Xuxi Zhang

Data analysis The heterogeneity between studies was evaluated by I 2 test. If I 2 ≤ 50%, the possibility of heterogeneity between studies was low and fixed effect model could be utilized. If I 2 ≥ 50%, there was heterogeneity between studies and the sources of heterogeneity should be analyzed. We performed continuous data meta ‐ analysis and adopted the weighted mean difference (WMD/MD) as the effect indicator. The criteria for significance is P < 0.05 and 95% CI not including 0. Sensitivity analysis was also performed to test the stability of the studies included. The possible publication bias was assessed by funnel plots with Begg and Egger tests with P < 0.1 as the criterion for significant publication bias. All statistical analyses were performed with Stata 11.0. GRADEpro 3.2 was utilized for the GRADE evidence strength assessment and table design. RESULTS Study selection and study characteristics The twenty articles in the review included a total of 4494 participants (Fig. 1). Table 1 and 2 present the characteristics and HbA 1c results of each study. The number of participants of each study ranged from 46 27 to 628 17 . One study 27 only included female participants and one 26 only included male participants. All participants had type 2 diabetes in 16 studies 13 ‐ 15, 17 ‐ 20, 22 ‐ 25, 27, 28, 30 ‐ 32 . In the remaining 4 studies 16, 21, 26, 29 , type of diabetes was unspecified but mean age was over 30 years old. In 3 studies 19, 24, 32 , all participants were Mexican American. Two studies 21, 27 reported that participants were African American, and three 16, 31, 15 reported participants were Hispanic, Spanish ‐ speaking or Latino adults. The intervention duration ranged from 1.25 months 16 to 24 months 14, 23 . Five studies 13, 15, 20, 24, 32 had more than one follow ‐ up interval which ranged from  ‐ 9 15 to 12 months 13 . Two studies 28, 30 also had more than one follow ‐ up interval but we only included one interval of each study because of the quality of results. Most studies 14, 17 ‐ 19, 21 ‐ 23, 25 ‐ 27, 29, 31 only had 0 ‐ month follow ‐ up interval. One study 16 only had 1.75 ‐ month follow ‐ up interval. Moreover, the details of intervention in IG and CG can be found in Supplementary Table S1. 6 133 Peer support among adults with type 2 diabetes