areas of Minas Gerais. Method: this is a cross-sectional study with 335 children. Anthropometric, demographic and blood pressure data were collected. The statistics analyzes were performed using the chi-square, t-student, Mann-Whitney and logistic regression tests, and the odds ratio was the association measure. Results: the prevalence of high blood pressure was significantly higher among rural children. In the urban area, the chance of high blood pressure was higher in children who had a high body mass index (2.97 [1.13-7.67]) and in the rural area, in those who had increased waist circumference (35.4 [3.0-406.2]) and the age range of 9-10 years (4.29 [1.46-12.6]). Conclusion: elevated body mass index and waist circumference were important anthropometric indicators for high blood pressure, as well as age in children living in rural area. The evaluation of body mass index and waist circumference, in addition to nutritional assessments, represents an important action for the screening of high blood pressure in children from different territorial contexts.

that could interfere with blood pressure, and being able to collaborate with blood collection procedures.
The evaluations of the children were carried out between January and July 2015 in the school premises, in reserved rooms and by a team of previously trained nurses.
The rural school had 129 eligible children, but one was excluded because did not accept that blood pressure was measured. In the urban school, 210 children were eligible, but three were excluded, two of them were also not accepted for BP measurement and one for using medication to control BP.
Demographic information, such as sex and age, and anthropometric data, such as height, body weight and waist circumference, were collected from all participating children. Body weight and height were determined in a single measurement using a 0.1 kg precision digital scale and a 0.1 cm precision portable stadiometer (Alturexata ® ). The children were weighed barefoot and wearing light clothing. To get the height, the children stood without shoes, with heels firmly resting on the floor and their knees extended.
Weight and height measurements were used to calculate Body Mass Index (BMI) in kg/m 2 using Anthro-Plus ® software (WHO, Geneva, Switzerland).
BMI was classified as high when Z score was greater than +1 (22) .
The waist circumference was evaluated twice with a non-elastic tape measuring the umbilical scar, with the child standing upright, with the abdomen naked and at the end of a normal expiration. For the final measure the average value of the two evaluations was used. Waist circumference was considered high in cases of percentile ≥ 90 for age and sex (23) . Height and waist circumference measures were used to calculate waist/ height ratio, considered high when ≥ 0.5 (23) .
Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were assessed using a calibrated mercury sphygmomanometer after each child h-ad rested for at least 15 minutes. Blood pressure was measured three times, with an interval of five minutes, in the right arm and with an appropriate size cuff for the child's arm. The arm was placed on a table with the palm facing upwards and the cubital fossa at the level of the lower sternum.
Andrade GN, Matoso LF, Miranda JWB, Lima TF, Gazzinelli G, Vieira EW.  There was no statistically significant difference between the two groups in relation to weight, height, BMI, waist circumference and DBP. However, rural children had significantly higher SBP values than those in the urban area (p <0.001) and lower waist/height ratio (p = 0.022) ( Table 1). The overall prevalence of high BP was 13.7% in the children studied. In the rural area, the prevalence of high BP (18.8%) and high SBP (17.2%) was significantly higher in relation to the urban area (10.6% and 3.4%, respectively) ( Figure 1).

Prevalence (%)
Rural (   After the adjusted analysis, in the rural area, age and waist circumference remained independently associated with the prevalence of elevated BP. In the urban area, children were more likely to present high BP when they had a high BMI (Table 2).
The variable "waist-to-height ratio" was not included in the final regression models due to the presence of collinearity with the variables waist circumference and elevated BMI. These variables were correlated by the Variance Inflation Factor (VIF) correlation test. *Adjusted analysis (urban area) -Nagelkerke R 2 = 0.051; †Adjusted analysis (rural area) -Nagelkerke R 2 = 0.250; ‡OR = odds ratio; §IC = 95% confidence interval; ǁp = Wald test value; ¶BMI = Body Mass Index.

Discussion
High waist circumference, central or abdominal fat index, as well as BMI, an indicator of body fat, were predictors of high BP in rural and urban children,
Recent changes in the eating patterns of Brazilian children, partly due to the improvement of general living conditions in both urban and rural areas, may be contributing to these associations (31) . It is believed that specific characteristics of the regions, such as economic, cultural and lifestyle differences, can mediate this association (14) .
In the rural area, the chance of older children aged 9 to 10 years old presenting with elevated BP was higher than in the age range from six to eight years. Although some studies also show a similar association (17,19) , it cannot be explained by the increase in age, since BP values were adjusted.
The general prevalence of high BP found in the studied children corroborates other studies developed with populations of the same age group (15,27) . In Brazil, this prevalence ranged from 3.8 to 40.6% (14)(15) . However, the higher prevalence among rural children compared to urban children is a result that deserves to be highlighted.
However, the scarcity of studies on the subject, considering both populations, makes comparisons difficult, indicating the need for greater attention to the urban/rural contrasts related to the health-disease process of children.
Blood pressure reading was recorded as the mean of three measurements performed on a single occasion.
Therefore, the probability of occurrence of failures