Journal of the American College of Nutrition, Vol. 26, No. 6, 663-668 (2007)
Published by the American College of Nutrition
The Use of Konjac Glucomannan to Lower Serum Thyroid Hormones in Hyperthyroidism
Adil Dogan Azezli, MD,
Taner Bayraktaroglu, MD and
Yusuf Orhan, MD
Istanbul University, Istanbul Faculty of Medicine, Department of Internal Medicine, Division of Endocrinology Metabolism and Nutrition
Address reprint requests to: Taner Bayraktaroglu, MD, Istanbul University, Istanbul Faculty of Medicine, Dept. Internal Medicine, Division of Endocrinology Metabolism and Nutrition, Fatih-Istanbul, TURKEY. E-mail: baytaner{at}yahoo.com
 |
ABSTRACT
|
|---|
Objective: Patients with hyperthyroidism occasionally need rapid restoration to the euthyroid state. In wiew of the increased enterohepatic circulation of thyroxine (T4) and triiodothyronine (T3) in thyrotoxicosis, and metabolic effects of konjac glucomannan in gastrointestinal system, we aimed to determine the activity of glucomannan in treatment of hyperthyroidism.
Methods: A prospective, randomized, placebo-controlled, one-blind study design was used with newly diagnosed 48 hyperthyroid patients (30 patients with Graves disease and 12 with multinodulary goitre). They were assigned to one of the following treatment groups: I) methimazole 2x10mg, propranolol 2x20mg, and glucomannan (Propol) 2x1.3gr daily for two months; II) methimazole 2x10mg, propranolol 2x20mg, and placebo powder daily for two months.
Results: No differences were detected from the point of view of the baseline thyroid hormone levels between groups (p > 0.05). Further analyses revealed that the patients receiving glucomannan at the end of the second, fourth and sixth weeks of the study had significantly lower serum T3, T4, FT3 and FT4 levels than the patients who received placebo (p < 0.05). TSH was not different between the two groups at any specific time (p > 0.05). At week 8, thyroid hormone levels were not shown any differences. The glucomannan-treated group had a more rapid decline in all four serum thyroid hormone levels than the placebo-treated group.
Conclusions: We believe our preliminary results indicate that glucomannan may be a safe and easily tolerated adjunctive therapeutic agent in the treatment of thyrotoxicosis. This combination therapy seems most effect during first weeks of treatment of a hyperthyroid patient.
Key words: konjac glucomannan, hyperthyroidism, thyroid hormones
 |
INTRODUCTION
|
|---|
There has been great interest in the physiobiological and pharmacological effects of dietary fibers, recently [1–4]. The effectiveness of gel-forming fibers which have the capacity to hydrate and the viscosity of digesta are dependent on fiber concentration, molecular weight, and also size-distribution of the fiber-gum particles. Konjac glucomannan have the highest viscosity among polysaccharides [5]. Glucomannan is a water-soluble dietary fiber derived from konjac root (Amorphophallus konjac). Its main constituent is highly viscous glucomannan, a polysaccharide chain containing glucose and mannose in a molar ratio of 1:1.6 with β-1–4 linkages. It has been shown that serum lipids and systolic blood pressure, postprandial glycemia, and body weight were significantly reduced by glucomannan [6–8]. It has been reported immunomodulatory effects such as production of interferon, cytokine and immunoglobulin of lymphocytes in experimental studies [4,9,10].
Hyperthyroidism (e.g. Basedow-Graves Disease, toxic multinodular goitre, toxic adenoma) is treated by reducing thyroid hormone synthesis, using antithyroid drugs, or by reducing the amount of thyroid tissue with radioiodine (131I) treatment or by surgery. During this time, while new hormone instead of syntesis is blocked, patients may stil be symptomatic from excessive circulating or peripherally stored hormone. Thyrotoxic crisis, or thyroid storm, is rare and presents as a life-threatening exacerbation of hyperthyroidism. Propranolol or longer acting beta blockers, such as atenolol, may be helpful to control adrenergic symptoms, especially in the early stages before antithyroid drugs take effect [11,12,13]. Preliminary studies suggest that bile acid sequestrants may be useful in treating exogenous thyroid-hormone-induced hyperthyroidism in emergent situations [14]. It was shown that exchange resins such as cholestiramine which blocks enterohepatic circulation are effective in producing a rapid and complete decline of thyroid hormones in patients with hyperthyroid Graves disease, when used in combination with methimazole and propranolol [15,16]. It may even be that rapidly correction of Graves disease may be affected by immunomodulatory effects of glucomannan.
There are no knowledges about the use of glucomannan in the treatment of thyrotoxicosis in the literature. Conjugation of thyroxine to glucuronides and sulphates is increased in thyrotoxic states, and there is increased biliary, urinary and fecal loss of both conjugated and fT4 [17]. In this study, it was examined the effects of konjac glucomannan which was used adjunctively with thionamides and a beta-blocker to lower serum iodothyronine levels in patients with hyperthyroidism.
 |
MATERIALS AND METHODS
|
|---|
A prospective, randomized, placebo-controlled, single blind study design was used with subjects. All subjects gave written informed consent to participat in the present study; which was approved by the local ethic commitees. Newly diagnosed 48 patients (12 men, 36 women; age 39.1 ± 12.0 years between 20–63 years) with 30 hyperthyroid Graves disease and 18 multinodular goitre were randomly assigned to one of the following treatment groups: I) methimazole (Thyromazol, Abdi Ibrahim Co., Turkey) 2x10mg, propranolol (Dideral, Sanofi Do
u Co., Turkey) 2x20mg, and glucomannan [Propol; white powder, viscosity min. 100 000 cps, minimum 95% fiber content, Behr, Wunderlich & Co. (GmBH & Co.), Germany] 2x1.3gr daily for two months (the added cost of treating hyperthyroid patients with glucomannan during this period is approximately 65 $ per patient); II) methimazole 2x10mg, propranolol 2x20mg, and placebo powder daily for two months. Twenty-four patients were enrolled in each group. No iodine was contained in the excipient of the glucomannan or in the placebo powder.
Patients with diabetes mellitus, disorders of kidney, pulmonary, gastrointestinal or liver were excluded. The diagnosis of hyperthyroidism was established with the presence of signs and symptoms of thyrotoxicosis accompanied by a diffuse goitre, an increased radioiodine uptake (>40%) at 24 h, with or without orbitopathy by degrees and autoantibodies to thyroid gland (TSH receptor antibody, anti-thyroid peroxydase and anti-thyroglobuline) in patients with Graves disease; and palpable-scintigraphyc hyperactive nodules in patients with multinodulary goitre, and biochemical evidence of hyperthyroidism (high circulating total and free T4 and total T3, as well as a supressed TSH levels).
Patients were followed two weeks interval for two months with thyroid function tests. All these samples were measured after an overnight fast. Serum total T3, total T4, free T3, free T4 and thyroid stimulating hormone (TSH) were measured by electrochemiluminescence immunoassay [ECLIA, Roche Elecsys 1010/2010 and Modular Analytics E170 (Elecsys module) immunoassay analyzers, Indianapolis, IN, USA] each after two weeks with inter- and intraassay CVs of 5.3% and 5.4%, 4.7% and 6.9%, 2.2% and 2.8%, 2.9% and 6.6%, 2.1% and 3.6% respectively. Normal laboratory ranges are as follows: total T3, 1.3–3.1 nmol/L; total T4, 66–181 pmol/L; free T3, 3.95–6.8 pmol/L; free T4, 12–22 pmol/L and TSH, 0.27–4.2 µIU/ml.
Statistical Analyses
Results are expressed as means ± SD. Significant differences between groups were analysed using "t tests" and "Mann-Whitney test" for continuous data that were not normally distributed. Statistical comparisons were evaluated by computer using the SSPS (Statistical Package for Social Science) for Windows Release 11.0.0 Standart version (Copyright SPSS Inc. 1989-2001). p-values < 0.05 were considered statistically significant.
 |
RESULTS
|
|---|
The two groups were matched in terms of age (39.8 ± 11.3 years in group with glucomannan, 38.4 ± 12.9 years in group with plasebo) and baseline thyroid hormone levels (p > 0.05, Table 1). It was shown the results on the change in body mass index, blood pressure and serum biochemical parameters (glucose, triglyceride and total cholesterol) during the experimental period in two groups in Table 1. Altough body mass indexes were significantly increased (p < 0.05), changes of blood pressures and biochemical parameters were not different than baseline (p > 0.05, Table 1). Body mass index, blood pressure, serum biochemical parameters with thyroid hormone levels were unsignificantly changed at baseline and during the study period between subjects with Graves disease and toxic multinodular goitre (data not shown).
View this table:
[in this window]
[in a new window]
|
Table 1. Changes in Body Mass Indexes, Blood Pressures and Serum Biochemical Parameters (Glucose, Triglyceride and Total Cholesterol) during the Experimental Period in Two Groups
|
|
The results of the thyroid hormone levels at the end of the second, fourth, sixth and eighth weeks of the study appear Table 2. No differences were detected from the point of view of the baseline thyroid hormone levels between groups (p < 0.05). Further analyses revealed that the patients receiving glucomannan at the end of the second, fourth and sixth weeks of the study had significantly lower serum T4, fT3 and fT4 levels than the patients who received placebo (p < 0.05). At the end of the second, fourth and sixth weeks, total and free thyroid hormone levels had decreased more in glucomannan group than in plasebo group (p < 0.05). The percent reduction in T3 and fT4 levels was not different between the two groups at the end of the eighth week (p > 0.05, Table 2). The serum T3 levels at 2 weeks, and T3, T4, T3/T4 ratio, free T3 and fT4 levels at 8 weeks are not different statistically between the groups. In sixth week, T3 to T4 ratio was significantly lower than placebo group. It was unsignificantly observed baseline, second, fourth and eighth weeks between groups. Means of TSH levels was not different between the two groups at any specific time (p > 0.05). TSH levels remained suppressed in the two groups throughout the second, fourth and sixth weeks. However, it was normalized in patients who receiving glucomannan at the end of eighth weeks. The glucomannan-treated group had a more rapid decline in all four serum thyroid hormone levels than the placebo-treated group. The TSH levels were early carried out insignificantly in patients who received glucomannan than others (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Hormone Levels and % Reduction at Baseline, the End of the Second, Fourth, Sixth and Eighth Weeks of the Study
|
|
All patients tolerated the glucomannan well. There was no increased bloating, flatulence or constipation with either drug or placebo. Patient's signs and syptoms were resolved and did not progress during study period.
 |
DISCUSSION
|
|---|
The physical-chemical structure of glucomannan may be an important determinant of its cholesterol and thyroid hormones-lowering ability. Exchanging anions, promotion of bile-acid excretion in the stool and/or blockage of cholesterol and thyroid hormones absorbtion without replacing to circulation are possible mechanisms similar to cholestyramine [6,18,19]. Exchange resins such as cholestyramine are effective in patients with hyperthyroid Graves disease when used in combination with methimazole and propranolol. Solomon et al and Mercado et al showed that cholestyramine produces a more rapid decline in thyroid hormone levels than antithyroid medications used alone [15,16]. Cholestyramine binds T4 in the intestine and increases its fecal excretion. This phenomena were proven in hamsters in the mid 1960s [20,21]. It has been shown that 50mg of cholestyramine can bind at least 3000µg of thyroxine [14].
Glucomannan delays gastric emptying, leading to a more gradual absorption of dietary sugar; this effect can reduce the elevation of blood sugar levels that is typical after a meal [22]. To the beneficial properties of this fiber, several disadvantages can be added as the production of flatulence, abdominal pain, esophageal obstruction, lower gastrointestinal obstruction or even the possible modification of the bioavailability of other drugs. Like other soluble fibers, glucomannan can bind to bile acids in the gut and carry them out of the body in the feces, which requires the body to convert more cholesterol into bile acids [23]. This can result in the lowering of blood cholesterol and other blood fats. Controlled [24] and double-blind [6,8,22] studies have shown that supplementation with several grams per day (8–13 grams per day) of glucomannan significantly reduced total blood cholesterol, low-density lipoprotein cholesterol, and triglycerides, and in some cases raised high-density lipoprotein cholesterol. There is convincing evidence that the effectiveness of gel forming fibers seems to relate mainly to their capacity to hydrate rapidly and thus increase the viscosity of digesta in the stomach and small intestine. These properties are dependent on fiber concentration, molecular weight, and also size distribution of the fiber-gum particles. Glucomannan is very important in this regard for, when properly selected, it may have the highest viscosity amongst polysaccharides [5]. In this study, serum lipid parameters were not significantly different between baseline and after two months, and in both groups, because of lipid disorders in hyperthyroid patients are too few.
In our study, it was used same doses of propranolol and methimazole during study, means of age and baseline thyroid hormone levels were not different between groups. The patients receiving glucomannan had significantly lower serum T3, T4, fT3 and fT4 levels than the patients who received placebo at the end of the second, fourth, sixth and eighth weeks of the study. At the second weeks, serum T3 levels were low in patients receiving glucomannan, but were not different between groups. All patients in two groups were get to be euthyroid status, but T3 and T4 levels of two patients receiving glucomannan and one in placebo group were in hypothyroid ranges at the end of the sixth weeks. Drugs and glucomannan were tolerated well. This study shows that giving glucomannan to hyperthyroid patients has a modest effect on accelerating the decline in serum thyroid hormones when these patients are treated with methimazole and propranolol. In this study, it was dedected that the rate at which the serum concentrations of throid hormone declined was faster in the glucomannan-treated group when compared with placebo. These differences were similar in magnitude to the effects of adjuvant therapy with cholestyramine studies which were reported in the 1990s. But we didn't examined a clear clinical advantage in the use of glucomannan adjuvant therapy compared with that of cholestyramine in clinical practice.
It is unclear what mechanism is involved in the effect of the compound to reduce thyroid hormone levels. It was suggested that glucomannan was not get into vascular compartmant, its effects on levels of thyroid hormone were introduced through its alterations in enterohepatic circulation. At the end of the sixth weeks, thyromazole blocked new hormone syntesis, therefore thyroid hormones which entered to enterohepatic circulation were reduced. Starting to rise of the TSH levels were early carried out in patients who received glucomannan than others, insignificantly. Its means clarified that the axis of the hypothalamo-hypophyseal-thyroid were recovered in a little early time. It is possible that dietary fibers also influence immune function as a consequence of changes in mucosal structure and gut microflora. It was speculated that fermentation by intestinal bacteria of dietary fibers leads to a more acidic environment in the intestine, and this situation inhibits the conversion of primary bile acids to their secondary counterparts [4, 25], and bile acids enhance immunoglobulin E production by mesenteric lymph node lymphocytes and suppress by mesenteric lymph node lymphocytes and suppress the production of IgA, IgG and IgM in an indirect manner, and their effect on cytokine production by immune cells partly participitates in this enhacement that can be encountered in disease states [26, 27]. It was shown that production of interferon
which is mostly immunomodulatorial cytokin was induced in the rats had been fed on glucomannan [9]. Multiple factors contribute to the etiology of Graves disease, including host genetic and environmental factors. It was indicated that Graves disease was a slowly progressing disease that involved activation and recruitment of thyrotropin receptor-specific T and B cells, and resulted in the production of stimulatory antibodies that could cause hyperthyroidism [28]. In the present study, the reason of the rapidly decline of thyroid hormones may be the immunomodulatory and physicochemical effects of glucomannan in hyperthyroid patients, but the both groups have Graves disease and non-autoimmune multinodulary toxic goitre. It can been studied these important features of glucomannan as an objective of a research in patients with Graves disease.
In conclusion, it was thought that glucomannan decreased levels of thyroid hormone through its effects on enterohepatic circulation. By means of this activity, thyroid hormone levels in patients received glucomannan was lower than placebo group. It was listed in order glucomannan effects on enterohepatic circulation, probably; glucomannan increases β-glucronidase activity, colonization of bifidobacterium, afterwards provides deconjugation [29,30], thus increases hydrolysis of the glucronid metabolits and fecal excretion, accelerates intestinal transit time, and increases fecal volume [31,32], decreases absorbable free water [33–35], binds directly thyroid hormones by features of ione exchanging, similar to cholestyramine [6,14,18–21]. It may be affect the progression of Graves disease by immunomodulatory action of glucomannan. But it is difficult to say that how much amount are each influences effective. It has been demonstrated that this product is highly effective in the treatment of diabetes mellitus, insulin resistance syndrome [22, 35] and obesity due to the satiety sensation that it produces; as a remedy for constipation, because it increases the faeces volume; as hypocholesterolemic agent, interfering in the transport of cholesterol and of bile acids and as hypoglycemic and hypoinsulinemic agent, probably, by delaying gastric emptying and slowering glucose delivery to the intestinal mucosa. Results of the present study indicate that glucomannan may be a safe and easily tolerated adjunctive therapeutic agent in the treatment of thyrotoxicosis during first weeks.
Received April 24, 2005.
Accepted February 6, 2007.
 |
REFERENCES
|
|---|
- Edwards CA: The physiological effects of dietary fiber. In Kritchevsky D, Bonefield C (eds): "Dietary Fiber in Health & Disease." St. Paul MN: Eagan Pres, pp58
–71,1995
.
- Mackeown-Eyssen GE, Bright-See E: Dietary factors in colon cancer: international relationships. Nutr Cancer6
:160
–170,1984
.[Medline]
- Schneeman BO, Tinker LF: Dietary fiber. Pediatric Nutr42
:825
–838,1995
.
- Lim BO, Yamada K, Nonaka M, Kuramoto Y, Hung P, Sugano M: Dietary Fibers Modulate Indices of Intestinal Immune Function in Rats. J Nutr127
:663
–667,1997
.[Abstract/Free Full Text]
- Vuksan V, Sievenpiper JL, Xu Z, Wong EY, Jenkins AL, Beljan Zdravkovic U, Leiter LA, Josse RG, Stavro MP: Konjac-Mannan and American ginsing: emerging alternative therapies for type 2 diabetes mellitus. J Am Coll Nutr20
:370S
–380S,2001
.[Abstract/Free Full Text]
- Arvill A, Bodin L: Effect of short-term ingestion of Konjac glucomannan on serum cholesterol in healthy men. Am J Clin Nutr61
:585
–589,1995
.[Abstract/Free Full Text]
- Doi K: Effect of Konjac fibre (glucomannan) on glucose and lipids. Eur J Clin Nutr49
:S190
–197,1995
.[Medline]
- Walsh DE, Yaghoubian V, Behfrooz A: Effect of glucomannan on obese patients: A clinical study. Int J Obesity8
:289
–293,1984
.[Medline]
- Yamada K, Tokunaga Y, Ikeda A, Okhura K, Mamiya S, Kaku S, Sugano M, Tachibana H: Dietary effect of guar gum and its partially hydrolyzed product on the lipid metabolism and immune function of Sprague-Dawley rats. Biosci Biotechnol Biochem63
:2163
–2167,1999
.[Medline]
- Onishi N, Kawamoto S, Nishimura M, Nakano T, Aki T, Shigeta S, Shimizu H, hashimoto K, Ono K: A new immunomodulatory function of low-viscous konjac glucomannan with a small particle size: its oral intake suppresses spontaneously occurring dermatitis in NC/Nga mice. Int Arch Allergy Immunol136
:258
–265,2005
.[Medline]
- Cooper DS: Antithyroid drugs. N Engl J Med311
:1353
–1362,1984
.[Abstract]
- Croxon MS, Hall TD, Nicoloff JT: Combinationdrug therapy for treatment of hyperthyroid Graves disease. J Clin Endocrinol Metab45
:623
–630,1977
.[Abstract/Free Full Text]
- Larsen PR, Davies TF: Thyroid physiology and diagnostic evaluation of patients with thyroid disorders. In Larsen PR, Kronenberg HM, Melmed S, Polonsky KS (eds): "Williams Textbook of Endocrinology," 10th ed. Philadelphia: W.B. Saunders Company, pp331
–373,2002
.
- Shakir MKM, Michaels RD, Hays HJ, Potter BD: The use of the bile acid sequestrants to lower serum thyroid hormones in iatrogenic hyperthyroidism. Ann of Intern Med118
:112
–113,1993
.[Abstract/Free Full Text]
- Solomon LB, Wartofsky L, Burman KD: Adjunctive cholestyramine therapy for thyrotoxicosis. Clin Endocrinol38
:39
–43,1993
.[Medline]
- Mercado M, Mendoza-Zubieta V, Bautista-Osorio R, Espinoza-de los Monteros AL: Treatment of hyperthyroidism with a combination of methimazole and cholestyramine. J Clin Endocrinol Metab81
:3191
–3193,1996
.[Abstract]
- Nicoloff JT: Physiologic and pathophysiologic implications of hormone binding. In Ingbar S, Braverman L (eds): "Werner's The Thyroid: A Fundemental and Clinical Text," 5th ed. New York: J.B. Lippincott, pp113
–134,1986
.
- Ikegami S, Hareda E, Tsuchihashi N, Nagayama S, Nishide E, Innami S: Effect of indigestible polysaccharides on pancreatic exocrine secretion and biliary output. J Nutr Sci Vitaminol30
:515
–523,1984
.[Medline]
- Kriyama S, Enishi A, Yoshida A, Sugiyama N, Shimahora H: Hypocholesterolemic activity and molecular weight of konjac mannan. Proc Japan Acad4
:231
–236,1972
.
- Bergman F, Heedman P, Van der Linden W: Influence of cholestyramine on absorbtion and excretion of thyroxine in Syrian hamsters. Acta Endocrinol (Copenh)53
:256
–263,1966
.[Abstract/Free Full Text]
- Northcutt RC, Stiel JN, Hollifield JW, Stant EG: The influence of cholestyramine on thyroxine absorbtion. JAMA208
:1857
–1861,1969
.[Abstract/Free Full Text]
- Vuksan V, Sievenpiper JL, Owen R, Swilley JA, Spadafora P, Jenkins DJ, Vidgen E, Brighenti F, Josse RG, Leiter LA, Xu Z, Novokmet R: Beneficial effects of viscous dietary fiber from Konjac-mannan in subjects with the insulin resistance syndrome: results of a controlled metabolic trial. Diabetes Care23
:9
–14,2000
.[Abstract]
- Wu J, Peng SS: Comparison of hypolipidemic effect of refined konjac meal with several common dietary fibers and their mechanisms of action. Biomed Environ Sci10
:27
–37,1997
.[Medline]
- Zhang MY, Huang CY, Wang X, Hong JR, Peng SS: The effect of foods containing refined Konjac meal on human lipid metabolism. Biomed Environ Sci3
:99
–105,1990
.[Medline]
- Jacob LR: Fiber and colon cancer. Gastroenterol Clin North Am17
:747
–760,1988
.[Medline]
- Lim BO, Yamada K, Sugano M: Effect of bile acids and lectins on immunoglobulin production in rat mesenteric lymph node lymphocytes. In Vitro Cell Develop Biol30
:407
–413,1994
.
- Lim BO, Yamada K, Yoshimura K, Watanabe T, Pham H, Taniguchi S, Sugano M: Free bile acids inhibit IgE production by mouse spleen lymphocytes stimulated by lipopolysaccaharide and interleukins. Biosci Biotech Biochem59
:624
–627,1995
.[Medline]
- Prabhakar BS, Bahn RS, Smith TJ. Current perspective on the pathogenesis of Graves disease and ophthalmopathy. Endocr Rev24
:802
–35,2003
.[Abstract/Free Full Text]
- Mizutani T, Mitsuoka T: Effect of konjac mannan on spontaneous liver tumorigenesis and fecal flora in C3H/Hc mal emice. Cancer Letters17
:27
–32,1982
.[Medline]
- Mizutani T, Mitsuoka T, Nishikawa Y: Influence of dietary konjac mannan on fecal microflora in f344 rats. Proc Japan Acad63
:273
–276,1987
.
- Gonzalez Canga A, Fernandez Martinez N, Sahagun AM, Garcia Vieitez JJ, Diez Liebana MJ, Calle Pardo AP, Castro Robles LJ, Sierra Vega M: Glucomannan: properties and therapeutic applications. Nutr Hosp19
:45
–50,2004
.[Medline]
- Kato K, Matsuda K: Studies on the chemical structure of konjac mannan. Part I. Isolation and characterization of oligosaccharides from the partial acid hydrolizate of mannan. Agric Biol Chem33
:1446
–1453,1969
.
- Kishida N, Okimasu S, Kamata N: Molecular weight and intrinsic viscosity of konjac gluco-mannan. Agric Biol Chem42
:1645
–1650,1978
.
- Maekaji K: The mechanism of gelation of konjac mannan. Agric Biol Chem38
:315
–321,1974
.
- Vuksan V, Jenkins DJ, Spadafora P, Sievenpiper JL, Owen R, Vidgen E, Brighenti F, Josse R, Leiter LA, Bruce-Thompson C: Konjac-mannan (glucomannan) improves glycemia and other associated risk factors for coronary heart disease in type 2 diabetes. A randomized controlled metabolic trial. Diabetes Care22
:913
–919,1999
.[Abstract]