ABC | Volume 112, Nº6, June 2019

Original Article Scheer et al Exercise at school and cardiovascular risk factors Arq Bras Cardiol. 2019; 112(6):775-781 Table 1 – Comparison of demographic characteristics, anthropometric data and cardiovascular risk factors between students from regular schools (RSch) and Olympic Experimental Gymnasiums (OEG) Variable OEG RSch p value Age (years) 12.6 ± 1.2 (n = 719) 13.3 ± 1.5 (n = 394) BMI (kg/m 2 ) 20.5 ± 4.3(n = 716) 21.4 ± 4.4 (n = 394) 0.001 Overweight 191/716 (26.7%) 133/394 (33.8%) 0.01 Altered glycemia 1/700 (0.1%) 0/393 (0.0%) NA Altered total cholesterol 270/714 (37.8%) 152/393 (38.7%) 0.78 Altered triglycerides 403/624 (64.6%) 253/387 (65.4%) 0.80 Pre-hypertension/hypertension 116/712 (16.3%) 112/393 (28.5%) < 0.001 Data are reported as mean ± standard deviation or absolute number (percentage). BMI: body mass index; NA: not applicable; p-value obtained by chi-square test or Student’s t test. Results A total of 1,113 students were screened, included, and evaluated. Of these, 719 were attending OEG schools and 394 were attending RSch. Table 1 shows comparative data between OEG and RSch students as well as the number of students analyzed for each category. RSch students had a higher BMI than OEG students (21.4 ± 4.4 vs 20.5 ± 4.3 kg/m 2 , p < 0.001), and overweight was also more prevalent in RSch students (33.8% vs 26.7%, p = 0.001). The prevalence of high BP was higher in RSch students than in OEG students (28.5% vs 16.3%, p = 0.013). RSch and OEG schools did not differ in the frequency of altered glycemia, total cholesterol or triglycerides. Despite our efforts to match the groups for sex, the χ 2 test indicated differences between groups, with a higher proportion of girls in RSch than in OEG schools (64.0% vs 49.4%, p < 0.001). Due to this difference, we performed a sensitivity analysis stratified by sex and the results are shown in Table 2. The difference observed in BMI between RSch and OEG students was limited to girls (21.9 ± 4.5 vs 20.6 ± 4.3 kg/m 2 , p = 0.001), with no difference when only boys were compared (20.6 ± 3.9 vs 20.3 ± 4.2 kg/m 2 , p = 0.564). There was a difference in the prevalence of overweight between RSch and OEG girls (35.7% vs 24.3%, p = 0.002), but not among boys (30.3% vs 29.0%, p = 0.777). There was no difference in any other parameter assessed. To provide estimates based on associations, a binary logistic regression model was used for each health marker, adjusted for age and sex, and the results are shown in Table 3. After adjusting for sex and age, RSch students were more likely to be overweight (OR 1.49, 1.13–1.98) and to have pre-hypertension and hypertension (OR 1.86, 1.36–2.34), while no differences were observed for altered total cholesterol or triglycerides. After stratifying by sex, due to the previously found gender differences between schools, the girls attending RSch had higher odds to be overweight (OR 1.89, 1.30–2.75) and pre‑hypertensive/hypertensive (OR 1.66, 1.10–2.51), with no association with altered total cholesterol and triglycerides. For boys, only the association between attending RSch and being pre-hypertensive/hypertensive was found to be significant (OR 2.20, 1.37–3.54). Discussion In this cohort study, we confirmed our primary hypothesis that students attending OEG schools had lower cardiovascular risk prevalence than students attending RSch. Our results are important because they highlight the association between a healthy school policy and reduction of CVD risk factors in adolescent students. A proportion-difference test showed that overweight and pre-hypertension/hypertension were less frequent in OEG schools. The prevalence of overweight students in OEG schools was approximately 11% lower. The difference was even lower for pre-hypertension/hypertension (approximately 13%). Glucose levels were not altered in most cases. Although no differences were found between schools, the lipid profile was altered in almost half of the students. The 2016 ERICA study, the largest cross-sectional study involving all regions of Brazil to assess the frequency of cardiovascular risk factors in students, also evaluated individuals of the same age as ours (12-14 years old) and of the same geographic region (southeast). 13,14 It should be noted that the ERICA study found a lower rate of pre‑hypertension/ hypertension (6.5%, 5.5–7.8%) and a lower overall prevalence of overweight (18.4%, 15.5–21.6%) in the subgroup of individuals of the same age range and residents of the same geographic region than those observed in our study. Conversely, Cureau et al. 14 reported a prevalence of 31.3% for high BP and 23.9% for overweight in a similar group. Several reasons may account for these differences, mostly regarding the criteria used to define pre-hypertension/hypertension and overweight, in addition to the use of different methodology, including the measurements obtained by aneroid and oscillometric devices. 15 As for the differences between OEG schools and RSch, Cureau et al. 14 provide useful findings to explain them, since the prevalence of physical inactivity in their cohort was estimated around 51%. Considering the attributable risk fraction of physical inactivity to develop hypertension and obesity, 16-18 it seems 777