Twelve RCTs (4,825 participants) were included: 11 assessed effects on HOMA-β (n=3,039) and eight assessed PI/IR (n=2,325). All trials were reported to be randomised and double blinded. No further details of trial validity were provided.
Overall, a statistically significant improvement in HOMA-β was found from sitagliptin compared with placebo (WMD 12.03%, 95% CI 9.45% to 14.60%), which indicated better preservation of β-cell function. A statistically significant reduction in PI/IR was found (-0.06, 95% CI -0.08 to -0.04). Analysis of trials that compared sitagliptin to active control regimens (four RCTs, n=1,425) found better improvement in HOMA-β in the active control group (WMD -5.64%, 95% CI -10.90 to -0.38). The same comparison for PI/IR (three RCTs, n not reported) found no significant difference between sitagliptin and active control regimens.
Subgroup analyses showed that when compared with placebo, a statistically significant improvement in HOMA- β was observed regardless of whether sitagliptin was administered alone or in combination with other drugs, trial duration (12 or 24 weeks) and dosage (50mg twice a day or 200mg/day). When sitagliptin was compared with sulfonylurea (two trials), a statistically significant worsening in HOMA- β was observed (WMD -9.25, 95%CI -16.85, -1.65). Statistically significant decreases in PI/IR were observed compared to placebo for sitagliptin administered alone or in combination, for trials of 24 weeks duration and for dosage of 200mg/day.
Sensitivity analysis (fixed-effects analysis) reportedly did not show a significant difference in results in any primary or subgroup analysis. Statistical heterogeneity was not significant for any primary analysis. Funnel plots were reported to show some evidence of asymmetry. Egger weighted regression models were not statistically significant for either primary analysis.