ABC | Volume 115, Nº2, August 2020

Original Article Jesus et al. Meta-analysis: body adiposity and apolipoproteins Arq Bras Cardiol. 2020; 115(2):163-171 heterogeneity. 20 In case of high heterogeneity (I 2 greater than 50%), meta-regression was used. Considering that less than ten studies were included in the meta-analysis, it was not possible to analyze publication bias with the Egger test and the Funnel plot. However, the comprehensive, sensitive and unrestricted search based on language and year contributed to reduce publication bias. For all analyses, a p-value less than 5% was adopted as statistically significant. They were performed in the statistical package Stata for Mac version 12 (Stata Corp, College Station, TX, USA), using the command “metan” to obtain the WMD. Results Results of the Systematic Review Selection of Studies In the systematic search, 7,116 articles were identified, of which 3,978 were duplicated. After reading heading and summary and verifying eligibility, 3,118 articles were excluded. Thus, 20 articles were selected for full reading, being 12 excluded for the following reasons: did not present data on ApoB, ApoA1 according to excess adiposity (three studies); sample composed of individuals with CNCD (hypertension, diabetes, metabolic syndrome or dyslipidemia – nine studies). In total, eight articles were chosen for the systematic review. Four had all the information about exposure and outcomes and were, therefore, included in the meta-analysis (Figure 1). Study Characteristics The main characteristics of the eight studies included in the systematic review are described in Table 1. As to origin, one study was carried out in Japan, two in Australia, one in the United States of America, two in England, one in Sweden, and one in Canada, having been published between 2001 and 2016. The sample size ranged from 59 to 7,589 individuals of both sexes, totaling 15,835 individuals, with mean age of 9 to 15.7 years. The minimum follow-up time was 12 months and the maximum was 144 months, periods that are sufficient for the phenomenon to occur. Risk of Bias The articles were evaluated using six questions from RTI, with all eight articles considered to be at low risk of bias. Figure 1 – Flowchart of the systematic review. 7,116 titles identified that presented the search strategy 3,138 articles selected 3,978 duplicates excluded 20 articles selected for full reading 8 articles selected for systematic review 4 articles included in the meta-analysis. 3,118 articles excluded after reading of heading and abstract 12 articles deleted after full reading 3 did not evaluate ApoB or adiposity 1 had different objective Inclusion Eligibility Screening Identification 165