Results from quality assessments
Table 2 reported the quality details of bias assessment. Briefly, participants random allocation was not mention in all included articles except in one study that explained the procedure of random sequence generation (31). All of 9 articles did not report allocation concealment. Most of articles represented low risk of bias for selective reporting; nonetheless, 2 articles had high risk of bias (30, 39) and 2 articles had unclear risk of bias (29, 37). All articles had a low risk of bias based on incomplete outcome data. Most of articles had a unclear or high risk of bias for blinding of personnel and participants except just one article that had low risk of bias as to participants and personnel blinding (31). Also, all of articles had a high or unclear risk for blinding outcome assessors. At last, about other potential threats to validity, all of articles had low risk of bias.