Figure 7.
Funnel plots of publication bias. Begg and Egger tests provided no evidence of significant publication bias in most outcome assessments except gastric volume (Egger test, P = 0.000; Begg test, P = N.S.), glucose (Egger test, P = 0.001; Begg test, P = 0.004) and homeostatic model assessment–insulin resistance (Egger test, P = 0.035; Begg test, P = N.S.).
