OFFICE OF DIETARY SUPPLEMENTS
CHROMIUM AND DIABETES WORKSHOP SUMMARY
November 4, 1999
Natcher Conference Center, National Institutes of Health
Fundamental to the mission of the Office of Dietary Supplements (ODS) is to promote activities and support clinically relevant basic research, including animal studies, as well as clinical studies of efficacy and safety of dietary supplements. ODSí interest in this topic emanates from a FY 99 congressional mandate to evaluate the role of chromium in diabetes. ODS was specifically urged to examine the effectiveness and efficacy of this treatment modality in collaboration with the United States Department of Agriculture. Appropriately, so is the need for a greater understanding and development of a research agenda for a disease that is the seventh leading cause of death in the United States. Diabetes is one of the most costly health problems in America. A number of the signs and symptoms of diabetes are shared in common with chromium deficiency. These include impaired glucose tolerance, fasting hyperglycemia, glucosuria, hypoglycemia, elevated circulating insulin, decreased insulin receptor number, and peripheral neuropathy. Through this workshop, the ODS hopes to better understand the determinants affecting the bioavailability of chromium, its clinical utility, safety and efficacy, and its potential role in the prevention and/or mitigation of diabetes.
The objectives of the workshop were as follows:
From this information it may be possible to define an appropriate research agenda identifying needed areas of investigation, including the rationale for the study design and implementation of a clinical trial.
The meeting summary presented herein reflects the highlights and salient issues of the topics discussed and/or findings in the current literature. A summary of each speakerís presentation is provided in addition to a concluding summary statement of the committeeís appraisal of the science base. The workshop did achieve a broad consensus on priority research areas that will be most productive in accelerating progress relevant to the use of chromium supplements in individuals at risk for, or diagnosed with, diabetes. A bulleted list of research questions and directives was compiled covering three broad areas of need: basic science, methodology, and clinical studies.
Scope and Impact of Diabetes Marian Parrott, MD, MPH
As spokesperson for the American Diabetes Association, Dr. Parrott provided an overview and update on the magnitude and cost of diabetes in the US today. The cause of diabetes is not completely understood, but genetic, environmental and behavioral factors all play a role. Type 1 diabetes, most commonly presenting in children, is an autoimmune disorder that is more common in persons with certain HLA-types. Type 2 diabetes is related to genetic factors as well as to diet and exercise. Type 2 diabetes is responsible for the enormous increase in diabetes over the past several decades. This is believed to be a result of the increase in obesity and sedentary lifestyles in the US and worldwide. Diabetes is the seventh leading cause of death in the US; 80% of the deaths are due to stroke and heart disease. The more diabetes-specific complications of end stage renal disease, neuropathy and retinopathy are also common. Neuropathy affects about 60-70% of people with diabetes after 10 to 15 years. Retinopathy, affecting 80-97% of patients after 15 years, can be largely prevented by good glycemic control and then early treatment of the disease with laser photocoagulation. Fifteen percent of diabetics will develop a foot ulcer at some point resulting in approximately 60,000 amputations a year. Health care costs and indirect costs have been estimated at $138 billion annually. Certain populations, including African Americans, American Indians, Asian/Pacific Islanders and Hispanics, are more susceptible to diabetes. Also, the prevalence of diabetes increases with age, and the aging of the population has contributed greatly to the rise in diabetes. Dr. Parrott reinforced the need for a well-balanced nutrition program for diabetics and cautioned about prematurely endorsing the use of dietary supplements for the management of diabetes.
What Is the Current State of the Knowledge?
What are the Limitations in Performing Basic Research Studies with Chromium?
Essentials of Chromium Bioavailability Barbara Stoecker, Ph.D.
Dr. Stoecker began her presentation by acknowledging that the exact forms of chromium in the American diet are not known. This has been a long-standing and perplexing question. Schwarz and Mertz in the 1950ís published a series of experiments, which provided evidence that chromium is an essential nutrient forming the active component of what was called the "glucose tolerance factor." Despite this, chromium remains the only essential transition metal whose mechanism of action is not known. Chromium exists in several valence states including metallic chromium (valence zero, Cr-0) which is inert; chromium III (Cr-III) which is stable and the biologically active form; and chromium-VI (Cr-VI) which when bound to oxygen is a strong oxidizing agent that is readily reduced to Cr-III in an acidic environment such as the stomach. Cr-VI is the form associated with industrial exposure and toxicity. Chromium supplements most frequently are available as Cr-III in the chloride or picolinate salt form. Chromium-III occurs in organic complexes with nicotinic acid. For purposes of this review, chromium will refer to chromium-III unless otherwise noted.
Absorption of chromium as chromium chloride is very low (0.4-2.5%) and is affected by dietary factors including ascorbic acid, fiber, and competition with other minerals. In the GI tract, chromium acts very much like iron. Different chromium salts as well as different forms of chromium have diverse solubilities that potentially affect absorption. In addition, the amount of chromium in a dose affects absorption, with a lower percentage being absorbed from higher doses. Once absorbed, chromium circulates largely bound to transferrin in the blood (normal concentration is 380-580 nmol/L). There is little storage of chromium in tissues and most is rapidly excreted in the urine and bile.
The National Research Council estimates that chromium intake in the U.S. is 25 to 33 ug/day, with more than 90% of persons having a daily intake less than 50 ug/day. An intake level of 5 ug/1,000 kcal has been shown to deplete subjects in well-controlled studies.
In animal trials, medications that inhibit prostaglandin synthesis, such as aspirin and indomethacin, markedly enhance uptake of 51Cr. Buffering compounds, such as magnesium or calcium carbonates or hydroxides, and a prostaglandin E2 analog reduce51Cr absorption. It has been shown that chromium competes for one of the binding sites on transferrin. Chromium concentrates in liver, spleen, kidney tissue, and bone. Dr. Stoecker noted that in her laboratory they were able to demonstrate that the dynamics of chromium absorption, distribution and clearance from 51CrCl3 and 51Cr-picolinate are different. It was found that every tissue compartment had higher concentrations of 51Cr-picolinate compared to 51CrCl3, except for bone. Dr. Stoecker noted that toxicity data are very scant. Trivalent chromium chloride is not well absorbed, and in vivo toxicity has not been observed. There have been only a few inconclusive case reports of toxicity attributed to chromium chloride and there have been a few reports of toxicity with chromium picolinate. For clinical trials, identification of subjects who are chromium depleted remains a problem, but the possibility for chromium to potentiate insulin action justifies intensive investigation. Furthermore, the widespread use of chromium supplements, particularly by diabetic patients, warrants extensive monitoring of their safety for long-term use.
Molecular Mechanism of Chromium Action John Vincent, Ph.D.
The demonstration of the molecular basis of the cellular action of chromium by a low molecular weight chromium (LMWCr) complex provides the biochemical rationale for chromium as an essential nutrient. Recently discovered in Dr. Vincentís laboratory is a naturally occurring oligopeptide of 10 amino acids, called LMWCr-binding substance that may function as a part of a novel insulin-signaling autoamplification mechanism by stimulating insulin receptor kinase activity up to several fold. Dr. Vincent notes that the mode of action of LMWCr is reminiscent of that of the calcium-binding signal protein, calmodulin. It appears that LMWCr is maintained in its active apo-oligopeptide form; in response to a chromium flux, LMWCr binds four chromic ions. The holo-oligopeptide is then capable of binding to the insulin receptor thereby amplifying insulin receptor kinase activity. Dr. Vincent suggested that this proposed mechanism also sheds some light on the potential of chromium-containing compounds to serve as nutrition supplements or therapeutics. Chromium picolinate is extremely stable, but concern has been expressed that if chromium picolinate can be reduced within cells (similar to Cr-VI) it could lead to the generation of hydroxyl radicals and potential DNA lipid damage. What is needed to attempt to answer this question of toxic effects are cultured cell studies under appropriate conditions with reasonable amounts of chromium picolinate, as well as long-term animal studies.
What Measures are Currently Available to Assess Chromium Status in Human Populations? Claude Veillon, PhD
At present, there are no good measures for assessing chromium status in humans. Chromium is present in biological tissues and fluids at extremely low levels, so many of the problems associated with finding a measure of status have been analytical in nature. Chromium levels are typically three orders of magnitude lower than zinc levels in biological samples, multiplying sensitivity requirements and contamination problems by the same amount. In most readily accessible biological samples from humans, like blood, serum/plasma, urine, etc., the levels of chromium are less than 1 ng/g (parts-per-billion) and in many cases closer to 0.1 ng/g. Only three analytical techniques have the required sensitivity to make measurements at these levels, namely, neutron activation analysis (NAA), mass spectrometry (MS), and graphite furnace atomic absorption spectrometry (GFAAS). The first two are not widely available, and the third is the one most susceptible to interference from the sample matrix. At the sub-parts-per-billion level, collecting samples without contaminating them, and generating sufficiently low analytical and reagent blanks, becomes extremely important and difficult. Verification of the accuracy of analytical determinations becomes imperative, yet few suitable reference materials are available.
Readily accessible fluids like urine and serum are frequently measured for chromium content, but these seem to reflect recent dietary intake. Whole blood, red blood cells and white blood cells have not been extensively studied. No chromium-containing enzymes have been identified. Hair chromium has been studied, but these measurements have not been widely accepted, partly because they have never addressed the fundamental question of endogenous vs exogenous chromium (or any other trace elements, for that matter). Insulin response does not appear to be a reliable indicator of chromium nutritional status. Balance studies, often used to measure absorption of minerals, are practically impossible for chromium because of the very low fraction absorbed (~ 0.5%) from the gastrointestinal tract.
Dr. Veillon concluded by stating that aside from assessing status, most of the problems associated with chromium in biological systems have been, and are, related to difficulties in accurately measuring chromium in samples. These problems must be solved before status can be assessed.
What Information is Available from Dietary Intake Studies and Clinical Studies regarding Chromium Intake and Supplementation? Richard Anderson, Ph.D.
The daily dietary intake of people in the US and most other countries is roughly 25 ug for women and 33 ug for men, which is sub-optimal based on the minimum estimated Safe and Adequate Daily Dietary Intake (ESADDI) of 50 ug (Anderson and Kozlovsky, 1985). The WHO/FAO/IAEA (1996), reported that the minimum population mean intake likely to meet the normative needs for chromium might be 33 ug/day. However, some data suggest that a chromium intake <20 ug/day is generally inadequate and a significant number of Americans are believed to be consuming < 20 ug/day. The frequency of actual chromium deficiency in the general population is unknown. Acute and subacute syndromes of chromium deficiency have been reported in patients receiving total parenteral nutrition (Jeejeebhoy et al., 1977; Anderson, 1995). The concentration of chromium in foods varies widely and there is considerable variation between batches or lots of the same foods. Therefore, chromium intakes cannot be predicted from databases. Survey data exist for a small number of isolated groups; however, no comprehensive studies exist determining intakes of individual subgroups. Presently, reliable clinical data are accumulating for the following groups: people with glucose intolerance, people with type 2 diabetes mellitus, people with gestational diabetes, and in people with steroid-induced diabetes, but typically little is known about their dietary intakes. In patients with mild hyperglycemia (as defined by 90-minute glucose values >5.56 but <11.1 mmol/L), chromium repletion after experimental dietary depletion improved glucose tolerance and reversed abnormal elevations in circulating insulin and glucagon (Anderson et al., 1991). Recently, Morris et al. (1999) reported the finding that mean levels of plasma chromium were approximately 33% lower and urine values almost 100% higher in 93 NIDDM patients compared to a group of healthy control subjects. These investigators also noted that in the early years of onset of NIDDM, plasma chromium values were inversely correlated with plasma glucose. This correlation however, was lost in patients with diabetes of more than 2 years duration. Ravina et al. (1999) studied a total of 50 patients with steroid-induced diabetes who received benefit from chromium supplementation that could not be controlled by insulin and or other drugs. Continual supplementation with chromium picolinate (200 ug, three times a day) also reduced the dose of required medications. Dr. Anderson confirms that data from these supplementation studies provide some evidence for the short-term and long-term efficacy of chromium, as evidenced by reductions in fasting and 2-hour glucose and insulin values, and long-term reductions in hemoglobin A1c concentrations utilizing varying doses of chromium (200, 500 or 1000 ug). Regarding toxicity, the safety factor for chromium is greater than that for almost any other nutrient. In 19 randomized controlled trials in which individuals received between 175 and 1000 ug/day chromium for duration of between 6 and 64 weeks, there was no evidence of any toxic effects (Jeejeebhoy, 1999). Dr. Anderson concluded by stating that large-scale studies are urgently needed to determine if chromium can prevent or delay the onset of these diseases.
Do chromium requirements change with disease state and pathology? Review of clinical data in patients (subjects) without diabetes mellitus? William Cefalu, MD
In an effort to examine chromium needs, chromium levels have been evaluated over a range of age and disease states, and many supplementation studies have been completed. Compelling evidence that chromium levels and metabolism may be altered with age comes from a recent study by Davies et al. (1997). Davies evaluated either hair (n=22,013), sweat (n=17,937) or serum (n=11,715) chromium levels in over 40,000 subjects from early childhood through the late seventies. He noted about a 40% drop in serum and hair chromium levels over this age range, as well as a significant drop in sweat chromium levels. Dr. Cefalu noted that the data from animal studies suggest that aging may be related to a decrease in cellular content and transport of chromium, reduced tissue retention and/or perhaps urinary retention (Wallach and Verch, 1986). Furthermore, some clinical studies suggest that daily chromium supplementation (400-1000 ug/day) may favorably affect insulin and glucose levels ñ not only in diabetics, but also in obese, insulin-resistant subjects (Cefalu et al, 1999).
Clinical studies conducted by Dr. Cefalu in 29 obese individuals with a high risk of developing type 2 diabetes were given either a placebo or chromium picolinate (1,000 ug/day) for eight months to evaluate glucose effectiveness and insulin sensitivity. Effectiveness and sensitivity were measured using the frequently sampled intravenous glucose tolerance test (FSIVGTT, modified minimal model), the administration of an oral glucose tolerance test (75 g), and the collection of 24-hour insulin and glucose profiles. Improvements were seen in the patients receiving chromium therapy vs. those receiving placebo without any significant changes in body fat distribution after 4 months and 8 months, suggesting that chromium picolinate can alter insulin sensitivity independent of a change in weight or body fat percentage. This may occur by directly affecting muscle insulin action (Cefalu et al., 1999).
The overriding question left to the committee was what is the best way to derive further information to determine whether a deficient chromium status plays a role in the incidence of diabetes in this country. Most perplexing is the fact that we do not yet have a good measure of chromium status, we cannot measure or assess deficiency in people, and we have not developed satisfactory animal models that can be extrapolated to humans. Hence, it becomes evident that the development of appropriate and sensitive biomarkers and outcomes measures are pivotal prior to initiation of formal clinical intervention studies. It is clear that, for the general public, current data do not warrant routine use of chromium supplements, whose risk-benefit function has not yet been adequately characterized. The consensus of the committee members was that small-scale, focused clinical studies might be warranted in certain population groups, such as those with glucose intolerance or newly diagnosed diabetes, since as convincing data are emerging in these groups. Efforts should be directed at developing and/or monitoring functional status indicators, such as insulin resistance or response to an oral glucose load. These studies should be designed and statistically powered to determine clinically relevant indicators. Multiple clinical indicators should be monitored in glucose-intolerant populations, enrolling both men and women from diverse ethnic and geographical origins. It is recommended that a systematic review of the literature be conducted and be made available to an advisory panel comprised of clinical experts in the fields of nutrition, endocrinology, diabetes, epidemiology, and statistics, in order to identify the key factors and components necessary for the conduct of a feasibility study. Key also to advancing research in this field are studies that further elucidate the mechanism of action of chromium-bound substances, to include an evaluation of their regulatory role and genetic characteristics. These studies should proceed with the same vigor as the clinical studies.
Priority research areas to be addressed include:
The following is a list of research questions and directives in need of answers, so that we may undertake a thorough evaluation of a therapeutic role for chromium in diabetes.
II. Research questions and directives that support issues in the development and refinement methodologies for assessing chromium status.
III. Research questions and directives derived from, and in support of, dietary and clinical studies utilizing chromium supplements in patients with and without diabetes.
Abraham, A.S., Sonneblick M, and Eini, M. (1981) Serum chromium and aging. Gerontology 27:326-328.
Anderson, R.A. (1989) Essentiality of chromium in humans. Sci. Total Environ. 86:75-81.
Anderson, R.A. (1995) Chromium and parenteral nutrition. Nutrition. 11:83-86.
Anderson, R.A. (1998) Chromium, glucose intolerance and diabetes. J. Am. Coll. Nutr. 17:548-555.
Anderson, R.A., Bryden, N.A., and Polansky, M.M. (1992) Dietary chromium intakeófreely chosen diets, institutional diets and individual foods. Biol. Trace Elem. Res. 32:117-121.
Anderson, R.A., Cheng, N., Bryden, N.A., et al. (1997) Beneficial effects of chromium for people with diabetes. Diabetes 46:1786-1791.
Anderson, R.A., and Koslovsky, A.S. (1985) Chromium intake, absorption and excretion of subjects consuming self-selected diets. Am. J. Clin. Nutr. 41:1177-1183.
Anderson, R.A., Polansky, M., Bryden, N., and Canary, J. (1991) Supplemental chromium effects on glucose, insulin, glucagon, and urinary chromium losses in subjects consuming controlled low-chromium diets. Am. J. Clin. Nutr. 54:909-916.
Cefalu, W., Bell-Farrow, A.D., Wang, Z.Q., et al. (1999) Effect of chromium picolinate on insulin sensitivity in vivo. J. Trace Elem. Exp. Med. 12:71-83.
Davies S, McLaren-Howard J, Hunniset A, and Howard, M. (1997) Age-related decreases in chromium levels in 51,665 hair, sweat, and serum samples from 40,872 patientsóimplications for the prevention of cardiovascular disease and type II diabetes mellitus. Metabolism 46:469-473.
Davis, C.M., and Vincent, J.B. (1997) Chromium oligopeptide activates insulin receptor tyrosine kinase activity. Biochem. 15:4382-4385.
Ding, W., Chai, Z., Duan, P., Feng, W., and Qian, Q. (1998) Serum and urine chromium concentrations in elderly diabetes. Biol. Trace Elem. Res. 63:231-237.
Hellerstein, M.K., (1998) Is chromium supplementation effective in managing type II diabetes? Nutr. Rev. 56:302-306.
Huang, B., Lin, S.Q., Chen, S.Y., Zhou, G., Yin, F., Lou, Z.P., and Bi, M.M. (1991) Hair chromium levels in patients with vascular disease. Biol. Trace Elem. Res. 29:133-137.
Jeejeebhoy, K.N. (1999) The role of chromium in nutrition and therapeutics and as a potential toxin. Nutr. Rev. 57:329-335.
Jeejeebhoy, K.N., Chu, R.C., Marliss, E.B., Greenberg, G.R., Bruce-Robertson, A. (1977) Chromium deficiency, glucose intolerance, and neuropathy reversed by chromium supplementation in a patient receiving long-term total parenteral nutrition. Am. J.Clin. Nutr. 30:531-538.
Jovanovic-Peterson L, Gutierrez M., Peterson, C. (1996) Chromium supplementation for gestational diabetes women improves glucose tolerance and decreases hyperinsulinemia. Diabetes Care 45 (suppl. 2) 237A.
Keim, K.S., Stoecker, B.J., and Henley, S. (1987) Chromium status of the rat as affected by phytate. Nutr. Res. 7:253-263.
Kozlovsky, K.S., Moser, P.B., Reiser, S., and Anderson, R.A. (1986) Effects of diets high in simple sugars on urinary chromium losses. Metabolism. 35:515-518.
Lukaski, H.C. (1999) Chromium as a Supplement. Annu. Rev. Nutr. 19:279-302.
Mertz, W. (1998) Interaction of chromium with insulin. A progress report. Nutr. Rev. 56:174-177.
Mertz, W. and Roginski, E.E. (1969) Effects of chromium (III) supplementation of growth and survival under stress in rats fed low protein diets. J. Nutr. 97:531-536.
Morris, B.W., Blumsohn, A., MacNeil, S., and Gray, T.A. (1992) The trace element chromium ñ a role in glucose homeostasis. Am. J. Clin. Nutr. 55:989-991.
Morris, B.W., Gray, T.A., and MacNeil, S. (1993) Glucose-dependent uptake of chromium in human and rat insulin-sensitive tissues. Clin. Sci. 84:477-482.
Morris, B.W., MacNeil, S., Hardesty, C.A., Heller, S., Burgin, C., Gray, T.A. (1999) chromium homeostasis in patients with Type II (NIDDM) diabetes. J. Trace. Elem. Med. Biol. 13:57-61.
Morris, B.W., MacNeil, S., Stanley, K., Gray, T.A., and Fraser, R. (1993) The inter-relationship between insulin and chromium in hyperinsulinaemic euglycaemic clamps in healthy volunteers. J. Endocrinol. 139:339-345.
National Research Council. Recommended Dietary Allowances. 10th ed. Washington, D.C., National Academy Press, 1989.
Offenbacher, E.G., Pi-Sunyer, F.X. (1980) Beneficial effects of chromium-rich yeast on glucose tolerance and blood lipids in elderly subjects. Diabetes 29:919-925.
Ravina, A., Slezak, L., Mirsky, N., Bryden, N.A., and Anderson, R.A. (1999) Reversal of corticosteroid-induced diabetes mellitus with supplemental chromium. Diabetic Med. 16:164-167.
Ravina, A., Slezak, L., Mirsky, N., and Anderson, R.A. (1999) Control of steroid-induced diabetes with supplemental chromium. J. Trace Elem. Exptl. Med. 12:375-378, 1999.
Schwarz, K., Mertz, W. Chromium (III) and the glucose tolerance factor. Arch. Biochem. Biophys. 85:292-295.
Seaborn, C.D., and Stoecker, B.J. (1990) Effects of antacid or ascorbic acid on tissue accumulation and urinary excretion of 51chromium. Nutr. Res. 10:1401-1407.
Seaborn, C.D., and Stoecker, B.J. (1992) Effects of ascorbic acid depletion and chromium status on retention and urinary excretion of 51Cr. Nutr. Res. 12:1229-1234.
Simonoff, M., Llabador, Y., Simonoff, G.N., Besse, P., and Conri, C. (1984) Cineangiographically determined coronary artery disease and plasma chromium levels for 150 subjects. Nucl. Instr. Meth. 231.
Simonoff, M., Llabador Y., Hamon, C., Mackenzie, Peers A., and Simonoff, G.N. (1984) Low plasma chromium in patients with coronary artery disease and heart disease. Biol. Trace Elem. Res. 6:431-439.
Veillon, C. and Patterson, K.Y. (1999) Analytical issues in nutritional chromium research. J. Trace Elem. Exp. Med. 12:99-109.
Vincent, J.B. (1999) Mechanisms of chromium action: low-molecular-weight chromium-binding substance. J. Am. Coll. Nutr. 18:6-12.
Wada, O., Wu, G.Y., Yamamoto, A., et al. (1983) Purification and chromium-excretory function of low molecular weight, chromium binding substances from dog liver. Environ. Res. 2:228-239.
Wallach, S. and Verch, R.L. (1986) Radiochromium distribution in aged rats. J. Am. Coll. Nutr. 5:291-298.
WHO/FAO/IAEA. 1996. Trace elements in human nutrition and health. Geneva: World Health Organization; p. 155-160.