- Email: [email protected]
Dietary supplementation with d -tagatose in subjects with type 2 diabetes leads to weight loss and raises high-density lipoprotein cholesterol
Available online at www.sciencedirect.com
Nutrition Research 30 (2010) 801 – 806 www.nrjournal.com
Dietary supplementation with D-tagatose in subjects with type 2 diabetes leads to weight loss and raises high-density lipoprotein cholesterol☆ Thomas W. Donner a,⁎, Laurence S. Magder b , Kiarash Zarbalian a a
Department of Internal Medicine, University of Maryland School of Medicine, Baltimore, MD 21201, USA b Department of Epidemiology, University of Maryland School of Medicine, Baltimore, MD 21201, USA Received 19 May 2010; revised 11 September 2010; accepted 13 September 2010
Abstract Oral D-tagatose (D-tag) attenuates the rise in plasma glucose during an oral glucose tolerance test in subjects with type 2 diabetes mellitus (DM) and reduces food intake in healthy human subjects. A reduction in food consumption and less weight gain occur in rats fed tagatose. This pilot study explored the metabolic effects of D-tag given daily to 8 human subjects with type 2 DM for 1 year. We hypothesized that this treatment period would lead to weight loss and improvements in glycated hemoglobin and the lipid profile. A 2-month run-in period was followed by a 12-month treatment period when 15 g of oral D-tag was taken 3 times daily with food. No serious adverse effects were seen during the 12-month treatment period. Ten of the initially12 recruited subjects experienced gastrointestinal side effects that tended to be mild and transient. When 3 subjects were excluded who had oral diabetes, medications added and/or dosages increased during the study and mean (SD) body weight declined from 108.4 (9.0) to 103.3 (7.3) kg (P = .001). Glycated hemoglobin fell nonsignificantly from 10.6% ± 1.9% to 9.6% ± 2.3% (P = .08). High-density lipoprotein cholesterol progressively rose from a baseline level of 30.5 ± 15.8 to 41.7 ± 12.1 mg/dL at month 12 in the 6 subjects who did not have lipid-modifying medications added during the study (P b .001). Significant improvements in body weight and high-density lipoprotein cholesterol in this pilot study suggest that D-tag may be a potentially useful adjunct in the management of patients with type 2 DM. © 2010 Elsevier Inc. All rights reserved. Keywords: Abbreviations:
Tagatose; Diabetes control; Weight loss agents; HDL cholesterol; Lipid modification; Bulk sweeteners CHD, coronary heart disease; DM, diabetes mellitus; D-Tag, D-tagatose; GlyHb, glycated hemoglobin; LDL, low-density lipoprotein.
1. Introduction Nonnutritive sweeteners are used commonly to help patients with type 2 diabetes mellitus (DM) reduce carbohydrate energy intake to affect better glycemic control
☆
Financial support was received from Maryland Industrial Partnerships, Glenn L. Martin Institute of Technology, University of Maryland, College Park, and Spherix Inc, Beltsville, Md. ⁎ Corresponding author. Division of Endocrinology and Metabolism, Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA. Tel.: +1 410 955 2908; fax: +1 410 614 9586. E-mail address: [email protected] (T.W. Donner). 0271-5317/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2010.09.007
and to assist with weight management. D-Tagatose (D-tag) is a ketohexose bulk sweetener with no net metabolizable energy when fed to rats [1,2]. D-tagatose received Generally Recognized as Safe status by the Food and Drug Administration in 2001 and entered the US market as a sweetener in 2003. In subjects with and without type 2 DM, a 75-g oral D-tag tolerance test led to no changes in plasma glucose or insulin levels. Pretreatment of type 2 DM subjects with doses of D-tag ranging from 10 to 75 g attenuated the rise in plasma glucose during an oral GTT in a dose-dependent manner [3]. These findings suggest that D-tag could potentially improve glycemic control in patients with type 2 DM by blunting postprandial hyperglycemia.
802
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
The long-term metabolic effects of D-tag have not previously been studied in humans. In a 90-day study in rats, D-tag given as 15% to 20% of daily energy intake led to less weight gain when compared with a control diet [4]. Reduced food intake has been observed in rats fed tagatose as 20% of daily energy intake [4] and in humans fed a single 29-g dose of tagatose when compared with 29 g of sucrose added to a breakfast meal [5]. If tagatose causes sustained reductions in energy intake, beneficial effects on weight, blood glucose control, and serum lipid profile could be seen. The objective of this pilot study was to investigate the metabolic effects of long-term administration of D-tag in obese subjects with type 2 DM. Following a 2-month run-in period, subjects had tagatose added to their daily meals and were followed on an every 2-month basis for 12 months for changes in weight, blood glucose control, and lipid profile. We hypothesized that daily oral tagatose would lower glycated hemoglobin (GlyHb) and lead to weight loss and improvements in the lipid profile, including lower low-density lipoprotein (LDL) cholesterol and triglycerides, and higher high-density lipoprotein (HDL) cholesterol.
2. Methods and materials 2.1. Subjects Eight subjects (4 male and 4 female) with type 2 DM for at least 1 year in duration were enrolled and ranged from 35 to 70 years in age. Four of the 12 initially screened subjects were excluded from analysis because they did not complete the study. Data analysis was performed on the 8 subjects who completed all 14 months of the study. Subjects were excluded from study entry if baseline GlyHb levels were lower than 8% (reference range, 4.4%-7.7%) or if they had gastrointestinal disorders. All investigational protocols were approved by the institutional review board of the University of Maryland for human subjects. Participants were enrolled only after giving written informed consent. At entry, 3 subjects were receiving diet therapy for diabetes, 4 were being treated with a sulfonylurea, and 1 was being treated with combination metformin and troglitazone. Those receiving medications had been on stable doses for at least 10 months before enrollment. Three subjects had oral diabetes medications added or adjusted by their primary care providers during the intervention phase of the study. These 3 subjects were included in all outcomes analyses, though subgroup analyses excluding these subjects were done for GlyHb and body weight, which are known to be affected by sulfonylureas and troglitazone. One subject had glipizide 5 mg added during month 4, and another subject had her glyburide dosage increased during month 8 of the intervention phase. The third subject had glyburide 2.5 mg added during month 3 and had the troglitazone dose increased from 200 to 400 mg daily during month 4 of the intervention phase. One subject had pravastatin added during month 8. A lipid subgroup analysis
was performed, which excluded this subject and the one whose troglitazone was increased during the study. A subgroup analysis was also performed on 3 subjects who had antihypertensive medications added during the study. 2.2. Materials D -Tagatose was prepared by Spherix Incorporated (Beltsville, Md). Samples provided to subjects were more than 99% pure by high-performance liquid chromatography analysis and were weighed and packaged in the University of Maryland Pharmacy before administration.
2.3. Study protocol After a 2-month run-in period, subjects were given 15-g packages of D -tag to be taken 3 times daily with nonstandardized meals. Dose-dependent, gastrointestinal side effects observed with large doses of D-tag have been attributed to an osmotic effect of this poorly absorbed sugar. Prior tolerance testing at the University of Maryland has shown the 15-g dosage to be typically well tolerated with minimal adverse gastrointestinal effects [3]. The D-tag was dissolved in liquids, used in baking, or added to prepared foods. For the duration of the 14-month study period, subjects were encouraged not to otherwise alter their dietary intake. All subjects enrolled were physically inactive and remained so throughout the 14-month observation period. At the initial visit and every 2 months for the next 14 months, body weight and vital signs were recorded, and fasting blood was sent for glucose, GlyHb, insulin, lipids, liver and kidney function, uric acid, bilirubin, phosphorus, calcium, magnesium, bicarbonate, chloride, sodium, potassium, total protein, albumin, and amylase. Subjects were questioned at each visit about the occurrence of any adverse events or side effects, and specifically about any gastrointestinal side effects. Compliance with D-tag was confirmed based on D-tag package counts performed every 2 months. 2.4. Laboratory evaluations Glycosylated hemoglobin values were measured by the affinity column method (Helena Glyco-Tek, Beaumont, Tex), which has a reference range of 4.4% to 7.7% and a mean intra-assay coefficient of variation of 2.95% [6]. Insulin levels were measured by the Coates-A-count RIA method (Diagnostics Products Corporation, Los Angeles, Calif). Chemistry profiles and lipids were performed at the University of Maryland Chemistry Laboratory using Beckman Coulter laboratory analyzers (Brea, Calif). Sodium, potassium, chloride, and calcium were measured by ionselective electrode. Bicarbonate was assayed by pH electrode. Glucose was measured by oxygen sensor. Blood urea nitrogen was measured by conductivity electrode. Magnesium was measured by calmagite method and phosphorous by phospho-molybdate. Uric acid was measured by enzymatic Trinder. Amylase was measured by
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
803
enzymatic DS. Aspartate aminotransferase and alanine aminotransferase levels were measured by the Henry method. Total protein, albumin, and creatinine were measured using the colorimetric methodology [7]. Total bilirubin levels were assayed by the Jendrassik-Grof methodology [8]. 2.5. Statistical analyses Mean values of physiologic measurements were calculated at each follow-up time. To assess whether there was a statistically significant trend in the means over time, we fit mixed-effects regression models including a random slope and random intercept for each study subject [9]. Mixedeffects models are alternatives to repeated-measures analysis of variance that appropriately account for the repeated measures on the same individuals but allow for varying numbers of observations per person. P values were based on Wald tests that based on these models. These models were fit using the “lme” software library for the R system for statistical computation and graphics. Three subjects had new medications added or medication dosages adjusted during follow-up that might have had an impact on their weight, GlyHb, or cholesterol. To avoid the potential confounding effects of these changes, we did not include the observations from these subjects that occurred after the start of the new medications for some of our analyses. Data are presented as means ± SD. Associations with a P value less than .05 are referred to as statistically significant. Baseline means were calculated by averaging both pretreatment means.
3. Results Eight subjects completed the full 14 months of the study protocol and were included in the efficacy analyses. Of these 8 subjects, 4 were male and 4 female, with a mean age of 50.7 ± 10.9 years. At baseline, study subjects were obese, with a mean body mass index of 36.7 ± 5.1 kg/m2, and in poor glycemic control, with a mean GlyHb of 11.2% ± 2.0%. Subjects were also dyslipidemic at baseline, with a mean total cholesterol of 224 ± 28 mg/dL; LDL cholesterol, 146 ± 3) mg/dL; triglycerides, 226 ± 98 mg/dL; and HDL cholesterol, 31 ± 14 mg/dL. Four subjects discontinued the study soon after D-tag therapy was initiated and were not included in data analyses. Two subjects withdrew during the first week of D-tag because of persistent gastrointestinal symptoms including diarrhea, flatulence and/or bloating. One subject with asthma withdrew after 2 months of daily D-tag because of a persistent, dry cough. The cough resolved after discontinuation of D-tag and recurred after reinitiation of D-tag. The fourth subject withdrew after moving to another state. During the 2-month run-in period, modest increases were observed for both mean body weight (+0.5 kg) and GlyHb (+1.1%). Thereafter, body weight and GlyHb remained
Fig. 1. Mean (±SEM) weight in all 8 treated subjects and in 5 subjects who did not have oral diabetes medication changes during the study. ⁎P = .01 vs baseline; #P = .001 vs baseline.
stable for the first 4 months of D-tag supplementation, and then weight progressively declined (Figs. 1 and 2). Seven of 8 subjects lost weight during the intervention phase, with a mean fall from 109 ± 14.7 kg at baseline to 105.3 ± 14.4 kg at month 12 (P = .01). The single subject who gained weight during the intervention period had been started on a sulfonylurea and had her troglitazone dose increased during the study. When this subject and 2 others who had a sulfonylurea added or dosage increased during the study were excluded from data analysis, a more substantial weight loss of 5.1 ± 3.1 kg during the 12-month intervention period was observed (Fig. 1, P = .001). After 12 months of D-tag, a decrease in GlyHb was observed in the 8 subjects who completed the study from 11.2% ± 2.0% to 9.5% ± 2.0% (Fig. 2). However, after the 3 subjects in whom hypoglycemic medications were added or increased during the study were excluded from analysis, a nonsignificant decrease in GlyHb was seen (10.6% ± 1.9% vs 9.6% ± 2.3%, P = .08). No significant changes were observed for fasting glucose or insulin levels during the intervention phase (data not shown). An increase in HDL cholesterol levels occurred throughout the 12-month intervention period in all 8 subjects (Fig. 3). Mean baseline HDL cholesterol rose from 31.1 ± 13.9 to 40.5 ± 10.4 mg/dL at month 12 (P = .01). After the
Fig. 2. Mean (±SEM) glycohemoglobin in all 8 treated subjects and in 5 subjects who did not have oral diabetes medication changes during the study.
804
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
A 48-year-old man who had experienced a cerebrovascular accident 4 years before study entry had an uncomplicated myocardial infarction 4.5 months into the intervention phase of the study. D-tagatose was not felt to have played a role in his myocardial infarction. He had numerous cardiovascular risk factors before study entry. He was allowed to restart D-tag after his hospitalization, but was excluded thereafter from analyses for lipids and blood pressure, since metoprolol and pravastatin had been added.
Fig. 3. Mean (±SEM) HDL cholesterol in all 8 treated subjects and in 6 subjects who did not have lipid-modifying medications added during the study. ⁎P = .01 vs baseline level; #P b .001 vs baseline level.
subjects in whom pravastatin was added (n = 1) or whose troglitazone dose was increased (n = 1) were excluded from data analysis, an even greater rise in HDL cholesterol occurred, from 30.5 ± 15.8 to 41.7 ± 12.1 mg/dL (P b .001). No changes were observed from baseline to month 12 in total cholesterol (225 ± 30 mg/dL vs 224 ± 54 mg/dL) or triglycerides (222 ± 101 mg/dL vs 246 ± 154 mg/dL) in the 6 subjects who did not have medications added that could affect lipid levels. No change was observed from baseline to month 12 in LDL cholesterol (154 ± 36 mg/dL vs 155 ± 44 mg/dL)) in 5 subjects who did not have medications added that could affect lipid levels or whose LDL cholesterol could not be calculated due to elevated triglycerides. In addition, no significant changes were observed in blood pressure during the 12-month intervention period. Three subjects had cardiovascular medications added during the study, 2 for hypertension (amlodipine and hydrochlorothiazide), and 1 who was normotensive but post–myocardial infarction (metoprolol). No significant changes in blood pressure were observed during the study when these 3 subjects were excluded from analysis. D-tagatose was not found to have toxic effects on renal or hepatic function, measures of which did not change during the 12-month intervention period. No other changes were observed in any of the other biochemical parameters tested during the intervention period (data not shown). During the first 2 weeks of D-tag therapy, gastrointestinal symptoms were reported by all 8 subjects who completed the study and in 10 of 12 subjects who initially enrolled in the study. These were typically mild and transient and included flatulence (n = 6), diarrhea (n = 4), and/or nausea (n = 1). Thereafter, adverse gastrointestinal effects resolved when the D-tag was taken as directed, with the exception of a single subject who reported persistent, mild flatulence during the first 6 months of therapy. Three subjects, including 1 who was on metformin throughout the study, reported bloating, flatulence, and/or diarrhea if more than 15 g of D-tag was consumed at a time or if the doses were spaced less than 3 hours apart. A single subject with chronic constipation noted more regular bowel habits throughout the 12-month treatment period.
4. Discussion This pilot study is the first to investigate the long-term metabolic effects of D-tag in humans with type 2 DM. No adverse metabolic or biochemical effects were noted among the 8 subjects who completed the study. Indeed, significant and beneficial changes were seen in weight and HDL cholesterol at the end of the 12-month intervention period, in the absence of any intentional changes in dietary intake or physical activity. No significant changes in GlyHb or other lipid parameters were seen. The weight loss occurred in 7 of the 8 subjects in the study. Weight loss occurred in one subject in whom a sulfonylurea had been added and in another whose sulfonylurea dosage had been increased. Indeed, the only subject who gained weight during the study did so only after being started on a sulfonylurea and having her thiazolidinedione dosage increased, 2 medications known to be associated with weight gain. When the 5 subjects who did not have changes in oral diabetes medications made during the study were separately evaluated, the mean weight loss observed following 12 months of D-tag was even greater (5.1 ± 3.1 kg vs 3.7 ± 3.4 kg for all 8 subjects). The hypothesis that daily tagatose would lead to weight loss was accepted. The etiology of weight loss that began after 4 months of D-tag in our study population is unclear. Subjects participating in the study did not change their dietary habits other than using D-tag as their sweetener. The 8 subjects were sedentary and did not modify their physical activity during the study. Kruger et al [4] reported less weight gain when rats were fed 15% to 20% of their diet as D-tag for 90 days. In that study, a decrease in mean weekly food consumption was reported in the rats fed 20% of their diet as D-tag. In this study, 3 subjects reported having mild early satiety, especially during the first month of D-tag use. One subject felt that this may have decreased his total daily energy intake throughout the study. Buemann et al similarly reported a perception of fullness in humans treated with 30 g D-tag/d for 2 weeks [10]. Lee and Storey reported appetite loss in humans given 20 g D-tag when compared with sucrose in chocolate [11]. Reduced energy intake has been reported after oral intake of compounds chemically similar to tagatose, including fructose in humans [12] and oligofructose in Wistar rats [13].
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
Two subjects in our study reported a reduction in simple sugar consumption due to substitution with D-tag. The reduction in sugar intake may have been underestimated in our subjects and may have contributed to a decrease in net utilizable energy intake. Other potential mechanisms of weight loss are only speculative. Studies have shown that only 26% of ingested D-tag is absorbed in the distal small bowel of pigs [14], leading to its low net energy content. It has been proposed that the osmotic effect of malabsorbed D-tag speeds its transit time through the gastrointestinal tract and that this may impair the absorption of other macronutrients [15]. The digestibility of sucrose after oral D-tag was reduced by 7.6% in pig small intestine [16]. A progressive and marked rise in HDL cholesterol was observed following dietary supplementation with D-tag. The hypothesis that daily tagatose would raise HDL cholesterol was accepted. The low-baseline HDL cholesterol in our population is seen commonly in obese patients with type 2 DM and has been shown to be a major risk factor for coronary heart disease (CHD) [17-20]. An aggregate analysis of 4 large epidemiologic studies found that for each 1-mg/dL increase in HDL cholesterol, one would expect a 2% decrease in CHD risk in men and a 3% decrease in CHD risk in women [21]. Reductions in body weight likely contributed to the improvement in HDL cholesterol in this study. Kaplan et al showed that energy restriction for 3 months in obese patients with type 2 DM leading to a 3-kg weight loss was associated with a 5.5-mg/dL increase in HDL cholesterol [22]. Wing et al [23] found a more modest 2.5-mg/dL increase in HDL cholesterol in obese type 2 DM patients only in those with weight loss exceeding 6.9 kg, following 1 year of energy restriction and increased exercise. Our subjects did not change their amount of physical activity or alcohol intake, and none discontinued cigarettes, all parameters known to affect HDL cholesterol. It is unknown whether the small amounts of D-tag that enter into the circulation, although known to be extensively metabolized in the liver, affect hepatic synthesis of HDL cholesterol. No changes in LDL cholesterol or triglycerides were seen, and the hypothesis that these parameters would be reduced is rejected. Interpreting the beneficial effects of D-tag on weight and HDL cholesterol is made difficult by the inherent limitations of a small, nonplacebo-controlled study. Adjustments of medications having metabolic effects by outside physicians during the study period also led to subsequent lipid and weight measurements being excluded from data analysis. There is no evidence in this study of a D-tag–associated deterioration of diabetic control. Among the 5 subjects who received 12 months of D-tag and had no oral diabetes medications added during the study, a nonsignificant 1% decrease in GlyHb was seen. The hypothesis that daily tagatose would lead to a reduction in GlyHb is rejected. Oral administration of D-tag has previously been shown to blunt hyperglycemia significantly following oral glucose in a dose-dependent manner in
805
subjects with type 2 DM, with doses as low as 10 g being efficacious [3]. Subjects in this study received 15 g of D-tag with meals for 12 months. The 1.1% increase in GlyHb that occurred during the 2-month run-in period demonstrates that study subjects were in a state of deteriorating glycemic control when they started D-tag. This may help explain why a significant decline in GlyHb was not seen during the study period. A larger, placebo-controlled study is needed to determine whether D-tag use in type 2 DM patients improves glycemic control, and if so, to what degree better glycemic control correlates with weight loss. A recent study in hypercholesterolemic mice showed that equivalent substitution of D-tag when compared with sucrose as dietary carbohydrate led to less obesity, hyperglycemia, hyperlipidemia, and atherosclerosis [24]. The 15 g of D-tag given with meals led to adverse gastrointestinal effects in 10 of 12 of the subjects initially enrolled into the trial. These symptoms were mostly transient and mild in severity. In 2 subjects, the effects were poorly tolerated and led to withdrawal from the study within 1 week. Poorly absorbed D-tag is thought to osmotically draw water into the colon, thereby softening the stool. The transient nature of most of the adverse gastrointestinal effects in our subjects may be due to adaptation to D-tag by colonic bacteria. Sprague-Dawley rats adapted to D-tag for 28 days were found to have progressively less frequent soft stools over time [25]. Pigs fed D-tag over a 15-day period were found to harbor a significantly greater number of D-tag– degrading colonic bacteria compared with pigs fed a control diet [15]. Future studies investigating the metabolic benefits of D-tag would benefit from an adaptation period using lower and more graded doses of D-tag to help minimize early, adverse gastrointestinal effects. In conclusion, this pilot study found that daily ingestion of D-tag with food in subjects with type 2 diabetes leads to weight loss and improvements in HDL cholesterol. These beneficial effects need to be confirmed in larger, placebocontrolled studies that are currently underway. The rise in HDL cholesterol appears to be disproportionate to the degree of weight loss observed during the study. No adverse effects on glycemic control or other biochemical parameters were seen. If beneficial effects of D-tag on weight and HDL cholesterol are confirmed in patients with type 2 DM, treatment with D-tag could help reduce cardiovascular disease in this high-risk population.
Acknowledgment We would like to thank Maryland Industrial Partnerships, Glenn L. Martin Institute of Technology, University of Maryland, College Park, and Spherix Incorporated (Beltsville, Md) for their financial support of this study. We are most grateful to Debra Ostrowski for her expert technical assistance and to Alan Shuldiner for his critical review of the manuscript.
806
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
References [1] Livesey G, Brown JC. D-Tagatose is a bulk sweetener with zero energy determined in rats. J Nutr 1996;126:1601-9. [2] Levin GV, Zehner LR, Saunders JP, Beadle JR. Sugar substitutes: their energy values, bulk characteristics and potential health benefits. Am J Clin Nutr 1995;62(Suppl):1161S-8S. [3] Donner TW, Wilber JF, Ostrowski D. D-Tagatose, a novel hexose: acute effects on carbohydrate tolerance in subjects with and without type 2 diabetes. Diab Obes Metab 1999;1:285-92. [4] Kruger CL, Whittaker MH, Frankos VH, Trimmer GW. 90-day oral toxicity study of D-tagatose in rats. Reg Tox Pharm 1999;29:S1-10. [5] Buemann B, Toubro S, Raben A, Blundell J, Astrup A. The acute effects of D-tagatose on food intake in human subjects. Br J Nutr 2000; 84:227-31. [6] Jeppson J, Approved IFCC. Reference method for the measurement of HbA1c in human blood. Clin Chem Lab Med 2002;40(1):78-89. [7] Tietz NW. Clinical guide to laboratory tests. 3rd ed. Philadelphia: WB Saunders; 1995. [8] Jendrassik L, Grof P. Biochem Z 1937;297:81. [9] Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics 1982;38:963-74. [10] Buemann B, Toubro S, Astrup A. D-Tagatose, a stereoisomer of D-fructose, increases hydrogen production in humans without affecting 24-hour energy expenditure or respiratory exchange ratio. J Nutr 1998; 128:1481-6. [11] Lee A, Storey DM. Comparative gastrointestinal tolerance of sucrose, lactitol, or D-tagatose in chocolate. Reg Tox Pharm 1999;29:S78-82. [12] Heacock P, Hertzler S, Wolf B. Fructose prefeeding reduces the glycemic response to a high-glycemic index, starchy food in humans. J Nutr 2002;132:2601-4. [13] Delzenne NM, Cani PD, Daubioul C, Neyrinck AM. Impact of inulin and oligofructose on gastrointestinal peptides. Br J Nutr 2005;93 (Suppl 1):S157-61.
[14] Johansen HN, Jensen BB. Recovery of energy as SCFA after microbial fermentation of D-tagatose. Int J Obes 1997;21(Suppl 2):550. [15] Jenkin AP, Menzies IS, Nukajam WS, Creamer B. The effect of ingested lactulose on absorption of L-rhamnose, D-xylose, and 3-Omethyl- D -glucose in subjects with ileostomies. Scand J Gastroenterol 1994;29:820-5. [16] Laerke HN, Jensen BB: D-Tagatose has low small intestinal digestibility but high large intestinal fermentability in pigs. J Nutr 1999; 129:1002-1009. 0363-6119 [17] Barter PJ, Rye KA. High density lipoproteins and coronary heart disease. Atherosclerosis 1996;121:1–12. [18] Castelli WP, Garrison RJ, Wilson PW, et al. Incidence of coronary heart disease and lipoprotein cholesterol levels. JAMA 1986;256: 2835-8. [19] Jacobs DR, Mebane IL, Bangdiwala SI, et al. High density lipoprotein cholesterol as a predictor of cardiovascular disease mortality in men and women: the follow-up study of the Lipid Research Clinics Prevalence Study. Am J Epidemiol 1990;131:32-47. [20] NIH Consensus Conference: Triglyceride, high density lipoprotein, and coronary heart disease. JAMA 1993;269:505-10. [21] Gordon DJ, Probstfield JL, Garrison RJ, et al. High density lipoprotein cholesterol and cardiovascular disease. Circulation 1989;79:8–15. [22] Kaplan RM, Wilson DK, Hartwell SL, Merino KL, Wallace JP. Prospective evaluation of HDL cholesterol changes after diet and physical conditioning programs for patients with type II diabetes mellitus. Diabetes Care 1985;8:343-8. [23] Wing RR, Koeske R, Epstein LH, Nowalk MP, Gooding W, Becker D. Long-term effects of modest weight loss in type II diabetic patients. Arch Intern Med 1987;147:1749-53. [24] Police SB, Harris JC, Lodder RA, Cassis LA. Effects of diets containing sucrose vs. D-tagatose in hypercholesterolemic mice. Obesity 2009;17(2):269-75. [25] Saunders JP, Zehner LR, Levin GV. Disposition of D-[U-14C]tagatose in the rat. Reg Tox Pharm 1999;29:S46-56.
Nutrition Research 30 (2010) 801 – 806 www.nrjournal.com
Dietary supplementation with D-tagatose in subjects with type 2 diabetes leads to weight loss and raises high-density lipoprotein cholesterol☆ Thomas W. Donner a,⁎, Laurence S. Magder b , Kiarash Zarbalian a a
Department of Internal Medicine, University of Maryland School of Medicine, Baltimore, MD 21201, USA b Department of Epidemiology, University of Maryland School of Medicine, Baltimore, MD 21201, USA Received 19 May 2010; revised 11 September 2010; accepted 13 September 2010
Abstract Oral D-tagatose (D-tag) attenuates the rise in plasma glucose during an oral glucose tolerance test in subjects with type 2 diabetes mellitus (DM) and reduces food intake in healthy human subjects. A reduction in food consumption and less weight gain occur in rats fed tagatose. This pilot study explored the metabolic effects of D-tag given daily to 8 human subjects with type 2 DM for 1 year. We hypothesized that this treatment period would lead to weight loss and improvements in glycated hemoglobin and the lipid profile. A 2-month run-in period was followed by a 12-month treatment period when 15 g of oral D-tag was taken 3 times daily with food. No serious adverse effects were seen during the 12-month treatment period. Ten of the initially12 recruited subjects experienced gastrointestinal side effects that tended to be mild and transient. When 3 subjects were excluded who had oral diabetes, medications added and/or dosages increased during the study and mean (SD) body weight declined from 108.4 (9.0) to 103.3 (7.3) kg (P = .001). Glycated hemoglobin fell nonsignificantly from 10.6% ± 1.9% to 9.6% ± 2.3% (P = .08). High-density lipoprotein cholesterol progressively rose from a baseline level of 30.5 ± 15.8 to 41.7 ± 12.1 mg/dL at month 12 in the 6 subjects who did not have lipid-modifying medications added during the study (P b .001). Significant improvements in body weight and high-density lipoprotein cholesterol in this pilot study suggest that D-tag may be a potentially useful adjunct in the management of patients with type 2 DM. © 2010 Elsevier Inc. All rights reserved. Keywords: Abbreviations:
Tagatose; Diabetes control; Weight loss agents; HDL cholesterol; Lipid modification; Bulk sweeteners CHD, coronary heart disease; DM, diabetes mellitus; D-Tag, D-tagatose; GlyHb, glycated hemoglobin; LDL, low-density lipoprotein.
1. Introduction Nonnutritive sweeteners are used commonly to help patients with type 2 diabetes mellitus (DM) reduce carbohydrate energy intake to affect better glycemic control
☆
Financial support was received from Maryland Industrial Partnerships, Glenn L. Martin Institute of Technology, University of Maryland, College Park, and Spherix Inc, Beltsville, Md. ⁎ Corresponding author. Division of Endocrinology and Metabolism, Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA. Tel.: +1 410 955 2908; fax: +1 410 614 9586. E-mail address: [email protected] (T.W. Donner). 0271-5317/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2010.09.007
and to assist with weight management. D-Tagatose (D-tag) is a ketohexose bulk sweetener with no net metabolizable energy when fed to rats [1,2]. D-tagatose received Generally Recognized as Safe status by the Food and Drug Administration in 2001 and entered the US market as a sweetener in 2003. In subjects with and without type 2 DM, a 75-g oral D-tag tolerance test led to no changes in plasma glucose or insulin levels. Pretreatment of type 2 DM subjects with doses of D-tag ranging from 10 to 75 g attenuated the rise in plasma glucose during an oral GTT in a dose-dependent manner [3]. These findings suggest that D-tag could potentially improve glycemic control in patients with type 2 DM by blunting postprandial hyperglycemia.
802
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
The long-term metabolic effects of D-tag have not previously been studied in humans. In a 90-day study in rats, D-tag given as 15% to 20% of daily energy intake led to less weight gain when compared with a control diet [4]. Reduced food intake has been observed in rats fed tagatose as 20% of daily energy intake [4] and in humans fed a single 29-g dose of tagatose when compared with 29 g of sucrose added to a breakfast meal [5]. If tagatose causes sustained reductions in energy intake, beneficial effects on weight, blood glucose control, and serum lipid profile could be seen. The objective of this pilot study was to investigate the metabolic effects of long-term administration of D-tag in obese subjects with type 2 DM. Following a 2-month run-in period, subjects had tagatose added to their daily meals and were followed on an every 2-month basis for 12 months for changes in weight, blood glucose control, and lipid profile. We hypothesized that daily oral tagatose would lower glycated hemoglobin (GlyHb) and lead to weight loss and improvements in the lipid profile, including lower low-density lipoprotein (LDL) cholesterol and triglycerides, and higher high-density lipoprotein (HDL) cholesterol.
2. Methods and materials 2.1. Subjects Eight subjects (4 male and 4 female) with type 2 DM for at least 1 year in duration were enrolled and ranged from 35 to 70 years in age. Four of the 12 initially screened subjects were excluded from analysis because they did not complete the study. Data analysis was performed on the 8 subjects who completed all 14 months of the study. Subjects were excluded from study entry if baseline GlyHb levels were lower than 8% (reference range, 4.4%-7.7%) or if they had gastrointestinal disorders. All investigational protocols were approved by the institutional review board of the University of Maryland for human subjects. Participants were enrolled only after giving written informed consent. At entry, 3 subjects were receiving diet therapy for diabetes, 4 were being treated with a sulfonylurea, and 1 was being treated with combination metformin and troglitazone. Those receiving medications had been on stable doses for at least 10 months before enrollment. Three subjects had oral diabetes medications added or adjusted by their primary care providers during the intervention phase of the study. These 3 subjects were included in all outcomes analyses, though subgroup analyses excluding these subjects were done for GlyHb and body weight, which are known to be affected by sulfonylureas and troglitazone. One subject had glipizide 5 mg added during month 4, and another subject had her glyburide dosage increased during month 8 of the intervention phase. The third subject had glyburide 2.5 mg added during month 3 and had the troglitazone dose increased from 200 to 400 mg daily during month 4 of the intervention phase. One subject had pravastatin added during month 8. A lipid subgroup analysis
was performed, which excluded this subject and the one whose troglitazone was increased during the study. A subgroup analysis was also performed on 3 subjects who had antihypertensive medications added during the study. 2.2. Materials D -Tagatose was prepared by Spherix Incorporated (Beltsville, Md). Samples provided to subjects were more than 99% pure by high-performance liquid chromatography analysis and were weighed and packaged in the University of Maryland Pharmacy before administration.
2.3. Study protocol After a 2-month run-in period, subjects were given 15-g packages of D -tag to be taken 3 times daily with nonstandardized meals. Dose-dependent, gastrointestinal side effects observed with large doses of D-tag have been attributed to an osmotic effect of this poorly absorbed sugar. Prior tolerance testing at the University of Maryland has shown the 15-g dosage to be typically well tolerated with minimal adverse gastrointestinal effects [3]. The D-tag was dissolved in liquids, used in baking, or added to prepared foods. For the duration of the 14-month study period, subjects were encouraged not to otherwise alter their dietary intake. All subjects enrolled were physically inactive and remained so throughout the 14-month observation period. At the initial visit and every 2 months for the next 14 months, body weight and vital signs were recorded, and fasting blood was sent for glucose, GlyHb, insulin, lipids, liver and kidney function, uric acid, bilirubin, phosphorus, calcium, magnesium, bicarbonate, chloride, sodium, potassium, total protein, albumin, and amylase. Subjects were questioned at each visit about the occurrence of any adverse events or side effects, and specifically about any gastrointestinal side effects. Compliance with D-tag was confirmed based on D-tag package counts performed every 2 months. 2.4. Laboratory evaluations Glycosylated hemoglobin values were measured by the affinity column method (Helena Glyco-Tek, Beaumont, Tex), which has a reference range of 4.4% to 7.7% and a mean intra-assay coefficient of variation of 2.95% [6]. Insulin levels were measured by the Coates-A-count RIA method (Diagnostics Products Corporation, Los Angeles, Calif). Chemistry profiles and lipids were performed at the University of Maryland Chemistry Laboratory using Beckman Coulter laboratory analyzers (Brea, Calif). Sodium, potassium, chloride, and calcium were measured by ionselective electrode. Bicarbonate was assayed by pH electrode. Glucose was measured by oxygen sensor. Blood urea nitrogen was measured by conductivity electrode. Magnesium was measured by calmagite method and phosphorous by phospho-molybdate. Uric acid was measured by enzymatic Trinder. Amylase was measured by
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
803
enzymatic DS. Aspartate aminotransferase and alanine aminotransferase levels were measured by the Henry method. Total protein, albumin, and creatinine were measured using the colorimetric methodology [7]. Total bilirubin levels were assayed by the Jendrassik-Grof methodology [8]. 2.5. Statistical analyses Mean values of physiologic measurements were calculated at each follow-up time. To assess whether there was a statistically significant trend in the means over time, we fit mixed-effects regression models including a random slope and random intercept for each study subject [9]. Mixedeffects models are alternatives to repeated-measures analysis of variance that appropriately account for the repeated measures on the same individuals but allow for varying numbers of observations per person. P values were based on Wald tests that based on these models. These models were fit using the “lme” software library for the R system for statistical computation and graphics. Three subjects had new medications added or medication dosages adjusted during follow-up that might have had an impact on their weight, GlyHb, or cholesterol. To avoid the potential confounding effects of these changes, we did not include the observations from these subjects that occurred after the start of the new medications for some of our analyses. Data are presented as means ± SD. Associations with a P value less than .05 are referred to as statistically significant. Baseline means were calculated by averaging both pretreatment means.
3. Results Eight subjects completed the full 14 months of the study protocol and were included in the efficacy analyses. Of these 8 subjects, 4 were male and 4 female, with a mean age of 50.7 ± 10.9 years. At baseline, study subjects were obese, with a mean body mass index of 36.7 ± 5.1 kg/m2, and in poor glycemic control, with a mean GlyHb of 11.2% ± 2.0%. Subjects were also dyslipidemic at baseline, with a mean total cholesterol of 224 ± 28 mg/dL; LDL cholesterol, 146 ± 3) mg/dL; triglycerides, 226 ± 98 mg/dL; and HDL cholesterol, 31 ± 14 mg/dL. Four subjects discontinued the study soon after D-tag therapy was initiated and were not included in data analyses. Two subjects withdrew during the first week of D-tag because of persistent gastrointestinal symptoms including diarrhea, flatulence and/or bloating. One subject with asthma withdrew after 2 months of daily D-tag because of a persistent, dry cough. The cough resolved after discontinuation of D-tag and recurred after reinitiation of D-tag. The fourth subject withdrew after moving to another state. During the 2-month run-in period, modest increases were observed for both mean body weight (+0.5 kg) and GlyHb (+1.1%). Thereafter, body weight and GlyHb remained
Fig. 1. Mean (±SEM) weight in all 8 treated subjects and in 5 subjects who did not have oral diabetes medication changes during the study. ⁎P = .01 vs baseline; #P = .001 vs baseline.
stable for the first 4 months of D-tag supplementation, and then weight progressively declined (Figs. 1 and 2). Seven of 8 subjects lost weight during the intervention phase, with a mean fall from 109 ± 14.7 kg at baseline to 105.3 ± 14.4 kg at month 12 (P = .01). The single subject who gained weight during the intervention period had been started on a sulfonylurea and had her troglitazone dose increased during the study. When this subject and 2 others who had a sulfonylurea added or dosage increased during the study were excluded from data analysis, a more substantial weight loss of 5.1 ± 3.1 kg during the 12-month intervention period was observed (Fig. 1, P = .001). After 12 months of D-tag, a decrease in GlyHb was observed in the 8 subjects who completed the study from 11.2% ± 2.0% to 9.5% ± 2.0% (Fig. 2). However, after the 3 subjects in whom hypoglycemic medications were added or increased during the study were excluded from analysis, a nonsignificant decrease in GlyHb was seen (10.6% ± 1.9% vs 9.6% ± 2.3%, P = .08). No significant changes were observed for fasting glucose or insulin levels during the intervention phase (data not shown). An increase in HDL cholesterol levels occurred throughout the 12-month intervention period in all 8 subjects (Fig. 3). Mean baseline HDL cholesterol rose from 31.1 ± 13.9 to 40.5 ± 10.4 mg/dL at month 12 (P = .01). After the
Fig. 2. Mean (±SEM) glycohemoglobin in all 8 treated subjects and in 5 subjects who did not have oral diabetes medication changes during the study.
804
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
A 48-year-old man who had experienced a cerebrovascular accident 4 years before study entry had an uncomplicated myocardial infarction 4.5 months into the intervention phase of the study. D-tagatose was not felt to have played a role in his myocardial infarction. He had numerous cardiovascular risk factors before study entry. He was allowed to restart D-tag after his hospitalization, but was excluded thereafter from analyses for lipids and blood pressure, since metoprolol and pravastatin had been added.
Fig. 3. Mean (±SEM) HDL cholesterol in all 8 treated subjects and in 6 subjects who did not have lipid-modifying medications added during the study. ⁎P = .01 vs baseline level; #P b .001 vs baseline level.
subjects in whom pravastatin was added (n = 1) or whose troglitazone dose was increased (n = 1) were excluded from data analysis, an even greater rise in HDL cholesterol occurred, from 30.5 ± 15.8 to 41.7 ± 12.1 mg/dL (P b .001). No changes were observed from baseline to month 12 in total cholesterol (225 ± 30 mg/dL vs 224 ± 54 mg/dL) or triglycerides (222 ± 101 mg/dL vs 246 ± 154 mg/dL) in the 6 subjects who did not have medications added that could affect lipid levels. No change was observed from baseline to month 12 in LDL cholesterol (154 ± 36 mg/dL vs 155 ± 44 mg/dL)) in 5 subjects who did not have medications added that could affect lipid levels or whose LDL cholesterol could not be calculated due to elevated triglycerides. In addition, no significant changes were observed in blood pressure during the 12-month intervention period. Three subjects had cardiovascular medications added during the study, 2 for hypertension (amlodipine and hydrochlorothiazide), and 1 who was normotensive but post–myocardial infarction (metoprolol). No significant changes in blood pressure were observed during the study when these 3 subjects were excluded from analysis. D-tagatose was not found to have toxic effects on renal or hepatic function, measures of which did not change during the 12-month intervention period. No other changes were observed in any of the other biochemical parameters tested during the intervention period (data not shown). During the first 2 weeks of D-tag therapy, gastrointestinal symptoms were reported by all 8 subjects who completed the study and in 10 of 12 subjects who initially enrolled in the study. These were typically mild and transient and included flatulence (n = 6), diarrhea (n = 4), and/or nausea (n = 1). Thereafter, adverse gastrointestinal effects resolved when the D-tag was taken as directed, with the exception of a single subject who reported persistent, mild flatulence during the first 6 months of therapy. Three subjects, including 1 who was on metformin throughout the study, reported bloating, flatulence, and/or diarrhea if more than 15 g of D-tag was consumed at a time or if the doses were spaced less than 3 hours apart. A single subject with chronic constipation noted more regular bowel habits throughout the 12-month treatment period.
4. Discussion This pilot study is the first to investigate the long-term metabolic effects of D-tag in humans with type 2 DM. No adverse metabolic or biochemical effects were noted among the 8 subjects who completed the study. Indeed, significant and beneficial changes were seen in weight and HDL cholesterol at the end of the 12-month intervention period, in the absence of any intentional changes in dietary intake or physical activity. No significant changes in GlyHb or other lipid parameters were seen. The weight loss occurred in 7 of the 8 subjects in the study. Weight loss occurred in one subject in whom a sulfonylurea had been added and in another whose sulfonylurea dosage had been increased. Indeed, the only subject who gained weight during the study did so only after being started on a sulfonylurea and having her thiazolidinedione dosage increased, 2 medications known to be associated with weight gain. When the 5 subjects who did not have changes in oral diabetes medications made during the study were separately evaluated, the mean weight loss observed following 12 months of D-tag was even greater (5.1 ± 3.1 kg vs 3.7 ± 3.4 kg for all 8 subjects). The hypothesis that daily tagatose would lead to weight loss was accepted. The etiology of weight loss that began after 4 months of D-tag in our study population is unclear. Subjects participating in the study did not change their dietary habits other than using D-tag as their sweetener. The 8 subjects were sedentary and did not modify their physical activity during the study. Kruger et al [4] reported less weight gain when rats were fed 15% to 20% of their diet as D-tag for 90 days. In that study, a decrease in mean weekly food consumption was reported in the rats fed 20% of their diet as D-tag. In this study, 3 subjects reported having mild early satiety, especially during the first month of D-tag use. One subject felt that this may have decreased his total daily energy intake throughout the study. Buemann et al similarly reported a perception of fullness in humans treated with 30 g D-tag/d for 2 weeks [10]. Lee and Storey reported appetite loss in humans given 20 g D-tag when compared with sucrose in chocolate [11]. Reduced energy intake has been reported after oral intake of compounds chemically similar to tagatose, including fructose in humans [12] and oligofructose in Wistar rats [13].
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
Two subjects in our study reported a reduction in simple sugar consumption due to substitution with D-tag. The reduction in sugar intake may have been underestimated in our subjects and may have contributed to a decrease in net utilizable energy intake. Other potential mechanisms of weight loss are only speculative. Studies have shown that only 26% of ingested D-tag is absorbed in the distal small bowel of pigs [14], leading to its low net energy content. It has been proposed that the osmotic effect of malabsorbed D-tag speeds its transit time through the gastrointestinal tract and that this may impair the absorption of other macronutrients [15]. The digestibility of sucrose after oral D-tag was reduced by 7.6% in pig small intestine [16]. A progressive and marked rise in HDL cholesterol was observed following dietary supplementation with D-tag. The hypothesis that daily tagatose would raise HDL cholesterol was accepted. The low-baseline HDL cholesterol in our population is seen commonly in obese patients with type 2 DM and has been shown to be a major risk factor for coronary heart disease (CHD) [17-20]. An aggregate analysis of 4 large epidemiologic studies found that for each 1-mg/dL increase in HDL cholesterol, one would expect a 2% decrease in CHD risk in men and a 3% decrease in CHD risk in women [21]. Reductions in body weight likely contributed to the improvement in HDL cholesterol in this study. Kaplan et al showed that energy restriction for 3 months in obese patients with type 2 DM leading to a 3-kg weight loss was associated with a 5.5-mg/dL increase in HDL cholesterol [22]. Wing et al [23] found a more modest 2.5-mg/dL increase in HDL cholesterol in obese type 2 DM patients only in those with weight loss exceeding 6.9 kg, following 1 year of energy restriction and increased exercise. Our subjects did not change their amount of physical activity or alcohol intake, and none discontinued cigarettes, all parameters known to affect HDL cholesterol. It is unknown whether the small amounts of D-tag that enter into the circulation, although known to be extensively metabolized in the liver, affect hepatic synthesis of HDL cholesterol. No changes in LDL cholesterol or triglycerides were seen, and the hypothesis that these parameters would be reduced is rejected. Interpreting the beneficial effects of D-tag on weight and HDL cholesterol is made difficult by the inherent limitations of a small, nonplacebo-controlled study. Adjustments of medications having metabolic effects by outside physicians during the study period also led to subsequent lipid and weight measurements being excluded from data analysis. There is no evidence in this study of a D-tag–associated deterioration of diabetic control. Among the 5 subjects who received 12 months of D-tag and had no oral diabetes medications added during the study, a nonsignificant 1% decrease in GlyHb was seen. The hypothesis that daily tagatose would lead to a reduction in GlyHb is rejected. Oral administration of D-tag has previously been shown to blunt hyperglycemia significantly following oral glucose in a dose-dependent manner in
805
subjects with type 2 DM, with doses as low as 10 g being efficacious [3]. Subjects in this study received 15 g of D-tag with meals for 12 months. The 1.1% increase in GlyHb that occurred during the 2-month run-in period demonstrates that study subjects were in a state of deteriorating glycemic control when they started D-tag. This may help explain why a significant decline in GlyHb was not seen during the study period. A larger, placebo-controlled study is needed to determine whether D-tag use in type 2 DM patients improves glycemic control, and if so, to what degree better glycemic control correlates with weight loss. A recent study in hypercholesterolemic mice showed that equivalent substitution of D-tag when compared with sucrose as dietary carbohydrate led to less obesity, hyperglycemia, hyperlipidemia, and atherosclerosis [24]. The 15 g of D-tag given with meals led to adverse gastrointestinal effects in 10 of 12 of the subjects initially enrolled into the trial. These symptoms were mostly transient and mild in severity. In 2 subjects, the effects were poorly tolerated and led to withdrawal from the study within 1 week. Poorly absorbed D-tag is thought to osmotically draw water into the colon, thereby softening the stool. The transient nature of most of the adverse gastrointestinal effects in our subjects may be due to adaptation to D-tag by colonic bacteria. Sprague-Dawley rats adapted to D-tag for 28 days were found to have progressively less frequent soft stools over time [25]. Pigs fed D-tag over a 15-day period were found to harbor a significantly greater number of D-tag– degrading colonic bacteria compared with pigs fed a control diet [15]. Future studies investigating the metabolic benefits of D-tag would benefit from an adaptation period using lower and more graded doses of D-tag to help minimize early, adverse gastrointestinal effects. In conclusion, this pilot study found that daily ingestion of D-tag with food in subjects with type 2 diabetes leads to weight loss and improvements in HDL cholesterol. These beneficial effects need to be confirmed in larger, placebocontrolled studies that are currently underway. The rise in HDL cholesterol appears to be disproportionate to the degree of weight loss observed during the study. No adverse effects on glycemic control or other biochemical parameters were seen. If beneficial effects of D-tag on weight and HDL cholesterol are confirmed in patients with type 2 DM, treatment with D-tag could help reduce cardiovascular disease in this high-risk population.
Acknowledgment We would like to thank Maryland Industrial Partnerships, Glenn L. Martin Institute of Technology, University of Maryland, College Park, and Spherix Incorporated (Beltsville, Md) for their financial support of this study. We are most grateful to Debra Ostrowski for her expert technical assistance and to Alan Shuldiner for his critical review of the manuscript.
806
T.W. Donner et al. / Nutrition Research 30 (2010) 801–806
References [1] Livesey G, Brown JC. D-Tagatose is a bulk sweetener with zero energy determined in rats. J Nutr 1996;126:1601-9. [2] Levin GV, Zehner LR, Saunders JP, Beadle JR. Sugar substitutes: their energy values, bulk characteristics and potential health benefits. Am J Clin Nutr 1995;62(Suppl):1161S-8S. [3] Donner TW, Wilber JF, Ostrowski D. D-Tagatose, a novel hexose: acute effects on carbohydrate tolerance in subjects with and without type 2 diabetes. Diab Obes Metab 1999;1:285-92. [4] Kruger CL, Whittaker MH, Frankos VH, Trimmer GW. 90-day oral toxicity study of D-tagatose in rats. Reg Tox Pharm 1999;29:S1-10. [5] Buemann B, Toubro S, Raben A, Blundell J, Astrup A. The acute effects of D-tagatose on food intake in human subjects. Br J Nutr 2000; 84:227-31. [6] Jeppson J, Approved IFCC. Reference method for the measurement of HbA1c in human blood. Clin Chem Lab Med 2002;40(1):78-89. [7] Tietz NW. Clinical guide to laboratory tests. 3rd ed. Philadelphia: WB Saunders; 1995. [8] Jendrassik L, Grof P. Biochem Z 1937;297:81. [9] Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics 1982;38:963-74. [10] Buemann B, Toubro S, Astrup A. D-Tagatose, a stereoisomer of D-fructose, increases hydrogen production in humans without affecting 24-hour energy expenditure or respiratory exchange ratio. J Nutr 1998; 128:1481-6. [11] Lee A, Storey DM. Comparative gastrointestinal tolerance of sucrose, lactitol, or D-tagatose in chocolate. Reg Tox Pharm 1999;29:S78-82. [12] Heacock P, Hertzler S, Wolf B. Fructose prefeeding reduces the glycemic response to a high-glycemic index, starchy food in humans. J Nutr 2002;132:2601-4. [13] Delzenne NM, Cani PD, Daubioul C, Neyrinck AM. Impact of inulin and oligofructose on gastrointestinal peptides. Br J Nutr 2005;93 (Suppl 1):S157-61.
[14] Johansen HN, Jensen BB. Recovery of energy as SCFA after microbial fermentation of D-tagatose. Int J Obes 1997;21(Suppl 2):550. [15] Jenkin AP, Menzies IS, Nukajam WS, Creamer B. The effect of ingested lactulose on absorption of L-rhamnose, D-xylose, and 3-Omethyl- D -glucose in subjects with ileostomies. Scand J Gastroenterol 1994;29:820-5. [16] Laerke HN, Jensen BB: D-Tagatose has low small intestinal digestibility but high large intestinal fermentability in pigs. J Nutr 1999; 129:1002-1009. 0363-6119 [17] Barter PJ, Rye KA. High density lipoproteins and coronary heart disease. Atherosclerosis 1996;121:1–12. [18] Castelli WP, Garrison RJ, Wilson PW, et al. Incidence of coronary heart disease and lipoprotein cholesterol levels. JAMA 1986;256: 2835-8. [19] Jacobs DR, Mebane IL, Bangdiwala SI, et al. High density lipoprotein cholesterol as a predictor of cardiovascular disease mortality in men and women: the follow-up study of the Lipid Research Clinics Prevalence Study. Am J Epidemiol 1990;131:32-47. [20] NIH Consensus Conference: Triglyceride, high density lipoprotein, and coronary heart disease. JAMA 1993;269:505-10. [21] Gordon DJ, Probstfield JL, Garrison RJ, et al. High density lipoprotein cholesterol and cardiovascular disease. Circulation 1989;79:8–15. [22] Kaplan RM, Wilson DK, Hartwell SL, Merino KL, Wallace JP. Prospective evaluation of HDL cholesterol changes after diet and physical conditioning programs for patients with type II diabetes mellitus. Diabetes Care 1985;8:343-8. [23] Wing RR, Koeske R, Epstein LH, Nowalk MP, Gooding W, Becker D. Long-term effects of modest weight loss in type II diabetic patients. Arch Intern Med 1987;147:1749-53. [24] Police SB, Harris JC, Lodder RA, Cassis LA. Effects of diets containing sucrose vs. D-tagatose in hypercholesterolemic mice. Obesity 2009;17(2):269-75. [25] Saunders JP, Zehner LR, Levin GV. Disposition of D-[U-14C]tagatose in the rat. Reg Tox Pharm 1999;29:S46-56.