Are Chromium (Cr) Supplements Efficacious for Glycemic Control in Type 2 Diabetes (T2DM)?

Abstract

Chromium (Cr)‐containing dietary supplements are claimed to improve glycemic control in T2DM. We conducted a systematic literature search of randomized clinical trials (RCT) of adults with T2DM to document changes in glycemic control from pre‐ to post‐supplementation with Cr. The search included studies from 1/1991–1/2014 with the following terms: blood glucose, blood sugar, glucose metabolism disorders, or metabolic syndrome; randomized clinical trial, systematic review or meta‐analysis; trials with adults (>18yr.); T2DM defined by self‐report, clinical diagnosis or biochemical determination of elevated fasting plasma glucose [FPG ] or hemoglobin A1c [HbA1c];with and without concurrent treatment with hypoglycemic agents; and RCT design having a placebo or comparator control along with the Cr dietary supplement. Exclusions were studies in children, or solely in patients with T1DM; presence of unstable chronic disease and/or acute conditions; studies of combination therapies without a separate Cr supplement; observational, non‐randomized or un‐blinded studies; and those not reporting results pre vs. post‐supplementation. We defined elevated FPG as >126 mg/dL, or elevatedHbA1c as > 6.5%, and treatment goals as FPG ≤ 130 mg/dL, or ≤7.0% HbA1c, and a clinically substantial decrease as ≥0.5% HbA1c (or 60 mg/dL, a criterion based on reduction achieved by successful drug therapy; no similar criterion was available for a single FPG). A total of 20 RCT (22 arms) were identified. Many different doses and forms of Cr were used. 16 of 20 RCT continued hypoglycemic medications and lifestyle modifications during Cr supplementation, which confounded the interpretation of the cause of changes during supplementation. Many studies failed to show decreases compared to controls and in others decreases were small. Glycemic treatment goals were reached after Cr supplementation in 5/20studies using the mean FPG criterion, 3/14 using HgA1c, and 1/14 using both FPG and HgA1c. Mean HbA1c declined by ≥0.5% in 5/14 RCT, but in 4/5 of those with a drop concurrent hypoglycemic medications were also administered. Neither dose nor duration of Cr supplementation appeared to be associated with effectiveness. Also, we discovered that many of the RCT had features that threatened their quality, including heterogeneous study populations, small study size, high drop‐out rates, and short study duration. Moreover, in many studies the advantages of randomization for causal inference were lost due to high dropout rates and analyses solely of completers (per protocol analysis) rather than by intent to treat. The low strength of the evidence to date makes it unlikely that Cr supplements are efficacious for glycemic control. Future meta‐analyses should be selective and use similar inclusion criteria, similar forms of Cr, and include only high quality studies to provide scientifically sound recommendations.Support or Funding InformationFunding: Office of Dietary Supplements, National Institutes of Health.