Information on 20 studies was tabulated, but only 15 RCTs with sufficient data were included in the review (618 people: 193 with type 2 diabetes and 425 who were either in good health or had glucose impairment).
Eleven RCTs analysed data from 90% or more of the patients randomised. Overall, 15% of the randomised patients were lost to follow-up or excluded from analyses. Only 4 RCTs reported baseline chromium levels.
Fasting glucose (14 RCTs; 38 diabetic and 425 nondiabetic patients).
Overall, there was no statistically-significant difference in fasting glucose between chromium supplements and control; the mean difference was 0.027 mmol/L (95% CI: -0.09, 0.15). No significant heterogeneity was found (P=0.97). There was also no statistically-significant difference in fasting glucose between chromium supplements and control among nondiabetics; the mean difference was 0.028 mmol/L (95% CI: -0.086, 0.14). The results for diabetics (4 RCTs) were inconclusive. Three small RCTs found no significant difference between chromium and control. The fourth RCT (180 randomised, 155 analysed) found that 1,000 microg chromium significantly decreased fasting glucose compared with placebo; the mean difference was -1.70 mmol/L (95% CI: -2.41, -0.99). However, it found no significant difference with 200 microg chromium; the mean difference was -1.0 mmol/L (95% CI: -0.93, 0.73). This RCT was the only one conducted in a non-Western country.
Glucose at 120 minutes (5 RCTs; 8 diabetics and 133 nondiabetics).
Overall, there was no statistically-significant difference in glucose at 120 minutes between chromium supplements and control; the mean difference was 0.26 mmol/L (95% CI: -0.24, 0.76). No significant heterogeneity was found (P=0.98). There was also no statistically- significant difference in fasting glucose between chromium supplements and control among nondiabetics,; the mean difference was 0.042 mmol/L (95% CI: -0.43, 0.52). Among diabetics (4 RCTs), one RCT found that 1,000 microg chromium significantly decreased glucose compared with placebo, but it found no significant difference with 200 microg chromium. One small RCT found no significant difference between the treatments.
Fasting insulin (10 RCTs).
Overall, there was no statistically-significant difference in fasting insulin between chromium supplements and control; the mean difference was 0.28 mmol/L (95% CI: -7.0, 7.5). No significant heterogeneity was found (P=0.097). There was also no statistically-significant difference in fasting insulin between chromium supplements and control among nondiabetics; the mean difference was 0.25 mmol/L (95% CI: -6.98, 7.48). The results for diabetics (2 RCTs) were mixed with one study (8 patients) finding no significant difference and the other larger study (155 patients) finding that chromium significantly reduced fasting insulin compared with placebo.
Insulin at 120 minutes (5 RCTs; 8 diabetics and 133 nondiabetics).
There was no statistically-significant difference in insulin between chromium supplements and control, both overall and in nondiabetics. The mean difference overall was 11.1 pmol/L (95% CI: -69.0, 91.2); no significant heterogeneity was found (P=0.15). The mean difference for nondiabetics was 5.5 pmol/L (95% CI: -74.0, 85.1). Among diabetics (2 RCTs), one RCT (non Western population) found that chromium (1,000 and 200 microg) significantly reduced insulin at 120 minutes compared with placebo; the mean differences were -63 picomol/L (95% CI: -79.6, -46.4) and -63 picomol/L (95% CI: -78.3, -47.7) for doses of 1,000 and 200 microg chromium, respectively. A second small RCT found no significant difference.
HbA1c (3 RCTs).
One RCT of 33 healthy patients found no significant difference between chromium and control, as did another RCT of 24 patients with glucose intolerance. One RCT (155 diabetics) found that chromium (1,000 and 200 microg) significantly reduced HbA1c in comparison with placebo with a dose-relationship response; the mean differences were -1.90% (95% CI: -2.34, -1.46) and -1.00% (95% CI: -1.55, -0.45) for doses of 1,000 and 200 microg chromium, respectively.
Chromium formulation, chromium dose and exercise level did not influence the effect of chromium on fasting glucose or insulin.
None of the RCTs reported any adverse events with chromium supplements.