Table 4.

Summary of primary outcomes.

#	Study	Baseline data	Diabetes development	Glycaemic control
(1)	Knowler et al. [17]	Baseline characteristics for intervention and control had no significant difference.	Diabetes incidence: 4.8 cases per 100 person-years in the intervention group and 11.0 cases per 100 person-years in the control group. Cumulative incidence of diabetes was 14.4% for the intervention group and 28.9% for the control group at the end of the study period.	No primary or secondary outcome
(2)	Kosaka et al. [18]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 3.0% for intervention group and 9.3% for control group. The development of diabetes reduced by 67.4% in the intervention group compared to the control group.	At the end of 4-year study, improvement in OGTT was 53.8% for intervention and 33.9% for control
(3)	Lindström et al. [19]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 9% for intervention group and 20% for control group.	2 h plasma glucose Intervention: baseline = 8.9 ± 1.5, 1 year = −0.9 ± 1.9, and 3 years = −0.5 ± 2.4. Control: baseline = 8.9 ± 1.5, 1 year = −0.3 ± 2.2, and 3 years = −0.1 ± 2.2
(4)	Moore et al. [20]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 13% for intervention group and 7% for control group.	Cumulative incidence of prediabetes at the end of study was 45% for intervention and 67% for control. Moving from prediabetes to nondiabetes was 43% for intervention and 26% for control. 2 h plasma glucose Intervention: baseline = 8.47 ± 1.39 and at 6 months = 7.79 ± 2.31. Control: baseline = 8.08 ± 1.78 and at 6 months = 7.98 ± 2.68
(5)	Penn et al. [21]	Baseline characteristics for intervention and control had no significant difference.	Diabetes incidence of 32.7 per 1000 person-years of follow-up in intervention group and 67.1 per 1000 person-years of follow-up in control group. The overall cumulative incidence of diabetes was 55% less in the intervention group compared to the control group.	Glycaemic control not a primary or secondary outcome
(6)	Ramachandran et al. [22]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 39.3% for intervention group and 55% for control group. Number of people who would need to be treated to prevent one case of diabetes in the intervention group was 6.4.	Glycaemic control not a primary or secondary outcome
(7)	Roumen et al. [23]	Age was higher in the control group than intervention: 54 years and 52 years, respectively. No other significant differences.	Cumulative incidence of diabetes was 18% for intervention group and 38% for control group.	2 h plasma glucose Intervention: baseline = 8.59 ± 0.24, at 1 year = 7.96 ± 0.29, and at 3 years = 8.55 ± 0.34. Control: baseline = 8.46 ± 0.23, at 1 year = 8.83 ± 0.29, and at 3 years = 9.35 ± 0.33
(8)	Saito et al. [24]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 12.2% for intervention group and 16.6% for control group.	Glycaemic control not a primary or secondary outcome. States that the fasting plasma glucose or 2 h plasma glucose levels significantly decreased more in intervention group than in control group
(9)	Xu et al. [25]	Baseline characteristics for intervention and control had no significant difference.	Cumulative incidence of diabetes was 14.6% for intervention group and 17.5% for control group.	Reverting to normal glucose levels at the end of study was 39.0% for intervention and 7.5% for control

OGTT: oral glucose tolerance test.