Impact of Sugar Beet Pectin and Polydextrose on Fasting And

ORIGINAL ARTICLE Impact of sugar beet pectin and polydextrose on fasting and postprandial glycemia and fasting concentrations of serum total and lipoprotein lipids in middle-aged subjects with abnormal glucose metabolism

U Schwab1, A Louheranta2,ATo¨rro¨nen3 and M Uusitupa2

1Department of Clinical Nutrition, University of Kuopio and Kuopio University Hospital, Kuopio, Finland; 2Department of Clinical Nutrition, University of Kuopio, Kuopio, Finland and 3Danisco Ltd, Espoo, Finland

Objective: To examine the effects of sugar beet pectin (SBP) and polydextrose (PDX) on fasting plasma glucose concentration, serum lipid profile and postprandial glycemia in middle-aged subjects with abnormal glucose metabolism. Design: A placebo-controlled, randomized, parallel double-blinded study. Subjects and setting: Subjects were recruited via newspaper announcements. Seventy subjects were recruited of which 66 completed the study. Intervention: The intervention period lasted for 12 weeks during which the subjects consumed a drink enriched with either SBP (n ¼ 22) or PDX (n ¼ 22) or without fiber enrichment (control group, n ¼ 22). The daily dose of the drinks was 4 dl. The subjects were also given nutrition counseling. Postprandial glycemia was examined in 24 subjects (n ¼ 8 in each group) at 0 and 12 weeks after a standardized breakfast. Results: Fasting plasma glucose concentration did not change in the SBP and PDX groups, whereas it increased in the control group (P ¼ 0.007). On the contrary, the glycosylated hemoglobin A1c increased marginally but significantly (Pp0.05) in the intervention groups without a change in the control group. In postprandial glycemia, no differences between the groups were found. In both the SBP and PDX groups, fasting serum High-density lipoprotein (HDL)-cholesterol concentration increased (Pp0.05) without a change in the control group. Total to HDL-cholesterol ratio decreased in all groups (Pp0.05). Conclusions: It was found that SBP and PDX do not have positive effects on fasting or postprandial plasma glucose concentrations or serum lipid profile in subjects with abnormal glucose metabolism. Sponsorship: Danisco Ltd, Espoo, Finland. European Journal of Clinical Nutrition (2006) 60, 1073–1080. doi:10.1038/sj.ejcn.1602421; published online 8 March 2006

Keywords: diet; fiber; glucose; human; lipid; pectin

Introduction

Correspondence: Dr U Schwab, Department of Clinical Nutrition, University of Type 2 diabetes mellitus (DM2) is becoming to one of the Kuopio, PO Box 1627, 70211 Kuopio, Finland. major causes of clinical and public health problems around E-mail: [email protected] the world. It has been predicted that there will be at least 350 Guarantors: U Schwab and M Uusitupa. Contributors: US contributed to study plan, subject recruitment, data analysis million people in the world with type 2 diabetes by the year and interpretation, and was the principal author of the manuscript. 2030 (WHO, 2003). Type 2 diabetes mellitus is a disease, that AL was involved in subject recruitment, counseling of the subjects and data usually develops gradually through impaired glucose toler- analysis. AT was involved in study plan, planning of the fiber drinks and data ance (IGT). A person with a high risk of developing DM2 has interpretation. MU contributed to study plan and data interpretation. Received 30 June 2005; revised 21 December 2005; accepted 16 January several characteristics of metabolic syndrome, that is, central 2006; published online 8 March 2006 obesity, elevated blood pressure, and low serum High-density Sugar beet pectin and polydextrose U Schwab et al 1074 lipoprotein (HDL)-cholesterol and elevated serum triglycer- Table 1 Baseline characteristics of the subjectsa ide concentrations. Furthermore, the disease has a strong Control Polydextrose SBP genetic background, but sedentary lifestyle and increasing n ¼ 22 n ¼ 22 n ¼ 22 prevalence of obesity largely explain the current epidemic of DM2. Type 2 diabetes mellitus increases markedly the risk of Gender (M/F) 10/12 10/12 9/13 cardiovascular diseases. Research into primary prevention Type 2 DM (yes/no) 1/21 5/17 4/18 7 7 7 should thus not only focus on impaired glucose metabolism, Age (years) 53 8557519 Weight (kg) 79.8711.3 80.1712.8 83.3710.5 but also on obesity, blood pressure and serum lipid 2 BMI (kg/m ) 28.873.5 28.473.2 29.473.4 concentrations. Three recent long-term lifestyle intervention trials have Blood pressure (mmHg) shown that it is possible to prevent or delay the development Systolic 136714 133716 134710 Diastolic 84710 82788578 of DM2 by changing lifestyles. In the Finnish Diabetes fP-glucose (mmol/l) 6.671.0 6.370.7 6.570.7 Prevention Study, besides weight reduction and physical fS-cholesterol (mmol/l) 5.5971.00 5.4770.85 5.6571.00 activity also an increased intake of dietary fiber and a HDL-cholesterol (mmol/l) 1.1570.30 1.3270.38 1.1770.28 modification of the quality of dietary fat contributed to risk LDL-cholesterol (mmol/l) 3.5370.88 3.4170.65 3.8371.03 fS-triglycerides (mmol/l) 1.9171.03 1.0370.74 1.5470.88 reduction (Tuomilehto et al., 2001; Lindstro¨m et al., 2003). The study by Chandalia et al. (2000) showed a beneficial Abbreviations: BMI ¼ body mass index; DM ¼ diabetes mellitus; HDL ¼ high- effect of dietary fiber on glucose homeostasis in subjects with density lipoprotein; LDL ¼ low-density lipoprotein; SBP ¼ sugar beet pectin. a DM2. Intake of soluble fiber has been shown to improve Measured at screening. postprandial glucose metabolism, and guar gum, a soluble fiber, has been shown to improve glucose concentrations in DM2 patients (Aro et al., 1981; Fuessl et al., 1987; Groop B-GHbA1c p8%, for serum concentrations of total choles- et al., 1993). terol o7.5 mmol/l and for total triacylglycerols (TG) The aim of the study was to examine the effects of sugar o4 mmol/l. The subjects without diagnosed DM2 partici- beet pectin (SBP) and polydextrose (PDX) as a part of the pated in an oral glucose tolerance test. Those with normal daily diet on fasting concentrations of plasma glucose and fasting and 2 h plasma glucose concentrations (p6 and serum total and lipoprotein lipids as well as on postprandial o7.8 mmol/l, respectively) were excluded. So, subjects with glycemia in middle-aged subjects with abnormal glucose IGT and IFG were included. Subjects with lipid-lowering metabolism (IGT, impaired fasting glucose (IFG) or diet- medication, oral hypoglycemic medication or insulin treat- treated DM2). ment were excluded. Subjects with unstable angina, history of myocardial infarction, coronary artery bypass graft, percutaneous transluminal coronary angioplasty within the Subjects and methods past 6 months, transient ischemic attacks or malignancy were excluded as well. Subjects The subjects were not allowed to use any nutrient Seventy subjects (31 male and 39 female subjects) with supplements during the intervention or within 1 month of normal liver, kidney and thyroid functions were recruited for the beginning of the intervention. the study. Two of the subjects dropped out at the very The subjects gave their written informed consent, and the beginning of the study, one owing to lack of motivation and study protocol was approved by the Ethical Committee of the other owing to skin problems with unknown etiology. the Hospital District of Northern Savo. Two subjects dropped out at the very end of the study, one owing to excessive alcohol intake at the very end of the study and one owing to scheduling problems. Therefore, the Study design final number of subjects was 66 (29 male and 37 female The study was a placebo-controlled, randomized, parallel subjects) of which 10 had diet-treated DM2. Seventeen of the double-blinded study. The length of the intervention period subjects had hypertension and five reported to have was 12 weeks. Fiber enrichment was consumed as a drink. gallstones. Forty subjects were on a regular medication The daily dose was 4 dl (containing 16 g fiber per day), which (antihypertensive medication, postmenopausal hormonal was ingested as two 2 dl portions with main meals. During replacement therapy). The subjects were asked to keep their the first week of the intervention, the subjects consumed medication constant during the study. The subjects were only half of the daily dose (2 dl). The subjects were given randomly assigned to one of the three groups. The baseline nutrition counseling by a nutritionist aiming to meet the characteristics of the subjects in each group are presented in generally recommended dietary goals for the prevention of Table 1. cardiovascular diseases and DM2 (DNSG, 2000; NCEP, 2001). For body mass index (BMI) the inclusion criterion The subjects met a clinical nutritionist at the visits of 0, 2, 4 was p35 kg/m2 and for age it was 30–65 years. The inclusion and 8 weeks. The nutrition counseling concentrated on the criterion for fasting plasma glucose was p8 mmol/l, for modification of the quality of fat and increasing the intake of

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1075 dietary fiber. The diet counseled was eucaloric and active All venous blood samples were drawn after an overnight weight loss was not an aim. fast (12 h). For the analysis of serum total and lipoprotein The subjects in the SBP group consumed a drink enriched lipid concentrations, samples were ultracentrifuged for 18 h in SBP obtained from Danisco Sugar (Helsinki, Finland and at the density of 1.006 kg/l to remove very low-density Nakskov, Denmark). The PDX group consumed a drink lipoprotein (VLDL). High-density lipoprotein in the infra- enriched in PDX formed by Danisco Sweeteners (Helsinki, nate was separated from LDL by precipitation of LDL with Finland and Nakskov, Denmark). The market product of PDX dextran sulfate and magnesium chloride (Penttila¨ et al., is known as Litesses. The control group consumed a drink 1981). Low-density lipoprotein-cholesterol concentration without fiber enrichment. All the drinks had the same was calculated as the difference between the mass of ingredients: water, fructose 70 g/l, sodium bentsoate 0.11 g/l, cholesterol in the infranate and HDL. Enzymatic colori- potassium sorbate 0.23 g/l and citric acid 0.5 g/l. Artificial metric methods with commercial kits (cholesterol: CHOD- flavor was raspberry in the SBP drink and blackcurrant in the PAP, triglyceride GPO-PAP; Roche Diagnostics GmbH, PDX and control drinks. The SBP drink contained 40 g/l SBP Mannheim, Germany) were used to determine cholesterol of which 76% was soluble fiber and the PDX drink contained and TG concentrations from whole serum and separated 40 g/l PDX. The drinks were produced by Trensums food AB lipoproteins with an automated instrument (Kone Specific (Tingsryd, Sweden). Clinical Analyzer, Kone Ltd, Espoo, Finland). The subjects visited the research unit at the beginning of Plasma glucose concentration was analyzed using the the study (0 week) and four times during the intervention: at glucose dehydrogenase method (Granutest 500, Merck 2, 4, 8 and 12 weeks. Body weight was measured at each visit. KGaA, Darmstadt, Germany). Plasma insulin concentration Blood pressure, concentrations of plasma glucose, plasma was analyzed with a radioimmunoassay (RIA) method insulin and serum total and lipoprotein lipids, oxidized Low- (Phadeseph Insulin RIA 100, Pharmacia Diagnostics, Uppsala, density lipoprotein (LDL), and GHbA1c proportion were Sweden). measured at 0-, 4-, 8- and 12-week visits. Blood hemoglobin, The proportion of B-GHbA1c, plasma concentrations of C- hematocrit, leukocytes, thrombocytes, liver enzymes, serum peptide, sodium, potassium and calcium as well as serum sodium, potassium and calcium were measured at 0-, 4- and concentrations of gamma-glutamyl transferase (g-GT), ala- 12-week visits. nine aminotransferase (ALT), asparagine aminotransferase Postprandial glycemia was examined in 24 volunteers (ASAT), alkaline phosphatase (ALP), thyroid-stimulating (eight subjects from each group) at the beginning and at the hormone (TSH) and creatinine were analyzed using the end of the intervention period (0 and 12 weeks, respectively) methodology currently in use at the Kuopio University after a standardized breakfast ingested after a 12-h overnight Hospital, Kuopio, Finland. fast. The content of the breakfast was 40 g white bread, 40 g The concentration of oxidized LDL was measured as the cucumber, 160 g orange juice and 2 dl drink consumed baseline level of conjugated dienes in lipids extracted from during the intervention. Blood samples were drawn and a LDL (Ahotupa et al., 1996). visual analogue scale was filled just before the ingestion of The visual analogue scale used to evaluate the feelings of the breakfast (0 min) and five times after the ingestion of the hunger, satiety, fullness and desire to eat during the breakfast (at 15, 30, 60, 120 and 180 min). postprandial test was based on the publication by Flint The subjects kept a 4-day food record just before the et al. (2000). intervention to monitor their habitual diet and to enable The laboratory personnel were unaware of the randomiza- individual nutrition counseling. During the intervention tion group during the intervention period and laboratory period, the subjects kept 4-day food records during the third analyses. and seventh weeks. The food records were calculated using the Micro-Nutricas dietary analysis software based on Finnish food analyses and international food composition Statistical analyses tables (Rastas et al., 1997). The data were analyzed with the SPSS/PC þ statistics The subjects also filled a questionnaire of the appearance program (V11.0, SPSS, Chicago, IL, USA). Before further of side effects in gastro-intestinal tract at 0-, 4- and 12-week analyses, normal distribution of the variables was checked visits. Changes in feelings of thirst were also monitored. with the Kolmogorov–Smirnov test with Lilliefors signifi- cance correction. Variables with abnormal distribution (TG,

HDL-cholesterol and GHbA1c) were log-transformed and the Methods log-values were used in further analyses. The postprandial Body weight was measured using the same calibrated area under the curve (AUC) of plasma glucose, insulin and C- electronic scale throughout the study. Blood pressure was peptide was calculated using a Canvas software (v. 6.0, measured from right arm after 5 min of rest in the sitting Deneba Systems Inc., Miami, FL, USA). Two-factor repeated position using a mercury sphygmomanometer. Two mea- measures ANOVA (GLM in SPSS) was used to test the changes surements were performed 5 min apart. The mean of these within time and group, and to examine the interaction two measurements was used for further analyses. between time and group. Paired samples t-test was used for

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1076 two-tailed comparisons within the groups. Student’s t-test the control group (Table 4). However, the proportion of was used for between-group comparisons. Friedman’s test GHbA1c increased marginally but statistically significantly and Wilcoxon matched-pairs signed-ranks test were used to in the intervention groups with no change in the control test the changes in variables that did not become normal by group. log-transformation. All data are expressed as mean7s.d. P-level o0.05 was considered as statistically significant. Serum total and lipoprotein lipids, and oxidized LDL There were no significant changes in serum total cholesterol Results concentration in either of the intervention groups, but there was a significant (Po0.05) decrease in the control group. Dietary intake Serum HDL-cholesterol concentration increased signifi- There were favorable changes in the intake of fat, especially cantly in the intervention groups, whereas serum LDL- in the intake of saturated fat, carbohydrates, fiber and cholesterol concentration decreased significantly in the cholesterol in all groups (Table 2). The intake of fat, saturated control and PDX groups. Furthermore, total to HDL- fat and dietary cholesterol decreased, whereas the intake of cholesterol ratio decreased significantly in all groups carbohydrates and dietary fiber increased. (Table 5). There were no changes in the concentration of serum total triglycerides. Concentration of oxidized LDL or the ratio of oxidized LDL to total cholesterol did not change Body weight and blood pressure during the study in any of the groups (data not shown). There were small but significant decreases (Po0.05) in body weight and BMI in the control and PDX groups (Table 3). Systolic blood pressure decreased significantly in the SBP Postprandial glycemia group and diastolic blood pressure decreased significantly in There were no differences in AUC of plasma glucose, C- the control group and in the SBP group (Table 3). peptide or insulin between or within the groups during the study. Area under the curve of plasma glucose was 2.171.5 vs 2.671.6 mmol/l min in the control group, 2.370.9 vs Plasma glucose concentration and proportion of glycosylated 2.972.6 mmol/l min in the PDX group and 2.270.9 vs

HbA1c 3.272.9 mmol/l min in the SBP group (0 week vs 12 weeks, Fasting plasma glucose concentration did not change in the respectively). In the visual analogue scale, no differences intervention groups, but increased significantly (Po0.05) in were found between or within the groups in the feeling of

Table 2 Intake of energy, energy nutrients, cholesterol, fiber and fructose during the study

Control Polydextrose SBP

0 week 12 weeks 0 week 12 weeks 0 week 12 weeks

Energy (kJ) 764271423 797871398 746471980 797872443 760272113 776871309

Fat (E%)a 31.375.7 26.376.4 32.177.0 27.874.9 32.576.2 28.474.7 Saturated 11.872.8 9.272.4 12.172.9 9.872.5 12.373.0 9.872.5 Monounsaturated 10.372.4 9.173.2 10.873.5 9.571.9 10.972.5 9.871.6 Polyunsaturated 5.771.9 5.371.4 5.271.2 5.471.1 5.671.3 5.671.0

Protein (E%) 18.574.1 18.873.9 18.772.5 18.571.3 17.972.8 17.872.5

Carbohydrates 47.677.6 53.276.4 46.877.1 51.375.3 46.474.9 51.375.0 Fructose (E%) 3.071.4 9.671.8 3.671.4 9.772.3 2.971.4 9.472.1

Fiber (g)b 25.177.6 29.576.3 26.9712.4 28.7712.3 23.978.8 26.478.8 Water soluble 5.471.5 6.671.4 6.072.4 6.472.7 5.371.7 6.071.9 Water insoluble 11.074.0 13.073.6 12.977.0 13.476.7 11.074.4 11.874.2 Cellulose 4.071.2 5.171.1 4.571.6 4.771.7 3.971.4 4.671.8 Lignine 2.870.8 3.370.9 3.271.6 3.471.8 2.871.1 3.271.4

Cholesterol (mg) 2337110 205790 258795 227780 242760 210760 Alcohol (E%) 2.774.3 1.672.2 2.474.1 2.473.5 3.177.3 2.572.9

Abbreviation: SBP ¼ sugar beet pectin. a% of energy intake. bWithout fiber enrichment.

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1077 Table 3 Body weight, BMI and blood pressure during the study, n ¼ 22 in each group

0 week 4 weeks 8 weeks 12 weeks Pa Ub

Body weight (kg) Control 79.5711.1 78.9710.9 78.8710.6 78.6710.6c *NS Polydextrose 80.4712.9 80.1712.8 79.8713.1 79.4713.0c SBP 83.4710.8 83.3710.8 83.4710.6 83.1710.4

BMI (kg/m2) Control 28.773.4 28.573.4 28.573.4 28.473.4c *NS Polydextrose 28.373.2 28.273.1 28.173.3 27.973.4c SBP 29.473.4 29.373.4 29.473.4 29.373.3

Blood pressure (mm Hg) Systolic Control 135714 133715 131715 134714 ** NS Polydextrose 135717 136716 140716 135718 SBP 134713 133713 131713 129715c

Diastolic Control 8679827980798179c * 0.064 Polydextrose 8578857884788479 SBP 8679847882788077c

Abbreviation: BMI ¼ body mass index; SBP ¼ sugar beet pectin. aGLM, *significant effect of time without interaction of time and group, **significant interaction of time and group. bUnivariate analysis of variance; difference between groups, starting level as a covariate. cPo0.05 within the group, 0 week vs 12 week.

Table 4 Concentration of plasma glucose and percentage of glycosy- trations of sodium, potassium or calcium or serum concen- lated hemoglobin during the study, n ¼ 22 in each group trations of liver enzymes (g-GT, ASAT, ALT or ALP) or 0 week 4 weeks 8 weeks 12 weeks Pa Ub creatinine during the study (data not shown). fP-glucose (mmol/l) Control 6.470.9 6.270.9 6.571.1 6.670.8c NA NA Polydextrose 6.370.9 6.570.8 6.670.9 6.471.0 SBP 6.470.5 6.470.5 6.270.7 6.570.7 Discussion

d B-GHbA1c (%) The subjects of the present study had either a diet-treated 7 7 7 7 Control 6.0 0.6 5.9 0.5 6.0 0.5 6.0 0.5 * NS type 2 diabetes or they were at high risk of developing Polydextrose 5.970.5 6.070.6 6.170.6 6.170.7e SBP 5.770.4 5.870.3 5.970.4 5.870.4e diabetes owing to an abnormal glucose metabolism (IGT or IFG). Increased intake of soluble fiber has been found to Abbreviation: NA ¼ not analyzed; SBP ¼ sugar beet pectin. improve glucose homeostasis and postprandial glucose aGLM, *significant effect of time without interaction of time and group. metabolism (Fuessl et al., 1987; Groop et al., 1993). Guar bUnivariate analysis of variance; difference between groups, starting level as a covariate. gum is the most widely studied form of soluble fiber. In the cP ¼ 0.007, Friedman test. present study, the effects of PDX and SBP were studied. d Log-transformed. Polydextrose, trade name Litesses, is a glucose polymer eP 0.05, 0 week vs 12 weeks, paired samples t-test. p poorly absorbed by mammals, and it is widely used by food industry. It is readily soluble in water. However, unlike other soluble fibers, there is no gel formation by PDX in water hunger, satiety, fullness or desire to eat except that there was (Oku et al., 1991; Murphy, 2001). Sugar beet pectin contains a significant (Po0.05) difference in the feeling of hunger at 73% dietary fiber (non-starch polysaccharides) of which one- 180 min within the PDX group (6.073.6 vs 3.372.5, 0 week third is water-soluble fiber, mainly pectin. vs 12 weeks, respectively) (data not shown). The subjects received nutrition counseling concentrating on the intake of fat, carbohydrates and dietary cholesterol as well as the quality of fat. Although active weight loss was not Other blood samples an aim, there was a small but significant decrease in body There were no changes in hematocrit or blood concentra- weight in the PDX and control groups, whereas no change tions of hemoglobin, leukocytes or thrombocytes (data not was found in the SBP group. However, there were no shown). Neither were there any changes in plasma concen- significant differences in body weight between the groups.

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1078 Table 5 Concentrations of serum total and lipoprotein lipids, and oxidized LDL during the study, n ¼ 22 in each group

0 week 4 weeks 8 weeks 12 weeks Pa Ub

fS-cholesterol (mmol/l) Control 5.6170.94 5.4670.99 5.5471.10 5.3271.03c *NS Polydextrose 5.4370.87 5.3770.84 5.3670.97 5.2770.82 SBP 5.7171.08 5.5971.02 5.6870.85 5.5470.68

fS-HDL-cholestrol (mmol/l)d Control 1.1470.32 1.1770.36 1.1570.34 1.1770.31 * NS Polydextrose 1.2570.32 1.2770.33 1.2970.31 1.3270.35c SBP 1.1270.23 1.1370.22 1.1570.25 1.1970.25c

fS-LDL-cholestrol (mmol/l) Control 3.6470.71 3.4270.83 3.4570.87 3.2870.79c *NS Polydextrose 3.5870.80 3.3770.68 3.3870.85 3.2970.71c SBP 3.8871.04 3.7071.02 3.7570.87 3.6570.77

Total to HDL-cholesterol ratio Control 5.2071.40 5.0071.45 5.1671.71 4.8571.55c *NS Polydextrose 4.5671.34 4.4471.46 4.3471.22 4.2271.22c SBP 5.3171.49 5.1471.32 5.1371.18 4.8571.16c

fS-triglycerides (mmol/l)d Control 1.8470.95 1.9471.12 1.9671.19 1.9671.46 NS Polydextrose 1.3070.51 1.6271.48 1.6171.03 1.4670.71 SBP 1.5971.21 1.7171.07 1.8571.20 1.6070.25

Abbreviation: HDL ¼ high-density lipoprotein; LDL ¼ low-density lipoprotein; SBP ¼ sugar beet pectin. aGLM, *significant effect of time without interaction of time and group. bUnivariate analysis of variance; difference between groups, starting level as a covariate. cPp0.05, 0 week vs 12 weeks, paired samples t-test. dLog-transformed.

The fasting concentration of plasma glucose did not There were no significant differences between the groups change in the PDX and SBP groups, but increased in the in serum total or lipoprotein lipid concentrations. Serum control group. In previous studies, guar gum (Aro et al., 1981; LDL-cholesterol concentration decreased in the PDX Fuessl et al., 1987; Groop et al., 1993) as well as other sources and control groups. The concentration of HDL-cholesterol of soluble fiber (Vuksan et al., 1999; Chandalia et al., 2000; increased in both SBP and PDX groups, whereas the total to Pins et al., 2002) have been reported to result in either a HDL-cholesterol ratio decreased in all groups without significant or a nonsignificant decrease in blood or plasma a difference among groups. The increase of HDL-cholesterol glucose concentrations. There were no significant differences is usually difficult to achieve by dietary means except by in GHbA1c between the groups. However, it increased weight loss. An increase of HDL-cholesterol concentration slightly in the SBP and PDX groups. The drinks were asked by 0.03 mmol/l has been estimated to decrease the mortality to be ingested with main meals. This together with the owing to atherosclerosis by 2–3% (von Eckardstein et al., increased intake of fructose during the intervention might 2001). In this study, HDL-cholesterol concentration play a role, although according to the most recent recom- increased significantly by 0.07 mmol/l in both PDX and mendations the intake of fructose up to 30 g/day appears to SBP groups, however, without a significant difference have no deleterious effects on concentrations of plasma as compared with the control group. Previous studies insulin or serum total and lipoprotein lipids in subjects with have shown a small decrease by soluble fiber in serum total DM2 (Mann et al., 2004). The daily dose (4 dl) of the drink and LDL-cholesterol concentrations without a decreasing contained 28 g fructose. In animal studies, increased intake effect on HDL-cholesterol (Brown et al., 1999; Knopp of fructose has been found to cause insulin resistance or et al., 1999; Anderson et al., 2000; Vuksan et al., 2000). worsen existing insulin resistance (Blakely et al., 1981). In Regarding concentrations of serum triglycerides, the results humans, the metabolic effect of increased fructose intake is are controversial; in some studies dietary fiber has been not clear (Elliott et al., 2002). Hallfrisch et al. (1983) have shown to have no effect on the concentration of found that a diet containing 15% of energy as fructose has serum triglycerides (Vuksan et al., 1999, 2000), whereas in detrimental effects on glucose metabolism both in normal some other studies a decreasing effect of dietary fiber has and hyperinsulinemic men. Reduced insulin sensitivity and been reported (Chandalia et al., 2000). The most commonly insulin binding to its receptors has also been reported (Beck- reported negative effect of fructose on serum lipid profile has Nielsen et al., 1980). been the increased concentration of serum triglycerides

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1079 (Bantle et al., 2000). This effect was not seen in the present psyllium intake adjunctive to diet therapy in men and women study. with hypercholesterolemia: meta-analysis of 8 controlled trials. Am J Clin Nutr According to previous studies, the effect of soluble fiber on 71, 472–479. Aro A, Uusitupa M, Voutilainen E, Hersio K, Korhonen T, Siitonen O blood pressure is controversial. A decreasing effect on (1981). Improved diabetic control and hypocholesterolaemic systolic blood pressure has been found in some (Vuksan effect induced by long-term dietary supplementation with et al., 1999; Pins et al., 2002) but not all studies (Vuksan et al., guar gum in type 2 (insulin-independent) diabetes. Diabetologia 2000). In the present study, systolic blood pressure decreased 21, 29–33. Bantle J, Raatz S, Thomas W, georgopoulos A (2000). Effects of dietary in the SBP group only (6 mm Hg). The DASH Study (Dietary fructose on plasma lipids in healthy subjects. Am J Clin Nutr 72, Approaches to Stop Hypertension) resulted in 3.5 mm Hg 1128–1134. loss in normotensives and 11.4 mm Hg loss in hypertensives. Beck-Nielsen H, Pedersen O, Lindskov HO (1980). Impaired Diastolic blood pressure decreased significantly and quite cellular insulin binding and insulin sensitivity induced by high-fructose feeding in normal subjects. Am J Clin Nutr 33, substantially in the SBP and control groups (6 and 5 mm Hg, 273–278. respectively). In the DASH Study, the magnitude of the Blakely SR, Hallfrisch J, Reiser S, Prather ES (1981). Long-term effects decrease of diastolic blood pressure was 2.1 mm Hg in of moderate fructose feeding on glucose tolerance parameters in normotensives and 5.5 mm Hg in hypertensives (Svetkey rats. J Nutr 111, 307–314. Bray GA, Nielsen SJ, Popkin BM (2004). Consumption of high- et al., 1999). In the control group, the decrease in diastolic fructose corn syrup in beverages may play a role in the epidemic of blood pressure might be partly explained by weight loss. obesity. Am J Clin Nutr 79, 537–543. There were no differences in postprandial concentrations Brown L, Rosner B, Willett WW, Sacks F (1999). Cholesterol-lowering of plasma glucose, insulin or C-peptide among the groups. In effects of dietary fiber: a meta-analysis. Am J Clin Nutr 69, 30–42. previous studies, PDX has been reported to have an acute Chandalia M, Garg A, Lutjohann D, von Bergmann K, Grundy SM, Brinkley LJ (2000). Beneficial effects of high dietary fiber intake decreasing effect on glucose absorption in rats. However, the in patients with type 2 diabetes mellitus 2. N Engl J Med 342, intestine seems to adapt to its effect (Yoshioka et al., 1996). 1392–1398. There are no previous studies in SBP, but guar gum has been The Diabetes Nutrition Study group (DNSG) of the European found to attenuate postprandial glycemia (Uusitupa et al., Association for the Study of Diabetes (EASD) (2000). Recommen- dations for the nutritional management of patients with diabetes 1984). A confounding factor in the present study might have mellitus. Eur J Clin Nutr 54, 353–355. been the relatively high fructose intake per day. High Elliott SS, Keim NL, Stern JS, Teff K, Havel PJ (2002). Fructose, fructose intake has been reported to lower circulating insulin weight gain, and the insulin resistance syndrome. Am J Clin Nutr concentrations (Elliott et al., 2002; Teff et al., 2004). In fact, 76, 911–922. Executive summary of the third report of the National Cholesterol fructose seems to be very poor in stimulating insulin Education Program (NCEP) Expert Panel on detection, evaluation, secretion (Bray et al., 2004). In addition to lowered and treatment of high blood cholesterol in adults (Adult postprandial serum insulin concentration, fructose also Treatment Panel III) (2001). J Am Med Assoc 285, 2486–2497. seems to attenuate postprandial serum leptin and ghrelin Flint A, Raben A, Blundell JE, Astrup A (2000). Reproducibility, power and validity of visual analogue scales in assessment of appetite et al concentrations (Teff ., 2004). As serum insulin and leptin sensations in single test meal studies. Int J Obes Relate Metal Disord concentrations have a key role in the regulation of eating, 24, 38–48. this may explain why there were no differences in the Fuessl HS, Williams G, Adrian TE, Bloom SR (1987). Guar sprinkled feelings of hunger, satiety, fullness and desire to eat in this on food: effect on glycaemic control, plasma lipids and gut hormones in non-insulin dependent diabetic patients. Diabet Med study. 4, 463–468. In summary, in subjects with abnormal glucose metabo- Groop PH, Aro A, Stenman S, Groop L (1993). Long-term effects of lism, SBP and PDX do not have positive effect on fasting or guar gum in subjects with non-insulin-dependent diabetes postprandial plasma glucose concentrations or plasma lipid mellitus. Am J Clin Nutr 58, 513–518. profile. Hallfrisch J, Ellwood KC, Michaelis OE, Reiser S, O’Dorisio TM, Prather ES (1983). Effects of dietary fructose on plasma glucose and hormone responses in normal and hyperinsulinemic men. J Nutr 113, 1819–1826. Acknowledgements Knopp RH, Superko RH, Davidson M, Insull W, Dujovne CA, Kwiterovich PO et al. (1999). Lond-term blood cholesterol-lowering effects of dietary fiber supplement. Am J Prev Med 17, 18–23. We thank laboratory personnel, Ms Pa¨ivi Turunen, Mrs Kaija Lindstro¨m J, Louheranta A, Mannelin M, Rastas M, Salminen V, Kettunen, Mrs Eeva Lajunen and Mr Sebastiaan Bol for Eriksson J et al. (2003). The Finnish Diabetes Prevention Study excellent technical assistance. (DPS). Lifestyle intervention and 3-year results on diet and physical activity. Diabetes Care 26, 3230–3236. Mann JI, De Leeuw I, Hermansen K, Karamanos B, Karlstro¨mB, Katsilambros N et al. (2004). Evidence-based nutritional ap- References proaches to the treatment and prevention of diabetes mellitus. Nutr Metab Cardiovasc Dis 14, 373–394. Ahotupa M, Ruutu M, Ma¨ntyla¨ E (1996). Simple methods of Murphy O (2001). Non-polyol low-digestible carbohydrates: food quantifying oxidation products and antioxidant potential of low applications and functional benefits. Br J Nutr 85, 47–53. density lipoproteins. Clin Biochem 29, 139–144. Oku T, Fuji Y, Okamatsu H (1991). Polydextrose as dietary fiber. Anderson JW, Allgood LD, Lawrence A, Altringer LA, Jerdack GR, Hydrolysis by digestive enzymes and its effects on gastrointestinal Hengehold DA et al. (2000). Cholesterol-lowering effects of transit time in rats. J Clin Biochem Nutr 11, 31–40.

European Journal of Clinical Nutrition Sugar beet pectin and polydextrose U Schwab et al 1080 Penttila¨ IM, Voutilainen E, Laitinen O, Juutilainen P (1981). (2001). Prevention of type 2 diabetes mellitus by changes in Comparison of different analytical and precipitation methods lifestyle among subjects with impaired glucose tolerance. N Engl J for the direct estimation of high-density lipoprotein cholesterol. Med 344, 1343–1350. Scand J Lab Invest 41, 353–360. Uusitupa M, Aro A, Korhonen T, Tuunainen A, Sarlund H, Penttila I Pins JJ, Geleva D, Keenan JM, Frazel C, O’Connor PJ, Cherney LM (1984). Blood glucose and serum insulin responses to breakfast (2002). Do whole-grain oat cereals reduce the need for antihy- including guar gum and cooked or uncooked milk in type 2 (non- pertensive medications and improve blood pressure control? JFam insulin-dependent) diabetic patients. Diabetologia 26, 453–455. Pract 51, 353–359. von Eckardstein A, Nofer JR, Assmann G (2001). High density Rastas M, Seppa¨nen R, Knuts L-R, Hakala P, Karttila V (eds.) (1997). lipoproteins and arteriosclerosis. Role of cholesterol efflux Nutrient Composition of Foods. Publications of the Social Insurance and reverse cholesterol transport. Arterioscler Thromb Vasc Biol Institution: Helsinki, Finland. 21, 13–27. Svetkey LP, Simons-Morton D, Vollmer WM, Appel LJ, Conlin PR, Vuksan V, Jenkins DJA, Spadafora P, Sievenpiper JL, Owen R, Vidgren Ryan DH, et al., for the DASH research group (1999). Effects of E et al. (1999). Konjac-Mannan (Glucomannan) improves glyce- dietary patterns on blood pressure: subgroup analysis of the mia and other associated risk factors for coronary heart disease in Dietary Approaches to Stop Hypertension (DASH) randomized type 2 diabetes. Diabetes Care 22, 913–919. clinical trial. Arch Intern Med 159, 285–293. Vuksan V, Sievenpiper JL, Owen R, Swilley JA, Spadafoda P, Jenkins Teff KL, Elliott SS, Tscho¨p M, Kiefer TJ, Rader D, Heiman M et al. DJA et al. (2000). Beneficial effects of viscous dietary fiber from (2004). Dietary fructose reduces circulating insulin and leptin, Konjac-Mannan in subjects with the insulin resistance syndrome. attenuates postprandial suppression of ghrelin, and increases Diabetes Care 23, 9–14. triglycerides in women. J Clin Endocrinol Metab 89, 2963–2972. Yoshioka M, Reiko D, Shimomura Y, Suzuki M (1996). Effects of Tuomilehto J, Lindstro¨m J, Eriksson JG, Valle TT, Ha¨ma¨la¨inen H, dietary polydextrose on in vitro intestinal absorption rate in rats. Ilanne-Parikka P, et al., Finnish Diabetes Prevention Study Group Nutr Res 16, 245–249.

European Journal of Clinical Nutrition