Study search and characteristics overview
We retrieved 2153 articles published before December 31, 2019 about the outcome of asthma from PubMed (613 articles), Web of Science (548 articles) and Cochrane Library (992 articles). Duplicated articles were removed automatically (n=358). After reviewing titles and abstracts from 1795 articles, 1782 studies were excluded, including 575 reviews, lecture materials, and randomized controlled study; 115 animal studies; 1021 studies unrelated to our research objective; 44 molecular mechanism studies; 25 articles duplicates excluded manually and 2 non-English studies. Of the remaining 13 full-text publications, five were further excluded due to various reasons, only abstract, incomplete statistics and irrelevant exposure. Finally, eight articles were included in this meta-analysis. (Figure 1)
We retrieved 1431 articles published before December 31, 2019 about the outcome of atopic dermatitis from PubMed (723 articles), Web of Science (240 articles), and Cochrane Library (468 articles). Duplicated articles were removed automatically (n=133). After reviewing the titles and abstracts of 1,298 articles, 1,291 studies were excluded, including 453 reviews, presentation materials and randomized controlled studies;106 animal studies; 707 studies unrelated to our research objective; 11 molecular mechanism studies; 13 repeated studies, and 1 non-English study. Of the \soutleft remaining seven full-text articles, one written in an abstract was excluded. In the end, six studies were included for analysis. (Figure 1)
We retrieved 1,374 articles published before December 31, 2019 on the outcome of food allergy from PubMed (697 articles), Web of Science (223 articles) and Cochrane Library (454 articles). Duplicated articles were automatically removed (n = 123). After reviewing the titles and abstracts of 1251 articles, 1243 studies were excluded, including 426 reviews, lecture materials and randomized controlled studies; 102 animal studies; 690 items unrelated to the research objective; 16 molecular mechanism studies; 9 repeated studies. Of the remaining eight full-text publications, one abstract was excluded. Finally, seven articles were included .(Figure 1)
To sum up, a total of eight articles were included. (Table 1)Several of them provided multiple sets of outcomes, with different effect estimates and 95% confidence intervals. Six studies were cohort studies; two were case-control studies. All studies were human population based. In the two case-control studies, the size of case groups ranged from 916 to 14,803, and the size of control groups ranged from 16,527 to 1,386,029; and the size of cohort studies ranged from 131 to 965,224.