Study screening and data extraction
Retrieved articles were independently screened according to titles and abstracts by two reviewers. Any discrepancy was resolved by discussion. The same process was employed for full-text screening. The full texts were assured by cross checking by consensus for validation and accuracy. The following data as the characteristics of various studies was extracted from eligible studies: first author’s name, publication year, research design, sample size, country, type of allergic diseases in the children, effect estimates (odds ratios, hazard ratios, risk ratios or prevalence ratios) and 95% confidence interval. All data was imported to Excel for analysis.