Está en la página 1de 11

Reviews/Commentaries/ADA Statements

R E V I E W A R T I C L E

Role of Chromium in Human Health and


in Diabetes
WILLIAM T. CEFALU, MD1 controversial and serves as the basis for
FRANK B. HU, MD, PHD2 this review.
Trivalent chromium is found in a
wide range of foods, including egg yolks,
whole-grain products, high-bran break-
fast cereals, coffee, nuts, green beans,

D
espite widespread use by patients contained a glucose tolerance factor
with diabetes and anecdotal reports (GTF) that prevented diabetes in experi- broccoli, meat, brewers yeast, and some
in the past regarding its efficacy, mental animals (1). This factor was even- brands of wine and beer (8,9). Chromium
until recently, data in humans concerning tually suggested to be a biologically active is also present in many multivitamin/
chromiums effects on insulin action in form of trivalent chromium that could mineral supplements, and there are also
vivo or on cellular aspects of insulin ac- substantially lower plasma glucose levels specific chromium picolinate (CrP) sup-
tion were scarce. Consequently, signifi- in diabetic mice (2). Interest regarding plements that contain 200 600 g chro-
cant controversy still exists regarding the chromium administration in patients mium per tablet (10). The U.S. National
effect of chromium supplementation on with diabetes was kindled by the observa- Academy of Sciences has established the
parameters assessing human health. Fur- tion in the 1970s that it truly was an es- Recommended Daily Allowances for
thermore, elucidating the cellular and sential nutrient required for normal chromium as 50 200 g/day for adult
molecular mechanisms by which chro- carbohydrate metabolism. A patient re- men and women (11), which is also the
mium supplements affect carbohydrate ceiving total parenteral nutrition (TPN) Estimated Safe and Adequate Daily Di-
metabolism in vivo is necessary before developed severe signs of diabetes, in- etary Intake (ESADDI) for chromium for
specific recommendations can be made cluding weight loss and hyperglycemia children aged 7 years to adulthood (7,12).
regarding its routine use in the manage- that was refractory to increasing insulin However, it appears that Americans nor-
ment of diabetes. This review focuses on dosing (3). Based on previous animal mally ingest 50 60% of the minimum
providing current information about this studies and preliminary human studies, suggested daily intake of 50 g (7). Re-
trace minerals specific mechanisms of ac- the patient was given supplemental chro- sults from one study (10) indicated that
tion and clinical trials in patients with mium. In the following 2 weeks, signs and daily chromium intakes for men and
diabetes. symptoms of diabetes were ameliorated, women in the U.S. were 33 and 25 g,
Chromium, one of the most common with markedly improved glycemic status respectively. Therefore, normal dietary
elements in the earths crust and seawater, and greatly reduced insulin requirements intake of chromium for adults may be
exists in our environment in several oxi- (exogenous insulin requirements de- suboptimal.
dation states, principally as metallic (Cr0), creased from 45 units/day to none). Other At dietary intakes 50 g/day, chro-
trivalent (3), and hexavalent (6) chro- studies (4,5) of the beneficial effects of mium absorption is 0.4%, but the triva-
mium. The latter is largely synthesized by chromium in patients receiving TPN have lent formulation also significantly
the oxidation of the more common and also been documented in the scientific lit- influences bioavailability. At a dose of
naturally occurring trivalent chromium erature. Chromium is now routinely 1,000 g/day, absorption of chromium
and is highly toxic. Trivalent chromium, added to TPN solutions (5). from chromium chloride (CrCl 3 ) is
found in most foods and nutrient supple- The results of these studies strongly 0.4%, whereas that from CrP may be as
ments, is an essential nutrient with very implicated chromium as a critical cofactor high as 2.8% (7,13,14). Once absorbed,
low toxicity. in the action of insulin (6,7). Whereas chromium is distributed widely in the
The interest in chromium as a nutri- chromium replacement in deficiency body, with the highest levels being found
tional enhancement to glucose metabo- states is well established, the role of chro- in the kidney, liver, spleen, and bone
lism can be traced back to the 1950s, mium supplementation to enhance glu- (14).
when it was suggested that brewers yeast cose metabolism in subjects is
BIOLOGIC ACTIONS OF
From the 1Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana; and CHROMIUM H o w c h r o m i u m
2
Harvard University School of Public Health, Boston, Massachusetts. serves as a cofactor for insulin action is
Address correspondence and reprint requests to William T. Cefalu, MD, Division of Nutrition and not fully understood. From several in vivo
Chronic Disease, The Pennington Biomedical Research Center, Louisiana State University, 6400 Perkins Rd., and in vitro studies (15), it was initially
Baton Rouge, LA 70808. E-mail: cefaluwt@pbrc.edu.
Received for publication 15 July 2004 and accepted in revised form 21 July 2004. thought that chromium potentiated the
F.B.H. has received grant support from Nutrition 21. actions of insulin as part of an organic
Abbreviations: CrP, chromium picolinate; CVD, cardiovascular disease; FPG, fasting plasma glucose; complex, GTF. More recent studies (15)
GTF, glucose tolerance factor; HPFS, Health Professionals Follow-up Study; LMWCr, lowmolecular weight have suggested that chromium may func-
chromium; MI, myocardial infarction; MW, molecular weight; TPN, total parenteral nutrition.
A table elsewhere in this issue shows conventional and Systeme International (SI) units and conversion tion as part of the oligopeptide low
factors for many substances. molecular weight (MW) chromium
2004 by the American Diabetes Association. (LMWCr)-binding substance (MW

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2741


Chromium in diabetes and cardiovascular risk

The controversy surrounding chro-


mium supplementation is due in part to
substantial variability in the results of
studies that have evaluated the effects of
chromium in patients with or without di-
abetes. Results from some trials (2126)
have indicated that chromium supple-
mentation increases muscle gain and fat
loss associated with exercise and im-
proves glucose metabolism and the serum
lipid profile in patients with or without
diabetes. In contrast, those from other
studies (2732) have indicated little or no
benefit of chromium on any of these
variables.
Recent meta-analyses (33,34) of re-
sults from studies that evaluated the ef-
fects of chromium supplementation have
suggested limited benefit in individuals
with or without diabetes. The major con-
clusions from these analyses were that
chromium has a very small effect versus
Figure 1Proposed mechanism of action for chromium and LMWCr potentiating the action of placebo in reducing body weight and that
insulin (15). the clinical relevance of this small de-
crease is debatable and should be inter-
preted with caution. It was also
1,500 Da), which is composed of gly- exists in insulin-dependent cells in the concluded that chromium has no effect
cine, cysteine, glutamic acid, and aspartic apo, or inactive, form. Binding with chro- on glucose metabolism or insulin concen-
acid. The interaction of chromium with mium ions converts inactive LMWCr to trations in individuals without diabetes
LMWCr and the manner in which this its holo, or active, form. It is proposed and that data for patients with diabetes
complex influences insulin metabolism is that LMWCr then participates as part of are currently inconclusive. It is important
considered in greater detail below. an insulin signal amplification system to note that these conclusions are based
(Fig. 1) as it binds to insulin-activated in- largely on data from patients without di-
Biochemistry sulin receptors and results in stimulating abetes and failed to include key positive
Very little chromium (2%) in the form its tyrosine kinase activity. The result of results for chromium supplementation
of inorganic compounds is absorbed but this process is the activation of insulin re- in diabetic patients and subjects with ges-
may be higher with certain organic for- ceptor kinase and potentiation of the ac- tational diabetes or the metabolic
mulations (14). Once absorbed, chro- tions of insulin (15,18,19). Importantly, syndrome.
mium is distributed to various tissues of LMWCr without bound chromium or in There is no clinically defined state of
the body, but appears to be most concen- the presence of other metal ions is ineffec- chromium deficiency, but diabetes has
trated in the kidney, muscle, and liver tive in activating insulin-dependent ki- been shown (32) to develop because of
(16). The principal carrier protein for nase activity and thus enhancing the low chromium levels in experimental an-
chromium is transferrin, which also plays actions of insulin (19). imals and in humans sustained by pro-
a critical role in the movement of chro- Chromium has also been demon- longed TPN. These results suggest that
mium from blood to LMWCr. It has been strated to inhibit phosphotyrosine phos- there may be a more general relationship
suggested that migration of transferrin re- phatase, the enzyme that cleaves between chromium levels and glucose
ceptors to the plasma membranes of insu- phosphate from the insulin receptor, and/or lipid metabolism. It has also been
lin-insensitive cells after insulin leading to decreases in insulin sensitivity. suggested (3537) that low chromium
stimulation is the initial step in this pro- Activation of insulin receptor kinase and concentrations and the associated impair-
cess. Transferrin containing the plasma- inhibition of insulin receptor phospha- ments in insulin, glucose, and lipid me-
bound chromium is postulated to bind to tase would lead to increased phosphory- tabolism may also result in increased
the transferrin receptors and is internal- lation of the insulin receptor and cardiovascular risk. In a cross-sectional
ized by endocytosis (Figs. 1 and 2). The increased insulin sensitivity (20). The bal- analysis (38), lower toenail chromium
pH of the internalized vesicle is reduced ance between kinase and phosphatase ac- levels have also been associated with in-
by ATP-driven proton pumps, chromium tivity may facilitate the role of insulin in creased risk of type 2 diabetes. Adequate
is released from transferrin, and the re- rapidly moving glucose into cells. In ad- dietary chromium intake may be espe-
sulting free chromium is postulated to be dition, it has been suggested (7) that chro- cially problematic in the elderly (39,40).
sequestered by LMWCr (15,17). With mium enhances insulin binding, insulin Consumption of refined foods, including
this step, chromium is transferred from receptor number, insulin internalization, simple sugars, exacerbates the problem of
transferrin to LMWCr, which normally and -cell sensitivity. insufficient dietary chromium because

2742 DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004


Cefalu and Hu

type 2 diabetic subjects compared with


control subjects. Another study reported
that chromium levels were reduced
50% in both diabetic men and women
compared with control subjects (42),
which was supported by Elmekcioglu et
al. (47), who reported significantly lower
chromium levels in the plasma of type 2
diabetic individuals compared with non-
diabetic healthy control subjects. Yet, an-
other study (48) suggested no alteration
of chromium levels in type 2 diabetes;
however, only 11 subjects were reported.

CLINICAL INTERVENTION
WITH CHROMIUM T h e m o s t
recent recommendations of the American
Diabetes Association state that at the
present, benefit from chromium supple-
mentation in persons with diabetes has
not been conclusively demonstrated
(49).

Review of the literature


A review of the literature has revealed nu-
merous conflicting studies evaluating
chromium supplementation and parame-
ters assessing carbohydrate metabolism
(21,30 32,50 60) (Table 1). Consider-
able differences in efficacy were reported
that essentially confused health care pro-
viders regarding routine use of chromium
in diabetic states. These earlier studies
(12) are difficult to interpret, as many
were open label and therefore generated
substantial bias. Additional concerns are
the lack of gold standard techniques to
assess glucose metabolism, use of differ-
ing doses and formulations, and hetero-
geneous study populations. Specifically,
the limitations of the earlier studies can be
classified as follows.
Study design. The use of a control group
Figure 2Proposed mechanism for the movement of chromium from the blood to LMWCr (18).
is of paramount importance when evalu-
Cr, chromic ion; M, metal ion; Tf, transferrin; Tf-R, Tf receptor.
ating the effect of chromium given the
possibility that patients who choose to use
these foods are not only low in dietary tion of chromium. One study reported chromium may be different from nonus-
chromium but also increase its loss from that in 40,800 patients from ages 1 to ers. Thus, only a randomized intervention
the body (41). Chromium losses are also 75 years, chromium levels in hair, can definitely establish the overall effects
increased during pregnancy and as a re- sweat, and blood diminished significantly of chromium on insulin action, as it is this
sult of strenuous exercise, infection, with age, with values decreasing from 25 design that controls for biases, whether
physical trauma, and other forms of stress to 40% depending on the tissue of interest known or unknown, that may confound
(40). Reduced chromium levels are re- (43). Additionally, it appears that diabetic the association and assessment of chro-
ported in the elderly and in patients with subjects may have altered chromium me- mium supplementation and carbohydrate
diabetes (42,43). However, one of the ma- tabolism compared with nondiabetic sub- metabolism. Unfortunately, many of the
jor problems with assessing chromium jects, as both absorption and excretion reported studies evaluating chromium
status in biological tissues and fluids is may be higher (44,45). Hair and blood supplementation were open-label studies
extremely low levels of chromium in these levels are reported (46) to be lower in di- (Table 1).
tissues. Regardless, recent studies have abetic subjects, with mean levels of Subject selection. The clinical charac-
demonstrated the successful determina- plasma chromium of 33% lower in 93 teristics of the study subjects varied tre-

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2743


2744
Table 1Comparison of chromium studies in humans

Results
Study Chromium supplement Technique
Reference type Study length (dose) Subjects n assessed Glucose Insulin HbA1c IS
Studies using CrCl, brewers yeast,
and CrN formulation
Trow et al. (50) OL 2 months Brewers yeast (100 g/day Type 2 12 OGTT NA NA
Cr3)
Sherman et al. (51) DB 4 months CrCl3 (150 g/day) Type 1/2/Nondiabetic 14 OGTT NA NA NA
Rabinowitz et al. (32) DB 4 months CrCl3 (150 g/day), brewers Type 2 43 Meal challenge NA NA
yeast (13 g/day Cr3)
Uusitupa et al. (30) DB 6 months Brewers yeast (160 g/day IGT, elderly 26 OGTT NA
Chromium in diabetes and cardiovascular risk

Cr3)
Uusitupa et al. (52) DB 6 weeks CrCl3 (200 g/day) Type 2 10 OGTT, HbA1c 2 NA
Potter et al. (21) OL 3 months CrCl3 (200 g/day) IGT 5 Hyperglycemic 1
clamp
Mossop (53) DB 3 months CrCl3 (600 g/day) Type 1/2 26 FBG 2 NA NA NA
Nath et al. (54) OL 2 months CrCl3 (500 g/day) Type 2 12 OGTT 2 2 NA NA
Glinsmann and Mertz (55) OL 18133 days CrCl3 (1803,000 g/day) Type 1/2 6 IVGTT, OGTT 2 NA NA NA
Wilson and Gondy (31) DB 3 months CrN (220 g/day) Nondiabetic, young 26 FBG/insulin 2* NA NA
Studies using CrP formulation
Anderson et al. (45) R 4 months CrP (200 or 1000 g/day) Type 2 180 OGTT, HbA1c 2 2 2 NA
Amato, Morales, and Yes (56) DB 2 months CrP (1,000 g/day) Nondiabetic, elderly 19 Minimal model NA
Jovanovic, Gutierrez, and DB 2 months CrP (320 or 640 g/day)* Gestational diabetes 20 OGTT, HbA1c 2 2 NA
Peterson (57)
Ravina et al. (58) OL 17 days CrP (600 g/day) Diabetes 3 FBG 2 NA NA NA
Morris et al. (59) OL 3 months CrP (400 g/day) Type 2 5 Insulin tolerance, 2 NA 1
HOMA
Cheng et al. (60) OL 110 CrP (500 g/day) Type 2 833 Fasting, postmeal 2 NA NA NA
months
Ghosh et al. (22) DB 3 months Cr3 (200 g/day) Type 2 50 FBG, HbA1c 2 NA 2 NA
Cefalu et al. (64) R 8 months CrP (1,000 g/day) Pre-diabetes 29 Minimal model 2 NA 1
Ravina et al. (68) OL 10 days CrP (200 g/day) Type 1/2 48/114 Insulin tolerance, NA NA 2 1
HbA1c
Lee and Reasner (70) DB, 2 months CrP (200 g/day) Type 2 30 FBG, HbA1c NA NA
Evans (78) DB 42 days CrP (200 g/day) Type 2 11 FBG, HbA1c 2 NA 2 NA
1, increased; 2, decreased; , no change; CrCl, chromium chloride; CrN, chromium nicotinate; DB, double blind; FBG, fasting blood glucose; IGT, impaired glucose tolerant; IS, insulin sensitivity; IVGTT,
intravenous glucose tolerance test; NA, not assessed; OGTT, oral glucose tolerance test; OL, open label; R, randomized. *In hyperinsulinemic patients only; -cell sensitivity to glucose.

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004


Cefalu and Hu

mendously as several studies grouped Assessment of chromium status. Many significant decreases in mean FPG, 2-h
type 1 and type 2 diabetic subjects to- of the earlier reported studies did not ad- glucose, and fructosamine. Chromium
gether in the evaluation of chromiums ef- dress the role of chromium blood levels at treatment also slightly reduced required
fect (Table 1). Indeed, even in studies in baseline or recorded changes, if any, with doses of antidiabetic drugs, and this de-
which only subjects with type 2 diabetes supplementation. In addition, objective cline achieved statistical significance for
were reported, subjects were assessed markers to measure compliance with the glibenclamide. Another group assessed
while on various therapies (e.g., diet, sul- regimen were not evaluated. the effects of jiangtangkang (8 g t.i.d.), a
fonylureas, metformin, insulin) and at Techniques to assess response. The chrysanthemum product high in chro-
different levels of glycemic control major limitation of the earlier studies, mium, on glucose and insulin metabolism
(32,50,55,58,60). It is well established however, may be the lack of sophisticated in 188 patients with type 2 diabetes (67).
that hyperglycemia secondary to glucose metabolic techniques used to assess car- After 2 months, jiangtangkang treatment
toxicity may contribute to attenuation in bohydrate metabolism. Many of the stud- reduced fasting and postprandial blood
insulin action (61), and the effect of med- ies evaluated response by fasting blood glucose and HbA1c without any corre-
ications to alter insulin action is well stud- levels only or used glucose tolerance or sponding change in plasma insulin. A 16-
ied (62,63). mixed-meal tests (Table 1). Although month, double-blind, randomized,
Dosage, formulation, duration of these tests are frequently used in clinical crossover trial (32) of chromium chlo-
study. The duration of supplementation studies, they do not provide the sensitiv- ride, brewers yeast that contained chro-
evaluated (ranging from 1 day to 8 ity required to precisely assess insulin ac- mium as GTF, brewers yeast extract
months) and the dose used (ranging from tion. Our literature search did not find without GTF, and a placebo in 43 patients
100 to 3,000 g daily) varied tremen- any study that evaluated the effect of with diabetes also demonstrated positive
dously in earlier studies. Studies that spe- chromium supplementation on insulin effects of chromium on glucose and insu-
cifically evaluated 200 g of chromium sensitivity by using the gold standard for lin metabolism. FPG and the glucose re-
chloride failed to elicit a clinical response assessing insulin action, i.e., the hyperin- sponse to either a standard meal or
in those with type 2 diabetes (Table 1). sulinemic-euglycemic clamp. One study tolbutamide were not significantly altered
Uusitupa et al. (52) demonstrated a posi- (65) used the euglycemic clamp, but only by any of the treatments, but ketosis-
tive effect at 200 g of the CrCl salt; how- to assess the relationship of blood chro- resistant patients experienced a signifi-
mium and insulin and did not supple- cant increase in postprandial insulin after
ever, the remaining variables in that study
ment the subjects. Another study (21) treatment with the brewers yeast that
did not appear to be altered by supple-
evaluated subjects with a hyperglycemic contained GTF. Results from an addi-
mental chromium. A more consistent
clamp and demonstrated that significant tional study indicated that chromium
clinical response is observed with daily
increases in glucose utilization were ob- supplementation has significant positive
supplementation of chromium 200 g/
served and associated with increases in effects on glucose and insulin metabolism
day for a duration of 2 months (Table
-cell sensitivity to glucose following in patients with diabetes. One study (68)
1). In addition, other forms of chromium, chromium supplementation. Of the other reported that 10 days of treatment with
especially CrP, appear to be more bio- three studies that evaluated a more so- CrP (200 g/day) significantly increased
available and clinically more effective phisticated technique to assess insulin ac- insulin sensitivity in patients with type 1
than chromium chloride in both human tion, one did not show an improvement in or 2 diabetes and also permitted reduc-
and animal studies. Evidence for a dose insulin sensitivity (56) using the minimal tions in dosages of insulin and/or oral an-
effect of CrP was provided by a study of model technique, in contrast to Cefalu et tidiabetic drugs in these patients.
Chinese type 2 diabetic subjects (45). al. (64). A separate study (59) demon- A large long-term study showed that
Short-term (2 months) and long-term (4 strated beneficial effects in type 2 diabet- 10 months of treatment with CrP (500
months) efficacy were observed, as evi- ics using the insulin tolerance test and g/day) in 833 patients with type 2 dia-
denced by reductions in fasting and 2-h homeostasis model assessment method. betes significantly improved both FPG
glucose and insulin values and long-term and postprandial plasma glucose versus
reductions in HbA1c concentrations uti- Individuals with diabetes baseline (Fig. 3) and reduced the inci-
lizing varying doses of CrP (200 or 1,000 Type 1 and 2 diabetes. Chinese patients dence of diabetes symptoms, including
g). The effectiveness of the 1,000-g with type 2 diabetes receiving CrP expe- fatigue, thirst, and frequent urination
dose in the Chinese study was reproduced rienced significant improvements in (60).
in a study of individuals with the meta- HbA1c, fasting plasma glucose (FPG), 2-h Not all studies have demonstrated
bolic syndrome (64). In a study (57) of 30 glucose (i.e., glucose levels 2 h after chal- significant positive effects of chromium
women with gestational diabetes receiv- lenge), and fasting and 2-h insulin (45). supplementation in patients with diabe-
ing placebo or 4 or 8 g kg1 day1 of Other investigators studied the effects of tes. One group (69) reported no signifi-
CrP, after 8 weeks the two groups taking brewers yeast (23.3 g chromium/day) cant effect of chromium supplementation
chromium had significantly lower glu- and chromium chloride (200 g chromi- (716 months of 250 g/day) versus pla-
cose and insulin levels. Finally, another um/day) on glucose tolerance, serum lip- cebo on serum glucose levels in 76 pa-
(58) observed that corticosteroid-treated ids, and antidiabetic drug dosage in a 16- tients aged 42 83 years (25 of whom had
subjects have accelerated chromium week, randomized, double-blind, type 2 diabetes) with atherosclerotic dis-
losses and that steroid-induced diabetes crossover trial that included 78 patients ease. These results are consistent with
was reversed with CrP supplementation with type 2 diabetes (23,66). Both forms those of another small-scale trial (70) that
at 600 g/day. of chromium supplementation resulted in indicated no significant effects of chro-

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2745


Chromium in diabetes and cardiovascular risk

study populations composed of both dia-


betic and nondiabetic patients.

Individuals with the metabolic


syndrome
Many patients with diabetes have addi-
tional metabolic abnormalities that, taken
together, constitute what has been re-
ferred to as the metabolic syndrome. The
National Cholesterol Education Program
Adult Treatment Panel III has defined the
metabolic syndrome as the presence of
three or more of the following conditions:
waist circumference 102 cm in men and
88 cm in women, serum triglyceride
level 150 mg/dl; HDL cholesterol 40
mg/dl in men and 50 mg/dl in women,
blood pressure 130/85 mmHg, or se-
rum glucose 110 mg/dl (71). Insulin re-
Figure 3Fasting (A) and sistance is a core feature of the metabolic
postprandial (B) glucose de- syndrome and is associated with in-
cline over time in patients with
type 2 diabetes treated for 10
creased cardiovascular disease (CVD)
months with 500 g/day CrP risk, even in the absence of glucose intol-
(60). b, significantly different erance (72). Several studies have evalu-
from baseline (a). ated the effects of chromium
supplementation in patients with compo-
mium supplementation (200 g/day for 2 Gestational diabetes. Chromium sup- nents of the metabolic syndrome.
months) versus placebo on either blood plementation has also shown to be effec- Cefalu et al. (64) assessed the effects
glucose or HbA1c in 30 patients with type tive in improving glucose and insulin of 8 months of treatment with CrP (1,000
2 diabetes. Similarly, another study re- metabolism in women with gestational di- g/day) or placebo on glucose tolerance,
ported that 6 weeks of supplementation abetes. A placebo-controlled study (57) of insulin sensitivity, and body fat in 29 sub-
with 200 g/day chromium in 10 patients 30 women with this condition treated jects with 125% of ideal body weight
with type 2 diabetes was not significantly with 4 or 8 g/kg CrP or placebo showed and a family history of diabetes. Study re-
different from placebo in improving glu- that 8 weeks of chromium supplementa- sults showed that CrP supplementation
cose tolerance or fasting or 2-h serum in- tion significantly decreased fasting levels significantly improved insulin sensitivity
sulin. However, 1-h serum insulin was of glucose, insulin, and C-peptide versus versus placebo (Fig. 4), but had no signif-
significantly lower with chromium sup- placebo. icant effects on glucose effectiveness,
plementation than with placebo (52). Steroid-induced diabetes. Ravina et al. body weight, abdominal fat, or BMI.
The lack of significant effects of chro- (58) showed that administration of chro- These investigators suggested that the
mium supplementation in these three mium can also reverse corticosteroid- positive effect of CrP on insulin sensitivity
studies may be related to the relatively induced diabetes. They treated three without a corresponding change in body
low chromium doses and specific formu- weight or BMI may indicate a direct effect
patients with steroid-induced diabetes
lations used for treatment as discussed of chromium on muscle insulin action. In
with 600 g/day CrP and reported that
above. Abraham et al. (69) treated pa- contrast, another study (30) reported no
fasting blood glucose values fell from 250
tients with 250 g/day CrCl3, Lee and significant changes in glucose or insulin
Reasner (70) administered 200 g/day to 150 mg/dl. The requirement for antidi- metabolism versus placebo after 6
CrP, and Uusitupa et al. (52) treated pa- abetic drugs was also reduced by 50% in months of treatment with Cr3-rich yeast
tients in their trial with 200 g/day CrCl3. these patients. (160 g/day) in a group of 26 elderly sub-
Thus, two of the three studies that failed Summary. Results from the trials noted jects with impaired glucose tolerance and
to document significant positive effects of above support the view that chromium moderate obesity (BMI 30 kg/m2 at
chromium on insulin or glucose metabo- supplementation, especially in the form baseline).
lism used a poorly absorbed inorganic of CrP, in patients with type 1, type 2,
formulation, and the third administered a gestational, or steroid-induced diabetes Individuals without diabetes
very low dose of CrP. These facts under- can improve both glucose and insulin me- Available data suggest the chromium sup-
score the point that chromium formula- tabolism. The reason why chromium sup- plementation has at best limited effects on
tion and dose must be carefully plementation was ineffective in some glucose and insulin metabolism in indi-
considered when evaluating results from studies is not clear, but it is worth noting viduals without diabetes. A small-scale
studies that have assessed its metabolic that all of these trials used relatively low study (56) that included 19 nonobese el-
effects in individuals with or without chromium doses (250 g/day), used derly subjects treated with 1,000 g/day
diabetes. different forms of chromium, or had CrP or placebo for 8 weeks indicated no

2746 DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004


Cefalu and Hu

a prospective study including 33,737


male health care professionals in the U.S.
who were free of chronic disease and pro-
vided toenail samples in 1987. During 7
years of follow-up, there were 367 con-
Figure 4Effects of chro- firmed myocardial infarctions (MIs). Two
mium supplementation on in- control subjects were matched to each
sulin sensitivity in overweight case subject. Study results showed that
subjects with a family history the risk for MI was significantly reduced
of diabetes treated for 8 months in men in the highest quintile for toenail
with 1,000 g/day CrP (64). Cr3. However, this relationship was
Data are means SD. *P only significant for subjects with BMI
0.05, **P 0.005 versus base- 25 kg/m2 (37).
line. In a second study conducted in the
HPFS (38), mean toenail chromium (mi-
significant effect of active treatment on in- (45), while another in elderly subjects crogram per gram) was 0.71 in healthy
sulin sensitivity. Another team (73) re- with impaired glucose tolerance demon- control subjects (n 361), 0.61 in dia-
ported that administration of 400 g/day strated significant reductions in BMI (30). betic subjects (n 688), and 0.52 in di-
chromium for 12 weeks in 44 moderately Of the eight double-blind, placebo- abetic men with prevalent CVD (n 198,
obese middle-aged women who were also controlled trials in individuals without di- P 0.003 for trend). In the cross-
participating in a weight-training and abetes, chromium supplementation sectional analysis, after adjustment of po-
walking program had no significant ef- showed decreases in weight and fat in tential confounders, the odds ratio (OR)
fects versus placebo on FPG, serum insu- three larger studies (26 29,56,73, between extreme quartiles was 0.74 (95%
lin, plasma glucagon, or serum C-peptide. 76,77). CI 0.49 1.11; P 0.18 for trend) com-
Chromium supplementation (220 g/day These results generally support the paring diabetic with healthy control sub-
chromium delivered as chromium nicoti- view that chromium supplementation has jects. A similar comparison between
nate) also had no significant effect versus at best modest effects on body weight or diabetic men with prevalent CVD and
placebo on fasting glucose or immunore- composition in individuals with diabetes healthy control subjects yielded an OR of
active insulin in 26 young volunteers. and perhaps more consistent positive ef- 0.45 (95% CI 0.24 0.84; P 0.003 for
However, chromium administration did fects in healthy volunteers. However, it trend). A nested case-control analysis
significantly reduce immunoreactive in- must be noted that most of the studies comparing diabetic men with incident
sulin levels in subjects with baseline con- addressing this question included only CVD with healthy individuals yielded
centrations 35 pmol/l (31). In contrast, small numbers of subjects and were of rel- similar results. These findings suggest
results from another trial (74) in which 24 atively short duration. that adequate chromium may be im-
elderly subjects (8 with diabetes) were portant for both diabetes and CVD
treated for 8 weeks with either 9 g/day Effects of chromium prevention.
Cr3-rich brewers yeast or Cr3-poor supplementation on the serum lipid The results of the HPFS are consistent
torula yeast indicated that the Cr3-rich profile with those from the European Commu-
supplement significantly improved glu- Many of the studies that evaluated the ef- nity Multicenter Study on Antioxidants,
cose tolerance and decreased insulin fects of chromium supplementation on Myocardial Infarction, and Breast Cancer
output. glucose and insulin metabolism also as- (EURAMIC), an incident, population-
sessed the effects of such treatment on se- based, case-control study conducted in
Chromium effects on body weight rum lipids. Results obtained in studies of eight European countries and Israel to de-
and composition patients with diabetes or glucose intoler- termine whether low toenail chromium
The prevalence of obesity in the U.S. is ance as well as those from normal subjects concentrations are significantly associ-
high, and more than one-half of all adults have indicated variable effects of chro- ated with increased risk for MI. The study
are currently overweight or obese. Obe- mium supplementation on one or more included 684 case subjects (men with a
sity significantly increases the risk for components of the serum lipid profile (22, first diagnosis of MI within 24 h of admis-
development of type 2 diabetes, hyper- 25,30 32,45,52,56,64,66 70,73, sion to the hospital) and 724 control sub-
tension, and CVD (75). Several studies 74,78 81) (Table 2). jects (men with similar demographic
have evaluated the effects of chromium characteristics, but without MI). Average
supplementation on body weight and Relationship between tissue toenail chromium was 1.10 mg/kg in the
composition in individuals with and chromium levels and disease state case subjects vs. 1.30 mg/kg in the control
without diabetes. Risk for coronary heart disease. Two subjects. Additional analysis indicated
Chromium supplementation has epidemiologic studies have evaluated the that the adjusted ORs for MI for chro-
variable effects on body weight and com- relationship between Cr3 levels in toenails mium quintiles 15 were 1.00, 0.82,
position in patients with diabetes (26 (a measure that can best reflect long-term 0.68, 0.60, and 0.59, respectively (82).
30,45,56,73,76,77). One study of intake of trace elements) and risk of cor- The results of EURAMIC thus indicate
patients with diabetes indicated no signif- onary heart disease in men. The Health that toenail chromium concentration has
icant effects on either body weight or BMI Professionals Follow-up Study (HPFS) is a clearly inverse relationship with MI risk

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2747


Chromium in diabetes and cardiovascular risk

1, increased; 2, decreased; apo, apolipoprotein; CO, crossover, Cr3, trivalent chromium, CrN, chromium nicotinate, DB, double blind; JKT, jiangtangkang; PC, placebo controlled; R, randomized; TG,
in men. This relationship remained signif-

Total cholesterol2, LDL cholesterol2,


icant after adjusting for age, BMI, HDL
cholesterol, diabetes, history of hyperten-

LDL cholesterol2, apoB2, HDL


sion, and smoking.

cholesterol1, apoA-I1
TG2, HDL cholesterol1
TG2, HDL cholesterol1
Key results

SAFETY OF CHROMIUM Most


of the concerns regarding the long-term

apoB2, apoA-I1
Total cholesterol2

Total cholesterol2

Total cholesterol2
HDL cholesterol1
HDL cholesterol1

LDL cholesterol2
safety of chromium supplementation
arise from results of several cell culture
studies using supraphysiological doses
No change

No change
No change
No change

No change
No change
No change
that suggested that chromium, particu-
larly in the form of CrP, may increase
TG2

TG2

DNA damage. However, there is currently


no evidence that chromium increases
DNA damage in vivo. There have also
been isolated reports (83) of serious ad-
CrCl3 (250 g/day), brewers yeast

Brewers yeast (160 g/day Cr3)


Chromium supplement (dose)

verse events, including kidney failure, as-


sociated with CrP treatment, but the
CrP (200 or 1,000 g/day)

relationship of chromium to these events


is not clear. Recent reviews of the safety of
(23.3 g/day Cr3)

GTF-Cr (600 g/day)

CrP by the Institute of Medicine (84) and


CrCl3 (250 g/day)

CrCl3 (200 g/day)

CrP (1,000 g/day)


CrP (1,000 g/day)
CrN (220 g/day)
CrP (200 g/day)

CrP (200 g/day)

CrP (200 g/day)

CrP (400 g/day)


CrP (400 g/day)
Cr3 (200 g/day)

Cr3 (200 g/day)

by Berner et al. (85) have concluded that


Brewers yeast

Brewers yeast

CrP is safe. Results from controlled clini-


cal trials (86) have shown that treatment
with chromium at doses up to 1,000 g/
day and for periods as long as 64 months
JKT

does not result in any toxic effects.

CONCLUSIONS A large body of


24 (8 with type 2 diabetes)

literature in both experimental animals


and humans indicates that chromium is
76 (25 with diabetes)
No. of subjects

78 (type 2 diabetes)

30 (type 2 diabetes)
180 (type 2 diabetes)
50 (type 2 diabetes)
188 (type 2 diabetes)
10 (type 2 diabetes)

11 (type 2 diabetes)

an essential element involved in the ac-


26 (all with IGT)

tion of insulin as demonstrated in the


studies of chromium deficiency. Al-
though chromium deficiency has not
been defined beyond that in patients re-
ceiving TPN, epidemiologic studies sug-
43

72
23
28

40
44
26
19
29

gest that tissue levels of chromium are


reduced among diabetic individuals, es-
Table 2Effects of chromium supplementation on serum lipids

pecially in those with existing CVD, com-


R, DB, PC, CO

R, DB, PC, CO

pared with healthy control subjects. Two


DB, PC, CO

DB, PC, CO

DB, PC, CO

DB, PC, CO

case-control studies have also found that


Design

R, DB, PC
R, DB, PC

R, DB, PC

R, DB, PC
R, DB, PC
R, DB, PC

lower toenail chromium levels predict


R, PC

R, PC

R, PC

R, PC

risk of MI in apparently healthy subjects.


DB

However, further epidemiologic studies


R

are needed to confirm these associa-


tions in different populations, and clinical
Preuss, Wallerstedt, and Talpur (81)

trials are needed to prove the causal


Abraham, Brooks, and Eylath (69)

relationship.
Offenbacher and Pi-Sunyer (74)
Chen S, Sun, and Chen X (67)

Amato, Morales, and Yen (56)

A more important question, however,


Press, Geller, and Evans (80)

is the role of chromium supplementation


Patients with diabetes or IGT

Individuals without diabetes

Wilson and Gondy (31)

outside of the rare deficiency states. It is


Riales and Albrink (79)
Rabinowitz et al. (32)
Lee and Reasner (70)

still controversial whether chromium


Anderson et al. (45)

Uusitupa et al. (52)

Uusitupa et al. (30)

Roeback et al. (25)

supplements should be recommended for


Bahijri et al. (66)

Ghosh et al. (22)

Cefalu et al. (64)


Volpe et al. (73)

glycemic control among diabetic patients.


Growing evidence suggests that chro-
Evans (78)

mium supplementation, particularly at


triglyceride.

higher doses and in the form of CrP, may


Study

improve insulin sensitivity and glucose


metabolism in patients with glucose intol-

2748 DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004


Cefalu and Hu

erance and type 1, type 2, gestational, and Accessed 3 February 2004 trial. Ann Intern Med 115:917924, 1991
steroid-induced diabetes and in some in- 11. RDA recommendations [Internet]. Avail- 26. Crawford V, Scheckenbach R, Preuss HG:
dividuals without diabetes. However, it able from http://lifestyler.com/lifestyler/ Effects of niacin-bound chromium sup-
must be recognized that most clinical rdachart.htm. Accessed 29 January 2004 plementation on body composition in
12. Anderson RA: Chromium, glucose intol- overweight African-American women.
studies have major limitations including erance and diabetes. J Am Coll Nutr 17: Diabetes Obes Metab 1:331337,1999
small size, short term, nonrandomized 548 555, 1998 27. Campbell WW, Joseph LJ, Anderson RA,
design, and different doses of chromium 13. Anderson RA: Chromium as an essential Davey SL, Hinton J, Evans WJ: Effects of
supplementation, which may explain the nutrient: the chromium file [Internet]. resistive training and chromium picoli-
high variability of the findings across Available from http://www.chromium- nate on body composition and skeletal
studies. Therefore, more clinical trials are asoc.com/publications/crfile6sep99.htm. muscle size in older women. Int J Sport
needed in the U.S. population to examine Accessed 30 January 2004 Nutr Exerc Metab 12:125135, 2002
the robustness of the results observed in 14. PDRhealth: Chromium [Internet]. Avail- 28. Lukaski HC, Bolonchuk WW, Siders WA,
other populations and appropriate doses. able from http://www.pdrhealth.com/ Milne DB: Chromium supplementation
Ideally, these trials should assess effects of drug_info/nmdrugprofiles/nutsupdrugs/ and resistance training: effects on body
chr_0073.shtml. Accessed 3 February composition, strength, and trace element
treatment on hard end points (e.g., type 2 2004 status of men. Am J Clin Nutr 63:954
diabetes and CVD) as well as metabolic 15. Vincent JB: Elucidating a biological role 965, 1996
parameters, although such trials would be for chromium at a molecular level. Acc 29. Hallmark MA, Reynolds TH, DeSouza
costly and time consuming because they Chem Res 33:503510, 2000 CA, Dotson CO, Anderson RA, Rogers
involve a large number of subjects and at 16. Hepburn DD, Vincent JB: Tissue and sub- MA: Effects of chromium and resistive
least several years of follow-up. Results cellular distribution of chromium picoli- training on muscle strength and body
from such long-term trials would also as- nate with time after entering the composition. Med Sci Sports Exerc
sess the safety of long-term chromium bloodstream. J Inorg Biochem 94:86 93, 28:139 144, 1996
supplementation. 2003 30. Uusitupa MI, Mykkanen L, Siitonen O,
17. Clodfelder BJ, Emamaullee J, Hepburn Laakso M, Sarlund H, Kolehmainen P,
DD, Chakov NE, Nettles HS, Vincent JB: Rassanen T, Kumpulainen J, Pyorala K:
The trail of chromium (III) in vivo from Chromium supplementation in impaired
References the blood to the urine: the roles of trans- glucose tolerance of elderly: effects on
1. Schwarz K, Mertz W: Chromium (III) and ferrin and chromodulin. J Biol Inorg Chem blood glucose, plasma insulin, C-peptide
the glucose tolerance factor. Arch Biochem 6:608 617, 2001 and lipid levels. Br J Nutr 68:209 216,
Biophys 85:292295, 1959 18. Vincent JB: The biochemistry of chro- 1992
2. Tuman RW, Doisy RJ: Metabolic effects of mium. J Nutr 130:715718, 2000 31. Wilson BE, Gondy A: Effects of chromium
the glucose tolerance factor (GTF) in nor- 19. Vincent JB: Mechanisms of chromium ac- supplementation on fasting insulin levels
mal and genetically diabetic mice. Diabe- tion: low-molecular-weight chromium- and lipid parameters in healthy, non-
tes 26:820 826, 1977 binding substance. J Am Coll Nutr 18:6 obese young subjects. Diabetes Res Clin
3. Jeejeebhoy KN, Chu RC, Marliss EB, 12, 1999 Prac 28:179 184, 1995
Greenberg GR, Bruce-Robertson A: Chro- 20. Davis CM, Sumrall KH, Vincent JB: A bi- 32. Rabinowitz MB, Gonick HC, Levin SR,
mium deficiency, glucose intolerance, ologically active form of chromium may Davidson MB: Effects of chromium and
and neuropathy reversed by chromium activate a membrane phosphotyrosine yeast supplements on carbohydrate and
supplementation, in a patient receiving phosphatase (PTP). Biochemistry 35: lipid metabolism in diabetic men. Diabe-
long-term total parental nutrition. Am J 1296312969, 1996 tes Care 6:319 327, 1983
Clin Nutr 30:531538, 1977 21. Potter JF, Levin P, Anderson RA, Freiberg 33. Althuis MD, Jordan NE, Ludington EA,
4. Freund H, Atamian S, Fischer JE: Chro- JM, Andres R, Elahi D: Glucose metabo- Wittes JT: Glucose and insulin responses
mium deficiency during total parenteral lism in glucose-intolerant older people to dietary chromium supplements: a
nutrition. JAMA 241:496 498, 1979 during chromium supplementation. Me- meta-analysis. Am J Clin Nutr 76:148
5. Anderson RA: Chromium and parenteral tabolism 34:199 204, 1985 155, 2002
nutrition. Nutrition 11:83 86, 1995 22. Ghosh D, Bhattacharya B, Mukherjee B, 34. Pittler MH, Stevinson C, Ernst E: Chro-
6. Kimura K: Role of essential trace elements Manna B, Sinha M, Chowdhury J, mium picolinate for reducing body
in the disturbance of carbohydrate metab- Showdhury S: Role of chromium supple- weight: meta-analysis of randomized tri-
olism. Nippon Rinsho 54:79 84, 1996 mentation in Indians with type 2 diabetes als. Int J Obes Relat Metab Disord 27:522
7. Anderson RA: Nutritional factors influ- mellitus. J Nutr Biochem 13:690 697, 529, 2003
encing the glucose/insulin system: chro- 2002 35. Horlick L: Dyslipidemia and metabolic
mium. J Am Coll Nutr 16:404 410, 1997 23. Bahijri SM, Mufti AM: Beneficial effects of factors in the genesis of heart attack and
8. Chromium [Internet]. Available from chromium in people with type 2 diabetes, stroke. Health Rep 6:94 99, 1994
http://www.nums.nwu.edu/nutrition/ and urinary chromium response to glu- 36. Khamaisi M, Wexler ID, Skrha J, Strojek
factsheets/chromium.pdf. Accessed 17 cose load as a possible indicator of status. K, Raz I, Milicevic Z: Cardiovascular dis-
September 2004 Biol Trace Elem Res 85:97109, 2002 ease in type 2 diabetics: epidemiology,
9. Chromium [Internet]. Available from 24. Anderson RA: Chromium as an essential risk factors and therapeutic modalities. Isr
http://www.nlm.nih.gov/medlineplus/ nutrient for humans. Reg Toxicol Pharma- Med Assoc J 5:801 806, 2003
ency/article/002418.htm. Accessed 17 col 26:S35S41,1997 37. Rimm EB, Guallar E, Giovannucci E, As-
September 2004 25. Roeback JR Jr, Hla KM, Chambless LE, cherio A, Stampfer MJ, Willet WC, Hu
10. Expert Group on Vitamins and Minerals: Fletcher RH: Effects of chromium supple- F: Toenail chromium levels and risk of
Review of chromium [Internet], 2002. mentation on serum high-density li- coronary heart disease among normal
Available from www.foodstandards.gov. poprotein cholesterol levels in men taking and overweight men. Presented at the
uk/multimedia/pdfs/reviewofchrome.pdf. beta-blockers: a randomized, controlled American Heart Association 42nd Annual

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2749


Chromium in diabetes and cardiovascular risk

Conference on Cardiovascular Disease, 51. Sherman L, Glennon JA, Brech WJ, glycemic indices, and abdominal fat com-
Epidemiology and Prevention, 2326 April Klomberg GH, Gordon ES: Failure of position in NIDDM. Drug Dev Res 44:17,
2002, Honolulu, Hawaii trivalent chromium to improve hypergly- 1998
38. Rajpathak S, Rimm EB, Li T, Morris JS, cemia in diabetes mellitus. Metabolism 17: 64. Cefalu WT, Bell-Farrow AD, Stegner J,
Stampfer MJ, Willett WC, Hu FB: Lower 439 442, 1968 Wang ZQ, King T, Morgan T, Terry JG:
toenail chromium in men with diabetes 52. Uusitupa MI, Kumpulainen JT, Vouti- Effect of chromium picolinate on insulin
and cardiovascular disease compared lainen E, Hersio K, Sarlund H, Pyorala KP, sensitivity in vivo. J Trace Elem Exp Med
with healthy men. Diabetes Care Koivistoinen PE, Lehto JT: Effect of inor- 12:71 83, 1999
27:22112216, 2004 ganic chromium supplementation on glu- 65. Morris BW, MacNeil S, Stanley K, Gray
39. Vaquero MP: Magnesium and trace ele- cose tolerance, insulin response, and TA, Fraser R: The inter-relationship be-
ments in the elderly: intake, status and serum lipids in noninsulin-dependent di- tween insulin and chromium in hyperin-
recommendations. J Nutr Health Aging abetics. Am J Clin Nutr 38:404 410, 1983 sulinaemic euglycaemic clamps in healthy
6:147153, 2002 53. Mossop RT: Effects of chromium (III) on volunteers. J Endocrinol 139:339 345,
40. Anderson RA: Chromium metabolism fasting glucose, cholesterol and choles- 1993
and its role in disease processes in man. terol HDL levels in diabetics. Cent Afr 66. Bahijri SM, Mira SA, Mufti AM, Ajabnoor
Clin Physiol Biochem 4:31 41, 1986 J Med 29:80 82, 1983 MA: The effects of inorganic chromium
41. Leeds AR: Glycemic index and heart dis- 54. Nath R, Minocha J, Lyall V, Sunder S, Ku- and brewers yeast supplementation on
ease. Am J Clin Nutr 76:286S289S, 2002 mar V, Kapoor S, Dhar KL: Assessment of glucose tolerance, serum lipids and drug
42. Ding W, Chai Z, Duan P, Feng W, Qian Q: chromium metabolism in maturity onset dosage in individuals with type 2 diabe-
Serum and urine chromium concentra- and juvenile diabetes using chromium 51 tes. Saudi Med J 21:831 837, 2000
tions in elderly diabetics. Biol Trace Elem and therapeutic response of chromium 67. Chen SH, Sun YP, Chen XS: Effect of
Res 63:231237, 1998 administration on plasma lipids, glucose jiangtangkang on blood glucose, sensitiv-
43. Davies S, McLaren Howard J, Hunnisett tolerance and insulin levels. In Chromium ity of insulin and blood viscosity in non-
A, Howard M: Age-related decreases in in Nutrition and Metabolism. Shapcott D, insulin dependent diabetes mellitus [in
chromium levels in 51,665 hair, sweat, Hubert J, Eds. Amsterdam, Elsevier/ Chinese]. Zhongguo Zhong Xi Yi Jie He Za
and serum samples from 40,872 patients: North Holland, 1979, p 213222 Zhi 17:666 668, 1997
implications for the prevention of cardio- 55. Glinsman WH, Mertz W: Effect of triva- 68. Ravina A, Slezak L, Rubal A, Mirsky N:
vascular disease and type II diabetes mel- lent chromium on glucose tolerance. Me- Clinical use of the trace element chro-
litus. Metabolism 46:469 473, 1997 tabolism 15:510 520, 1966 mium (III) in the treatment of diabetes
44. Anderson RA, Kozlovsky AS: Chromium 56. Amato P, Morales AJ , Yen SS: Effects of mellitus. J Trace Elem Exp Med 8:
intake, absorption and excretion of sub- chromium picolinate supplementation on 183190, 1995
jects consuming self-selected diets. Am J insulin sensitivity, serum lipids, and body 69. Abraham AS, Brooks BA, Eylath U: The
Clin Nutr 41:11771183, 1985 composition in healthy, nonobese, older effects of chromium supplementation on
45. Anderson RA, Cheng N, Bryden NA, Po- men and women. J Gerontol A Biol Sci Med serum glucose and lipids in patients with
lansky MM, Cheng N, Chi J, Feng J: Ele- Sci 55:M260 M263, 2000 and without non-insulin-dependent dia-
vated intakes of supplemental chromium 57. Jovanovic L, Gutierrez M, Peterson CM: betes. Metabolism 41:768 771, 1992
improve glucose and insulin variables in Chromium supplementation for women 70. Lee NA, Reasner CA: Beneficial effect of
individuals with type 2 diabetes. Diabetes with gestational diabetes mellitus. J Trace chromium supplementation on serum tri-
46:1786 1791, 1997 Elem Exp Med 12:9197, 1999 glyceride levels in NIDDM. Diabetes Care
46. Morris BW, MacNeil S, Hardisty CA, 58. Ravina A, Slezak L, Mirsky N, Bryden NA, 17:1449 1452, 1994
Heller S, Burgin C, Gray TA: Chromium Anderson RA: Reversal of corticosteroid- 71. Expert Panel on Detection, Evaluation,
homeostasis in patients with type II induced diabetes mellitus with supple- and Treatment of High Blood Cholesterol
(NIDDM) diabetes. J Trace Elem Med Biol mental chromium. Diabet Med 16:164 in Adults: Executive summary of the third
13:57 61, 1999 167, 1999 report of the National Cholesterol Educa-
47. Ekmekcioglu C, Prohaska C, Pomazal K, 59. Morris BW, Kouta S, Robinson R, Mac- tion Program (NCEP) Expert Panel on De-
Steffan I, Schernthaner G, Marktl W: Neil S, Heller S: Chromium supplemen- tection, Evaluation, and Treatment of
Concentrations of seven trace elements in tation improves insulin resistance in High Blood Cholesterol in Adults (Adult
different hematological matrices in pa- patients with type 2 diabetes mellitus. Treatment Panel III). JAMA 285:
tients with type 2 diabetes as compared to Diabet Med 17:684 685, 2000 2486 2497, 2001
healthy controls. Biol Trace Elem Res 79: 60. Cheng N, Zhu X, Shi H, Wu W, Chi J, 72. Kendall DM, Harmel AP: The metabolic
205219, 2001 Cheng J, Anderson R: Follow-up survey of syndrome, type 2 diabetes, and cardiovas-
48. Zima T, Mestek O, Tesar V, Tesarova P, people in China with type 2 diabetes mel- cular disease: understanding the role of
Nemecek K, Zak A, Zeman M: Chromium litus consuming supplemental chro- insulin resistance. Am J Manag Care
levels in patients with internal diseases. mium. J Trace Elem Exp Med 12:55 60, 8:S635S653, 2002
Biochem Mol Biol Int 46:365374, 1998 1999 73. Volpe SL, Huang HW, Larpadisorn K,
49. American Diabetes Association: Nutrition 61. Rossetti L, Giaccari A, DeFronzo RA: Glu- Lesser II: Effect of chromium supplemen-
principles and recommendations in dia- cose toxicity. Diabetes Care 13:610 630, tation and exercise on body composition,
betes (Position Statement). Diabetes Care 1990 resting metabolic rate and selected bio-
27 (Suppl. 1):S36 S46, 2004 62. Inzucchi SE, Maggs DG, Spollett GR, Page chemical parameters in moderately obese
50. Trow LG, Lewis J, Greenwood RH, Samp- SL, Rife FS, Walton V, Shulman GI: Effi- women following an exercise program.
son MJ, Self KA, Crews HM, Fairweather- cacy and metabolic effects of metformin J Am Coll Nutr 20:293306, 2001
Tait SJ: Lack of effect of dietary chromium and troglitazone in type II diabetes melli- 74. Offenbacher EG, Pi-Sunyer FX: Beneficial
supplementation on glucose tolerance, tus. N Engl J Med 338:867 872, 1998 effect of chromium-rich yeast on glucose
plasma insulin and lipoprotein levels in 63. Cefalu WT, Bell-Farrow AD, Wang ZQ, tolerance and blood lipids in elderly sub-
patients with type 2 diabetes. Int J Vitam McBride D, Dalgleish D, Terry JG: Effect jects. Diabetes 29:919 925, 1980
Nutr Res 70:14 18, 2000 of glipizide GITS on insulin sensitivity, 75. Sowers JR: Obesity as a cardiovascular

2750 DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004


Cefalu and Hu

risk factor. Am J Med 115:37S 41S, 2003 chloride supplementation on glucose tol- 2003, Boston, Massachusetts
76. Kaats GR, Blum K, Fisher JA, Adelman erance and serum lipids including high- 83. Wasser WG, Feldman NS, DAgati VD:
JA: Effects of chromium picolinate sup- density lipoprotein of adult men. Am J Chronic renal failure after ingestion of
plementation on body composition: a Clin Nutr 34:2670 2678, 1981 over-the-counter chromium picolinate
randomized, double-masked placebo- 80. Press RI, Geller J, Evans GW: The effect of (Letter). Ann Intern Med 126:410,
controlled study. Cur Ther Res 57:747 chromium picolinate on serum choles- 1997
756, 1996 terol and apolipoprotein fractions in hu- 84. Institute of Medicine (IOM): Chromium
77. Kaats GR, Blum K, Pullin D, Keith SC, man subjects. West J Med 152:41 45, picolinate: prototype monograph. In Di-
Woods R: A randomized, double- 1990 etary Supplements: A Framework for Evalu-
masked, placebo-controlled study of the 81. Preuss HG, Wallerstedt D, Talpur N: Ef- ating Safety. Washington, DC, Institute of
effects of chromium picolinate supple- fects of niacin-bound chromium and Medicine, 2004, p. B1B80
mentation on body composition: a repli- grape seed proanthocyanidin extract on 85. Berner TO, Murphy MM, Slesinski R: De-
cation and extension of a previous study. the lipid profile of hypercholesterolemic termining the safety of chromium tripico-
Curr Ther Res 59:379 388, 1998 subjects: a pilot study. J Med 31:227246, linate for addition to foods as a nutrient
78. Evans GW: The effect of chromium pico- 2000 supplement. Food Chem Toxicol 42:1029
linate on insulin controlled parameters in 82. Guallar E: Toenail chromium status and 1042, 2004
humans. Int J Biosoc Med Res 11:163180, cardiovascular disease risk in Europe. 86. Jeejeebhoy KN: The role of chromium in
1989 Presented at Chromium in Health and Dis- nutrition and therapeutics and as a poten-
79. Riales R, Albrink MJ: Effect of chromium ease, a CADRE Research Summit, 12 April tial toxin. Nutr Rev 57:329 335, 1999

DIABETES CARE, VOLUME 27, NUMBER 11, NOVEMBER 2004 2751