Реферат
Abstract Objective: To estimate the prevalence of childhood obesity in Brazil by means of a systematic review of representative studies. Sources: We searched for population-based studies that assessed obesity in Brazilian children aged < 10 years in MEDLINE, EMBASE, Scopus and other sources up to September, 2019. Paired researchers selected studies, extracted data and assessed the quality of these studies. Meta-analysis of prevalence and confidence interval (95% CI) was calculated, weighted by the population sizes using Freeman-Tukey double-arccosine transformation. Heterogeneity (I2) and publication bias were investigated by meta-regression and Egger's test, respectively. Summary of the findings: 53 studies were included (n = 122,395), which were held from 1986 to 2015 and limited mainly due to inadequate response rates. Prevalence of obesity in the threedecade period was of 8.2% ([95% CI]: 8.1-8.4%, I2 = 98.5%). Higher prevalence was observed in boys (9.7% [9.4-9.9%], I2 = 97.4%) than girls (7.3% [7.1 7.5%], I2 = 96.1%). Prevalence increased according to the decade (1990: 6.5% [6.0-7.0 %], I2 = 96.8%; 2000: 7.9% [7.7-8.0 %], I2 = 98.8%; 2010: 12.0% [11.5-12.6 %], I2 = 95.8%), and Brazilian region (Northeast: 6.4% [6.2−6.7%], I2 = 98.1%; North: 6.7% [6.3−7.2%], I2 = 98.8%; Southeast:10.6% [10.2−11.0%], I2 = 98.2%; South: 10.1 [9.7−10.4%], I2 = 97.7%). Heterogeneity was affected by age and region (p < 0.05) and publication bias was discarded (p = 0.746). Conclusion: For every 100 Brazilian children, over eight had obesity in the three-decade period and 12 in each 100 had childhood obesity in more recent estimates. Higher prevalence occurred in boys, recent decades and more developed Brazilian regions.