点击显示 收起
NutriGuard Research, 1051 Hermes Avenue, Encinitas, CA 92024 mccarty{at}pantox.com
Dear Sir:
The meta-analysis of Althuis et al (1) examining the effect of supplemental chromium on insulin, glucose, and glycated hemoglobin in controlled studies may be viewed as cogent evidence that functionally significant dietary chromium deficiency is uncommon in the United States. Most of the cited studies used doses of ≥ 200 µg Cr/d, which is roughly 7 times the average chromium intake of healthy Americans; one would certainly think that such a dose, consumed over several months, would be sufficient to correct a baseline dietary deficiency, and yet the results of most chromium supplementation studies have been paltry, at best.
However, an overview of the data is quite consistent with the possibility that high doses of well-assimiliated organic forms of chromium may indeed have clinical utility in selected populations. In a study by Anderson et al (2) in which Chinese subjects with diabetes were supplemented with ≤ 1000 µg Cr (as chromium picolinate)/d—the largest controlled study yet done with supplemental chromium, but omitted from the meta-analysis because it introduced heterogeneity into the data—glycated hemoglobin fell significantly, by an average of 30%, in those receiving the 1000-µg dose, and the reductions in fasting and postprandial glucose in that group were nearly as large. The fact that the response in the group receiving the 200-µg dose for 4 mo was equivocal, with no significant improvement in either fasting or postprandial glucose relative to the control subjects, strongly suggests that correction of a baseline dietary chromium deficiency was not responsible for the marked improvement in the 1000-µg dose group, and indeed there is no evidence that chromium nutrition is poorer in China than in the United States. (Whether some other reason—eg, genetics, body size, or diet macronutrient profile—might predispose Chinese diabetics to be more responsive to chromium than are their American counterparts remains to be seen.)
Using the same dose and form of chromium in middle-aged overweight US subjects who had first-degree relatives with diabetes, Cefalu et al (3) reported a significant increase of nearly 50% in insulin sensitivity, quantified by the minimal model method, during 4–8 mo of supplementation. Concurrent reductions of 20–25% in fasting and postprandial insulin were not statistically significant, possibly because of the small number of subjects (n = 15). Two other, somewhat shorter, controlled studies of older American subjects failed to observe an effect on insulin metabolism of 1000 µg Cr as chromium picolinate (4, 5).
The available data are thus consistent with the possibility that supranutritional doses of well-assimilated forms of chromium may indeed have useful efficacy in at least some target populations. This view is buttressed by several recent rat studies in which chromium picolinate had substantial metabolic effects (6–8); the control rats in these studies were fed normal (not chromium-depleted) diets, so it cannot be maintained that the observed responses reflected correction of dietary deficiency. It might be added that the chromium doses used in these studies, if corrected for relative body surface area, would be larger than the largest dose used in the Chinese clinical study. A supraphysiologic concentration of chromium picolinate (1 µmol/L) likewise can influence cellular function in vitro (9, 10).
The efficacy of supplemental chromium may prove to be analogous to that of vitamin E. In subjects who are not overtly vitamin E deficient, a nutritional dose of this vitamin will have no discernible effect on the oxidizability of LDL, whereas a megadose (eg, 800 IU) may have a substantial effect. The all-too-common presumption that nutrients can do nothing more than correct deficiency states is clearly wrong. The rodent and tissue culture studies cited above suggest that the physiologic effect of chromium is not always maximized at ordinary tissue concentrations; because we still do not know how chromium functions at the level of molecular biology, there are no firm grounds for assuming otherwise. In light of the fact that oral trivalent chromium has been proven to be safe in animals at any dose tested, future clinical studies with chromium should evaluate a range of daily doses of ≥ 1 mg, while attempting to identify those groups most apt to respond to high-dose chromium.
Note that my remarks pertain to dietary chromium deficiency. It is an open possibility that tissue deficiencies of chromium, attributable to metabolic perturbations that disrupt chromium transport, may be a contributory factor in various disorders. However, there is no reason to assume that ordinary dietary intakes of chromium could correct such tissue deficiencies.
REFERENCES