ABC | Volume 113, Nº1, July 2019

Original Article Rocha et al Dietary patterns and adiposity in children Arq Bras Cardiol. 2019; 113(1):52-59 Methods Study population and design This is a cross-sectional study, with a representative sample of 378 children from public and private schools in the urban area of the municipality of Viçosa, state of Minas Gerais, Brazil. The participants of this study came from the Schoolchildren Health Assessment Survey (PASE, Pesquisa de Avaliação da Saúde do Escolar ). The PASE aimed to investigate the cardiovascular health of children in the city of Viçosa, MG. The municipality of Viçosa is located in Zona da Mata Mineira, 227 km from the capital city, Belo Horizonte. Viçosa has a land area of 299 km 2 and 72,244 inhabitants, with 67.3% residing in the urban area. 15 In 2015, the municipality had 17 public schools and 7 private ones, which were attended by 8- and 9‑year‑old children in the urban area, totaling a population of 1,464 children enrolled in these schools. The sample was calculated using the OpenEpi statistical program (Version 3.01), considering a prevalence of 11.8% for obesity in the age group, 16 a tolerable error of 5%, plus 10% of losses and 10% of confounding factors, totalizing a sample size of 392 children. The final sample included 378 children, with a sample loss of 3.6%. The losses were due to non-compliance with all stages of the study. The sampling process was carried out in two stages. First, stratified casual sampling was performed, in which the number of children to be sampled from each school was proportional to the total number of students in each school. Subsequently, the lots were drawn by using a random number table, to complete the required number of students from the 24 public and private schools in the urban area that were attended by the assessed age group. The non-inclusion criteria for this study were the lack of contact with the parents or guardians after three attempts to contact them, children with some clinical or nutritional alterations that could interfere with food consumption, nutritional status and body composition, as well as children with physical, cognitive or multiple disabilities. This study was carried out in accordance with the guidelines established in the Declaration of Helsinki and was initiated only after approval by the Ethical Committee for Research in Human Beings of Universidade Federal de Viçosa (UFV) (Opinion n. 663.171/2014). The study was also approved by the Municipal Education Secretariat, Regional Superintendence of Education and the school principals. All parents and children were informed about the purpose of the study, and all the children’s parents/guardians signed the Free and Informed Consent form. Food consumption Food consumption assessment was performed by applying three 24-hour food recalls (R24h) on non-consecutive days, including one weekend day and a 15-day mean interval between them, based on information provided by the mother/guardian and by the child. For the children who consumed part of the food in the school environment, the researchers obtained the information from the schools, such as recipes and portions of the foods offered, in addition to confirming with the children what had been consumed. For those children who used to bring their food from home, the parents/guardians were asked about food and beverages offered and their amounts. The three food recalls were applied by trained nutritionists. The food consumption data obtained through the three R24h were tabulated and processed in Dietpro® 5i software, version 5.8. 17 Anthropometric data All anthropometric measurements were performed by a trained nutritionist, selected after calibration of the team members. Weight was measured using a digital electronic scale, with a capacity of 150 kg and sensitivity of 50 g (Tanita®, model BC 553, Arlington Heights, IL, USA). Height was measured using a vertical stadiometer, divided in centimeters and subdivided in millimeters (Alturaexata®, Belo Horizonte, MG, Brazil). The nutritional status was assessed by body mass index (BMI), with BMI cutoff points by age calculated in z-scores according to the World Health Organization (WHO) parameters (2007) 18 as thinness, normal weight, overweight and obesity. Excess weight was considered for the overweight and obesity categories. The waist circumference was obtained by measuring the midpoint between the iliac crest and the last rib using an inelastic measuring tape, divided in centimeters and subdivided in millimeters. Abdominal obesity was considered when the waist circumference was equal to the 90 th percentile of the sample itself, following the guidelines of the International Diabetes Federation (2007). 19 The waist-to-height ratio (WHtR) was obtained by the waist circumference measurement divided by height. The cutoff point ≥ 0.5 was used as the risk for the development of cardiovascular diseases. 20 The neck circumference (NC) was measured at the level of the thyroid cartilage using an inelastic measuring tape, divided in centimeters and subdivided in millimeters. The cutoff points proposed by Nafiu et al. (2010) 21 for the detection of excess body fat in children were used for NC classification. The children’ body composition was assessed through DXA by a specialized technician, obtaining the fat mass measurement. The children were evaluated in the morning, in fasting condition, in the supine position. Body fat was classified using the cutoff points proposed by Lohman (1992), 22 and the cutoff points for overweight risk and overweight were considered as excess body fat. Adjustment variables Potential adjustment variables were selected according to the previous literature. 23-25 The collection of these variables was performed by nutritionists, using a questionnaire created by the researchers themselves. The questionnaire was previously tested in a pilot study, and the sample consisted of children aged 8 and 9 years. 53