Five RCTs were included in the review with 3,796 participants (range 577 to 1,246). All the trials were randomised and double blind, but none detailed allocation concealment. Withdrawal rates were around 20% in all trials. Follow-up ranged from 18 to 24 weeks
Glycated haemoglobin (HbH1c): Compared with metformin monotherapy, initial combination therapy with dipeptidyl peptidase-IV inhibitors resulted in statistically significantly lower glycated haemoglobin from baseline (WMD -0.55%, 95% CI -0.63 to -0.46; Ι²=0%). Patients that received dipeptidyl peptidase-IV inhibitor initial combination therapy were more likely to achieve the glycated haemoglobin goal of below 7% than those who received metformin monotherapy (RR 1.55, 95% CI 1.43 to 1.67).
Fasting plasma glucose: Patients who received combination therapy reported a lower fasting plasma glucose than those on monotherapy (WMD -0.97 mmol/L, 95% CI -1.26 to -0.68). However, this analysis showed significant heterogeneity (I²=68%). Sensitivity analyses excluding studies using sitagliptin reduced heterogeneity (I²=0) and still reported a significant decline in fasting plasma glucose.
Compared to monotherapy, combination therapy reported a greater change from baseline for postprandial glycaemia (two trials) and index of beta-cell function (three trials), but there were no statistical differences between the two treatment groups for incidence of hypoglycaemia (three trials) and body weight increase (five trials).
There were no statistically significant differences between groups for any adverse events.
The trim-and-fill analysis suggested publication bias may not significantly affect the results.