Journal of Human Nutrition and Dietetics
RESEARCH PAPER The effects of a diet rich in inulin or wheat fibre on markers of cardiovascular disease in overweight male subjects L. Tripkovic,1 N. C. Muirhead,1 K. H. Hart,1 G. S. Frost1,2 & J. K. Lodge1,3 1
Faculty of Health and Medical Sciences,, Department of Nutritional Sciences, University of Surrey, Guildford, UK Department of Medicine, Imperial College, University of London, London, UK 3 Faculty of Health and Life Sciences, Northumbria University, Ellison Building, Newcastle-Upon-Tyne, UK 2
Keywords arterial stiffness, cardiovascular disease, dietary intervention, fibre, glucose control, humans, inulin, lipid status, wheat germ, whole grain. Correspondence John K. Lodge, Faculty of Health & Life Sciences, Ellison Building, Northumbria University, Newcastle-Upon-Tyne, Tyne & Wear NE1 8ST, UK. Tel.: +44 (0)191 243 7710 Fax: +44 (0)191 227 3519 E-mail: [email protected] How to cite this article Tripkovic L., Muirhead N.C., Hart K.H., Frost G.S. & Lodge J.K. (2014) The effects of a diet rich in inulin or wheat fibre on markers of cardiovascular disease in overweight male subjects. J Hum Nutr Diet. doi: 10.1111/jhn.12251
Abstract Background: Previous studies suggest that the beneficial health effects of a diet rich in whole grains could be a result of the individual fibres found in the grain. The present study aimed to investigate the influence of a diet high in either wheat fibre (as an example of an insoluble fibre) or inulin (a nondigestible carbohydrate) on markers of cardiovascular disease. Methods: Ten male participants classified as at higher risk of cardiovascular disease [mean (SD) body mass index 30.2 (3) kg m 2, mean (SD) waist circumference 106.4 (7) cm, mean (SD) age 39.8 (9) years] were recruited to a randomised, controlled, cross-over study comparing the consumption of bespoke bread rolls containing either inulin, wheat germ or refined grain (control) (15 g day 1) for 4 weeks with a 4-week washout period between each regime. At the end of each regime, participants underwent an oral glucose tolerance test (OGTT), measures of pulse wave velocity (PWV), 24-h ambulatory blood pressure (AMBP), plasma lipid status and markers of glucose control. Results: There was no difference in measures of glucose control, lipid status, 24-h AMBP or PWV after the intervention periods and no changes compared to baseline. There was no significant difference between OGTT glucose and insulin time profiles; however, there was a significant difference in area under the curves between the wheat fibre and control interventions when comparing change from baseline (control +10.2%, inulin +4.3%, wheat fibre 2.5%; P = 0.03). Conclusions: Only limited differences between the interventions were identified, perhaps as a consequence of the amount of fibre used and intervention length. The wheat germ intervention resulted in a significant reduction in glucose area under the curve, suggesting that this fibre may aid glucose control.
Introduction Increasing whole grain intake is considered to be associated with decreasing cardiovascular disease (CVD) (Mellen et al., 2008; Aune et al., 2013; Cho et al., 2013), although intervention studies have not been convincing (Andersson et al., 2007; Brownlee et al., 2010). Most intervention studies implementing an increase in whole grain intake ª 2014 The British Dietetic Association Ltd.
focus on the grains as a whole entity. However, data are becoming available to suggest the beneficial health effects of the individual fibres found within the grain, such as inulin and wheat fibre. Both inulin and wheat fibre are found specifically within the wheat kernel, with inulin being an example of a fructan that is a nondigestible soluble fibre, whereas wheat fibre is an example of an insoluble fibre. The amount of inulin and wheat fibre in 1
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
wholegrain wheat flour is very different, with approximately 0.6–1 g/100 g wheat for inulin but approximately 12 g/100 g of wheat fibre (Jones, 2008; Biesiekierski et al., 2011). Inulin is a generic term that covers a range of linear fructans with degrees of polymerisation, although, typically, a hydrolysed form obtained from Chicory is used in health studies (Flamm et al., 2001; Kaur & Gupta, 2002; Roberfroid, 2005). Evidence is available to suggest that these constituents of wholegrain may exert beneficial effects on cardiovascular health, thus lowering the risk of mortality and morbidity over time (Flamm et al., 2001; Kaur & Gupta, 2002). For example, wheat fibre may play an important role in improving peripheral insulin sensitivity because Weickert et al. (2006) demonstrated that whole-body glucose disposal was significantly improved in a group of overweight individuals who consumed greater quantities of insoluble fibre in an enriched bread compared to a control bread. By contrast, the health benefits of inulin appear to be broader but have been concentrated on their lipid-lowering effects (Davidson & Maki, 1999; Williams & Jackson, 2002; Letexier et al., 2003; Beylot, 2005; Russo et al., 2008), with studies showing significant reductions in plasma triacyglycerides and moderate reductions in total and low-density lipoprotein cholesterol (Jackson et al., 1999; Beylot, 2005). More recently, inulin has been shown to have a prebiotic effect, increasing markers of colonic fermentation (Fernandes et al., 2011) and affecting appetite (Tarini & Wolever, 2010). However, overall, the current body of evidence regarding inulin and wheat fibre and their respective potentials in exerting beneficial metabolic changes remains mixed and requires clarification. One question that remains is whether the separate fibres found within a whole grain kernel impart separate effects or whether they have an additive effect and therefore must be consumed together as an intact grain to be most effective. The present study aimed to investigate the effects of inulin or wheat fibre supplementation (15 g day 1; the mean intake of fibre in the UK and representing an intake that was shown to be well tolerated; Williams & Jackson, 2002) versus a refined grain control, on markers of CVD in men at an increased risk of the disease. Materials and methods Study design The study comprised a three-way, cross-over, controlled, randomised dietary intervention study in men aged 30– 55 years. Participants consumed (in a randomised order) bread rolls that contained either wheat fibre or inulin or refined wheat grain (control) on top of their habitual diet. The rolls were matched for nutrient profiles [per 100 g: wheat fibre 932 kJ, 43.4 g of carbohydrate (CHO), 2
8.5 g of protein, 1.3 g of fat, 7.7 g of fibre; inulin 1015 kJ, 47.4 g of CHO, 9.3 g of protein, 1.4 g of fat, 7.6 g of fibre; control 1088 kJ, 50.7 g of CHO, 9.9 g of protein, 1.5 g of fat, 2.5 g of fibre]. Participants were requested to consume three rolls per day, with each roll containing 5 g of the ‘active’ intervention ingredient (i.e. 15 g day 1); either wheat fibre, inulin or refined grain. The rolls were consumed for 28 days, followed by a 28day washout period immediately after. The participants attended four study sessions: one at the beginning (Day 0 – baseline) and then one at the end of each intervention stage (Study Day 1 to Day 28; Study Day 2 to Day 84; Study Day 3 to Day 140). The study received a favourable ethical opinion from the University of Surrey Ethics Committee (EC/2007/73/FHMS). Participants The changes in plasma insulin and glucose concentrations were the primary end-points of the study with sample size calculations based on the change in insulin. A sample size of 10 would have at least 80% power to detect a significant drop in insulin of 20 pM (a difference to detect of 25%) between the groups assuming that the common standard deviation is 16 pM or 20% (Robertson et al., 2005). Participants suitable for inclusion were adult males, aged 30–55 years, waist circumference >94 cm (37 inches), body mass index (BMI) of 25–35 kg m 2, not taking any prescription medicines or supplements within the past 6 months, not drinking more than 21 units of alcohol per week, and not regularly undertaking vigorous exercise or fitness training (no more than 3 9 30-min aerobic sessions per week). Participants were recruited from University staff and the surrounding area through e-mail and advertisement. Participants were randomly assigned to their starting intervention group and to the order in which they would receive the three intervention products using web-based randomisation software (http://www.randomization.com). Study protocol The week before commencing the study, participants completed a 3-day diet diary to record their dietary intake at baseline. The night before the study, the participants consumed a standard meal (chosen from a set list matched for macronutrient intake) in addition to refraining from strenuous exercise, alcohol or caffeine. Study session: all sessions All study sessions were held at the Clinical Investigation Unit, Faculty of Health and Medical Sciences, University ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
of Surrey and followed the same protocol. Participants arrived early in the morning (session start times between 07.30 h and 09.00 h) and in a fasted state (12-h fasted). Anthropometric measurements were taken first using standard methodologies. Height was measured via a wall-mounted Harpenden Stadiometer (Holtain Ltd, Crymych, UK) to the nearest 0.1 cm. Weight and body fat percentage were measured, and body mass index calculated, using a Tanita TBF-300 Body Composition Analyser (Tanita Europe BV, Amsterdam, The Netherlands). Waist and hip circumferences were measured using a flexible tape measure, to the nearest 0.1 cm, around the navel and widest part of the hips, respectively. Participants were then requested to rest in a quiet room with dimmed lighting for 10 min in preparation for the pulse wave velocity (PWV) measurements. After the resting phase, triplicate recordings of static, supine blood pressure were taken and the process of recording PWV commenced. The participant was then immediately cannulated and an initial 30-mL fasting baseline blood sample taken. An oral glucose tolerance test (OGTT) was then conducted. The participants consumed 75 g of anhydrous glucose dissolved in 400 mL of water, within the space of 10 min. With fasted blood samples already taken prior to consuming the glucose, further blood sampling of 5 mL continued every 15 min for a further 2 h. Once the blood sampling had been completed, participants were offered refreshments and were then fitted with the ambulatory blood pressure monitors (ABPM) to be worn for the next 24 h. Prior to leaving the Clinical Investigation Unit, participants were provided with an allocated quantity of bread rolls for the entire intervention period. Participants were required to consume their allocated rolls per day (containing 15 g of test ingredient) and were blinded to the contents of the roll. Participants were encouraged to incorporate the rolls into their habitual diet (e.g. by adapting their lunch by substituting their routine bread for the study rolls) as opposed to consuming them in addition to their habitual diet. There were no restrictions on what study rolls could be consumed with and the rolls could be warmed before eating, if preferred. Compliance was encouraged and monitored via regular telephone and face-to-face contact with participants during the course of the intervention periods and participants were encouraged to return any uneaten rolls. A 3-day diary was completed during the final 3 days of each intervention period. At the end of each intervention period, participants attended for a study session identical to that described above. The intervention period would then be followed by a washout phase of 28 days before the participant rotated on to another intervention stage. ª 2014 The British Dietetic Association Ltd.
Inulin versus wheat germ and CVD risk
Measurements Pulse wave velocity PWV is a measure of arterial stiffness and a marker for CVD (Safar et al., 2002). We have recently demonstrated good inter and intra-individual variation in PWV recordings that validate its use in intervention studies such as the present study (Tripkovic et al., 2014). PWV measurements ran to a standardised protocol taking into account the recommendations of the First International Consensus Conference on the Clinical Applications of Arterial Stiffness (Van Bortel et al., 2002). Blood pressure was recorded on the nondominant arm of the participant with calculation of the mean, as well as mean arterial pressure and pulse pressure. The route of the arteries was then traced and measured: carotid pulse to suprasternal notch (proximal) and suprasternal notch to femoral or radial pulse (distal). A three-lead electrocardiogram was attached, followed by the mean blood pressure recording and the proximal and distal measurements entered into the PWV computer software (SPHYGMOCOR 2000, version 7.1; AtCor Medical Inc., Itasca, IL, USA). The pulse points were then recorded sequentially with the sensor placed flat against the pulse point. For the radial–carotid recording, the radial pulse was assessed first followed by the carotid; for the carotid–femoral recording, the carotid pulse was measured first then the femoral pulse. The recording was saved once a steady, consistent pattern of pulse waveforms was achieved and maintained for at least 10 cardiac cycles. Duplicate recordings were made for all PWV outcomes. Twenty-four hour ambulatory blood pressure measurements An ABPM-04 ambulatory blood pressure recorder was used (PWS Instruments, Meditech Ltd, Budapest, Hungary). Automatic recordings of blood pressure were taken by the equipment every 30 min between 06.00 h and 22.00 h and every 1 h between 22.00 h and 06.00 h (as recommended by the British Hypertension Society). After the 24-h monitoring period, data were analysed using CARDIOVISIONS software (Meditech Ltd). Biochemistry Blood samples were taken into sodium oxalate tubes for glucose analysis and dipotassium ethylenediaminetetraacetic acid tubes for insulin and lipid profile analysis. All blood samples were centrifuged for 10 min at 3000 g and 4 °C, with the respective aliquots stored at 80 °C until required. Plasma glucose and lipid profiles [total cholesterol, high-density lipoprotein (HDL)-cholesterol, non-esterified fatty acids (NEFA) and triacylglycerides] were measured using enzymatic colourimetric methods via commercially available kits for the ILab650 (Instrumentation Laboratory, Milan, Italy). Quality control variation for all 3
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
markers was ≤3.1%. Plasma insulin was measured by radioimmunoassay using a commercially available kit (Millipore Corporation, Billerica, MA, USA). The minimum detection level for insulin was 2 lU mL 1 (12 pM), QCs did not fall beyond two SDs. The mean inter-assay CV was 8.6%, with a mean intra-assay CV of 8.5%. Statistical analysis Food diaries were analysed using WINDIETS RESEARCH (2005 version; Robert Gordon University, Aberdeen, UK) to derive mean daily intakes of energy and macronutrients (including fibre) at baseline and the end of each intervention period. Statistical analysis was conducted using SPSS, version 16 (SPSS Inc., Chicago, IL, USA) with P < 0.05 considered statistical significanct unless stated otherwise. The data were explored for normal distribution and descriptive statistics were performed. Detailed statistical analysis focussed on the effects of the interventions (i.e. treatment effect, within-subject, using repeated measures analysis of variance or Friedmans dependant on distribution of data) in addition to analysing a time-treatment effect (percentage change from baseline per intervention; analysed with repeated measures or Friedman’s test).
Results Ten overweight males participated in the present study. Table 1 shows the anthropometric characteristics of participants. During the course of the study, no participants reported ill-effects (changes in bowel habits were documented but not considered serious by the participants) and no-one withdrew from the study. Anthropometrics and dietary analysis There were no significant changes in any anthropometric parameters when comparing intervention stages. No significant differences in energy or macronutrient intake were detected when comparing the intervention phases, nor when comparing each intervention from baseline. Glucose control and lipid status Table 2 shows the effect of the intervention on plasma markers of glucose control and lipid status. Plasma glucose concentration increased by 3.6% following the control diet but only by 0.4% and 1.5% following the wheat
Table 1 Anthropometric and demographic characteristics of the subjects during the intervention Baseline Mean (SD) Age (years) Weight (kg) BMI (kg m 2) WC (cm) Body fat (%)
39.80 (9.59) 101.62 (13.40) 30.20 (3.02) 106.45 (7.62) 28.54 (4.14)
Wheat fibre
Inulin
Control
Mean (SD)
%ΔBS
Mean (SD)
103.06 30.70 106.98 28.56
1.62 1.79 0.54 0.34
101.99 30.39 105.12 29.41
(12.23) (2.63) (7.90) (4.21)
(12.15) (2.71) (6.28) (4.48)
%ΔBS
0.57 0.74 1.14 3.09
Mean (SD)
101.69 30.23 105.89 28.45
(12.86) (2.88) (7.40) (4.32)
%ΔBS
0.17 0.13 0.49 0.17
Data are the mean (SD). No significant differences found.BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; WC, waist circumference; %ΔBS, the percentage change between baseline and the intervention. Table 2 Influence of the intervention on plasma markers of glucose control and lipid status Baseline Mean (SD Glucose control HOMA-IR HOMA-% b Insulin (pM) Glucose (mM) Lipid status TG (mM) TC (mM) NEFA (mM) HDL-C (mM)
3.40 120.65 83.91 5.45 1.65 5.17 0.45 1.22
(1.77) (52.16) (41.15) (0.31) (0.42) (0.67) (0.12) (0.27)
Wheat fibre
Inulin
Control
Mean (SD)
%ΔBS
Mean (SD)
%ΔBS
Mean (SD)
%ΔBS
3.85 138.88 95.04 5.45
(1.37) (48.28) (31.95) (0.37)
21.39 30.58 21.34 0.41
3.68 118.92 88.52 5.53
(1.68) (42.52) (37.60) (0.28)
14.1 8.71 12.47 1.55
3.63 113.84 86.37 5.64
(1.35) (43.56) (31.35) (0.37)
14.93 5.09 11.61 3.61
(0.94) (0.78) (0.08) (0.17)
37.27 5.24 1.61 4.59
2.08 5.15 0.48 1.24
(0.86) (0.77) (0.14) (0.29)
29.8 0.57 9.49 2.26
1.79 4.97 0.48 1.17
(0.61) (0.63) (0.15) (0.25)
14.77 3.42 12.93 3.32
2.23 4.88 0.42 1.14
Data are the mean (SD). No significant differences found.HDL-C, HDL-cholesterol; HOMA, homeostasis assessment model; IR, insulin resistance; NEFA, non-esterified fatty acids; TC, total cholesterol; TG, triacylglycerides; % b, b-cell function (insulin sensitivity); %ΔBS, the percentage change between baseline and the intervention.
4
ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
fibre and inulin treatments. Plasma insulin concentration, however, increased by 21% following wheat fibre compared to 12% in the control and inulin groups. A similar pattern was observed with the homeostasis assessment model (HOMA) results, with larger percentage changes in the wheat fibre group and similar changes in the inulin and control groups (Table 2). However, when comparing the absolute values between the post-intervention results, these ‘between-intervention’ differences were not significant for either treatment or time-treatment effects. Table 2 also shows markers of lipid status at baseline and post intervention. Even though no significant treatment effect was identified for any of the lipid parameters, increases in triacylglyceride levels were observed in all treatment groups: 14.8% for the control, 29.8% for inulin and 37.3% for the wheat fibre. By contrast, total cholesterol showed more moderate fluctuations of 5.24 to 0.57% (also not significant) across the different treatment groups. This effect appears to have been largely driven by (a) 15
changes in HDL-cholesterol, which also showed mean decreases for the wheat fibre and control groups and a mean increase for the inulin group. Changes in NEFA were not significantly different across interventions. They were also not different when comparing percentage changes between intervention, although, in this case, relatively large mean increases were seen after the control and inulin interventions (approximately 10%) compared to a mean decrease in the wheat fibre intervention of 1.6%. Oral glucose tolerance test measurements The OGTT time-course data (Fig. 1) show changes between the interventions for both glucose and insulin. The OGTT glucose curve for the wheat fibre (‘WF’) was lower than that of the other intervention phases and baseline but, overall, no significant changes were found. This is also illustrated by the area under the curve (AUC) data BS WF IN
Glucose (mM)
10
CON
5
0 FAST
15
30
45
60
75
90
105
120
75
90
105
120
Time point
(b) 1500
BS WF IN
1000
Insulin (pM)
CON
Figure 1 Influence of the intervention on (a) glucose and (b) insulin profiles during an oral glucose tolerance test. Data are the mean (SD). No significant differences between treatments. BS, baseline; WF, wheat fibre ; IN, inulin; CON, control.
ª 2014 The British Dietetic Association Ltd.
500
0 FAST
15
30
45
60
Time point –500
5
L. Tripkovic et al.
Glucose AUC (mM per 120 min)
(a) 1500
BS
(a)
WF
Percentage change from baseline for insulin AUC
Inulin versus wheat germ and CVD risk
IN CON 1000
500
100
50
0
–50
–100
0 BS
WF
IN
WF
CON
Intervention
(b) BS
Percentage change from baseline for glucose AUC
Insulin AUC (pM per 120 min)
(b) 150 000
WF IN CON 100 000
50 000
0 BS
WF
IN
CON
Intervention
IN
40
20
*
0
–20
–40 WF
CON
IN
CON
Intervention
Intervention
Figure 2 Influence of the intervention on (a) glucose and (b) insulin area under the curve profiles from the oral glucose tolerance test data. Data are the mean (SD). AUC, area under the curve; BS, baseline; WF, wheat fibre; IN, inulin; CON, control.
Figure 3 Influence of the intervention on individual (a) glucose and (b) insulin area under the curve% from baseline values. *Significant difference between wheat fibre and control, P = 0.03. AUC, area under the curve; WF, wheat fibre; IN, inulin; CON, control.
shown in Fig. 2. The mean (SD) glucose AUC was 868 (138) mM 1 120 min for wheat fibre, 930 (159) mM 1 120 min for inulin and 991 (200) mM 1 120 min for control (trend for significance at P = 0.08). However, when calculating the difference between the interventions based on percentage change from baseline, as shown in Fig. 3, there was a significant difference between wheat fibre and control (P = 0.03). Although the insulin AUC data appear to mimic the glucose AUC in terms of visual differences between interventions, the error bars indicate a far wider spread of data and are highly nonsignificant (P = 0.89). Similarly, the triacylglyceride and NEFA AUC results also showed substantial variance and no significant differences between groups (data not shown).
For both radial–carotid and carotid–femoral PWV, there were no significant differences between the three intervention stages. This is despite (compared to baseline) the inulin intervention inducing a 12.84% reduction in radial–carotid PWV compared to a 1.34% increase for the wheat fibre intervention and a 4.24% increase for the control intervention.
Whole body markers of vascular status The effect of the interventions on arterial stiffness (PWV) and blood pressure (24-h AMBP) are shown in Table 3. With blood pressure, no significant differences were found between the intervention groups for blood pressure averaged over the day (Table 3) and for measurements averaged during the day or the night (data not shown). 6
Discussion Overall, the present study does not support the hypothesis that the increased consumption of a nondigestible carbohydrate (inulin) or an insoluble fibre (wheat fibre) will beneficially impact on metabolic dysfunction associated with CVD and specifically insulin sensitivity and dyslipidaemia, as previously demonstrated with these types of carbohydrate. The current literature is concentrated on data from intervention studies investigating glycaemic regulation and lipid profile responses to fibre consumption. Within the present study, although there were no significant intervention effects, there were still a number of changes in key outcomes. All three interventions increased fasting insulin ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
Table 3 Influence of the intervention on measurements of 24-h ambulatory blood pressure and arterial stiffness (pulse wave velocity) Baseline Mean (SD) Blood pressure SBP (mmHg) DBP (mmHg) Pulse wave velocity RC-PWV CF-PWV PP
Wheat fibre Mean (SD)
Inulin %ΔBS
Mean (SD)
Control %ΔBS
Mean (SD)
%ΔBS
124.8 (9.5) 72.9 (9.9)
126.8 (8.0) 76.5 (8.7)
1.77 6.84
125.2 (11.6) 73.3 (12.2)
0.28 1.88
121.5 (11.2) 70.9 (10.8)
2.65 1.37
8.39 (1.31) 7.79 (1.78) 52.90 (8.01)
8.46 (1.25) 7.66 (1.42) 50.30 (8.04)
1.34 0.17 4.61
7.84 (1.25) 6.91 (0.68) 51.90 (9.02)
12.84 1.40 1.67
8.72 (1.76) 7.36 (1.21) 50.60 (8.40)
4.24 3.79 4.09
Data are the mean (SD). No significant differences found.CF, carotid–femoral PWV; DBP, diastolic blood pressure; PP, pulse pressure; RC, radial– carotid PWV; SBP, systolic blood pressure; %ΔBS, the percentage change between baseline and the intervention.
levels and HOMA, although the effects were more dramatic following the wheat fibre intervention; however, it is not known whether these changes to glucose control are indicative of any longer-term changes to glucose processing. Weickert et al. (2006) found a significant improvement in whole body insulin insensitivity via the gold standard euglycaemic hyperinsulinaemic clamp methodology in response to an insoluble fibre (31 g day 1 for 3 days) intervention in 17 overweight/obese subjects (with normal glucose metabolism). However, although positive effects of fibre consumption were demonstrated, this was at a level of intake beyond ‘normal’ eating habits and at double the intake of the present study. Conversely, Jenkins et al. (1999, 2002) showed that a high-fibre (wheat bran) diet had no significant effects on any markers of CVD in diabetic (type 2 diabetes mellitus) subjects. This included markers of the lipid profile, inflammation and glycaemic control. However, when analysing the response to the oral glucose tolerance test in the present study, there was a significant difference between the response for the wheat fibre and control interventions from baseline. The control results showed a 10.2% increase from baseline, whereas the wheat fibre intervention reduced glucose AUC by 2.5%. This is a relatively small reduction in glucose AUC but is magnified when compared with the control, with a 13% difference between the two interventions. This result may indicate a benefit of consuming wheat fibre compared to refined grain, and, in the absence of changes to any other markers measured, also appears to support the current evidence base suggesting that the effects of wheat fibre are largely centred on glycaemic control (Weickert et al., 2005, 2006). The available literature on inulin tends to focus on an effect on the lipid profile (Davidson & Maki, 1999; Jackson et al., 1999; Williams & Jackson, 2002; Letexier et al., 2003; Beylot, 2005; Russo et al., 2008). In the present study, there were no statistically significant effects on lipid status; however, a potentially detrimental effect of ª 2014 The British Dietetic Association Ltd.
wheat fibre was indicated with a 4% reduction in HDLcholesterol and a 37% increase in triacylglycerides. Following the inulin intervention, there was a modest increase in HDL-cholesterol (2%), whereas the total cholesterol level was increased by a marginal 0.5%. The triacylglyceride levels following each intervention are of concern because, at baseline, all values were in the ‘normal’ range of 12.3 m s 1) (Mattace-Raso et al., 2006) and the relatively good baseline levels of vascular health in these study participants may explain the lack of effects seen following the dietary intervention. This is despite recruiting participants on the basis of recognised CVD risk factors (see below) and highlights the limitations of such non-invasive screening techniques. Similarly, there were no significant changes to 24-h monitoring of blood pressure. In some studies, high-fibre diets were shown to have a positive effect on blood pressure; for example, Behall et al. (2006) conducted a study assessing the impact of a diet containing barley, brown rice and whole wheat in normotensive, hypercholesterolaemic men and women and, over the course of the intervention, the diets lowered systolic and diastolic blood pressure (and therefore mean arterial pressure) compared to baseline. Despite design similarities between the study by Behall et al. (2006) and the present studies (i.e. comparing insoluble fibre versus soluble fibre versus a combination of the two), the findings are contradictory, with no beneficial blood pressure effects of soluble or insoluble fibre being observed in the present study. We recognise that the present study has several limitations. This baseline health status of the participants was despite a recruitment strategy designed to select those individuals considered to be at greater CVD risk as a result of a large waist circumferences and a high BMI (Guh et al., 2009). The present study was sufficiently powered in terms of insulin sensitivity; however, it remains relatively small in terms of absolute numbers of participants and no a priori specification of potential covariates was included in the study design, although subsequent analysis did not identify any significant differences in anthropometric or dietary variables. Compliance is an issue that could confound the dietary data and, indeed, the intervention because we are relying on participants to consume the required bread rolls and complete dietary records. The level of fibre intake per day used in the present study (15 g day 1) was found to be tolerable in previous studies (Williams & Jackson, 2002) and is fairly reflective of the amount of fibre recommended in the UK diet (18 g day 1) and that consumed within the recommended amount of whole grains (48 g day 1). Therefore, the levels are physiological and a realistic representation of possible intakes. However, 8
higher amounts have been used in previous studies and demonstrated positive effects. Because the amounts used in the present study did not exert a dramatic effect, it is possible that synergistic effects of a number of nutrients (e.g. consuming the whole grain) are more likely to force a positive metabolic change. Future studies could address this by comparing various types of fibre to whole grain, perhaps with higher levels of intake and extended intervention periods in higher-risk subjects. The beneficial effects of fibres and whole grain could also be assessed with respect to preventing metabolic changes rather than correcting them. Conclusions Overall, we found no significant effect of increasing consumption of either inulin or wheat fibre on clinical and whole body markers of CVD risk in overweight male participants. However, an interesting reduction in the glucose AUC for wheat fibre may indicate a potential beneficial effect on glycaemic control. This effect does support previous findings and is worth pursuing in further dietary intervention trials with a focus on the potential mechanisms of action of individual types of fibre.
Conflict of interests, source of funding and authorship The authors declare that they have no conflict of interests. The SLOWCARB consortium provided funding via PhD studentships to LT and NCM. JKL, KHH and GSF designed the study. LT and NCM co-ordinated the intervention, collected and analysed samples, and processed the data. All authors critically reviewed the manuscript and approved the final version submitted for publication.
References Andersson, A., Tengblad, S., Karlstrom, B., Kamal-Eldin, A., Landberg, R., Basu, S., Aman, P. & Vessby, B. (2007) Whole-grain foods do not affect insulin sensitivity or markers of lipid peroxidation and inflammation in healthy, moderately overweight subjects. J. Nutr. 137, 1401–1407. Aune, D., Norat, T., Romundstad, P. & Vatten, L.J. (2013) Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Eur. J. Epidemiol. 28, 845–858. ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Behall, K.M., Scholfield, D.J. & Hallfrisch, J. (2006) Whole-grain diets reduce blood pressure in mildly hypercholesterolemic men and women. J. Am. Diet. Assoc. 106, 1445–1449. Beylot, M. (2005) Effects of inulin-type fructans on lipid metabolism in man and in animal models. Br. J. Nutr. 93 (Suppl. 1), S163–S168. Biesiekierski, J.R., Rosella, O., Rose, R., Liels, K., Barrett, J.S., Shepherd, S.J., Gibson, P.R. & Muir, J.G. (2011) Quantification of fructans, galacto-oligosacharides and other short-chain carbohydrates in processed grains and cereals. J. Hum. Nutr. Diet. 24, 154–176. Brighenti, F., Casiraghi, M.C., Canzi, E. & Ferrari, A. (1999) Effect of consumption of a ready-to-eat breakfast cereal containing inulin on the intestinal milieu and blood lipids in healthy male volunteers. Eur. J. Clin. Nutr. 53, 726–733. Brock, D.W., Davis, C.K., Irving, B.A., Rodriguez, J., Barrett, E.J., Weltman, A., Taylor, A.G. & Gaesser, G.A. (2006) A high-carbohydrate, high-fiber meal improves endothelial function in adults with the metabolic syndrome. Diabetes Care 29, 2313–2315. Brownlee, I.A., Moore, C., Chatfield, M., Richardson, D.P., Ashby, P., Kuznesof, S.A., Jebb, S.A. & Seal, C.J. (2010) Markers of cardiovascular risk are not changed by increased whole-grain intake: the WHOLEheart study, a randomised, controlled dietary intervention. Br. J. Nutr. 104, 125–134. Cho, S.S., Qi, L., Fahey, G.C. Jr & Klurfeld, D.M. (2013) Consumption of cereal fiber, mixtures of whole grains and bran, and whole grains and risk reduction in type 2 diabetes, obesity, and cardiovascular disease. Am. J. Clin. Nutr. 98, 594–619. Davidson, M.H. & Maki, K.C. (1999) Effects of dietary inulin on serum lipids. J. Nutr. 129, 1474S–1477S. Fernandes, J., Vogt, J. & Wolever, T.M. (2011) Inulin increases short-term markers for colonic fermentation similarly in healthy and hyperinsulinaemic humans. Eur. J. Clin. Nutr. 65, 1279–1286. Flamm, G., Glinsmann, W., Kritchevsky, D., Prosky, L. & Roberfroid, M. (2001) Inulin and oligofructose as dietary fiber: a review of the evidence. Crit. Rev. Food Sci. Nutr. 41, 353–362. Forcheron, F. & Beylot, M. (2007) Long-term administration of inulin-type fructans has no significant lipid-lowering effect in normolipidemic humans. Metabolism 56, 1093– 1098. Guh, D.P., Zhang, W., Bansback, N., Amarsi, Z., Birmingham, C.L. & Anis, A.H. (2009) The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health 9, 88. Jackson, K.G., Taylor, G.R., Clohessy, A.M. & Williams, C.M. (1999) The effect of the daily intake of inulin on fasting lipid, insulin and glucose concentrations in middle-aged men and women. Br. J. Nutr. 82, 23–30. Jenkins, D.J., Kendall, C.W., Vuksan, V., Augustin, L.S., Mehling, C., Parker, T., Vidgen, E., Lee, B., Faulkner, D.,
ª 2014 The British Dietetic Association Ltd.
Inulin versus wheat germ and CVD risk
Seyler, H., Josse, R., Leiter, L.A., Connelly, P.W. & Fulgoni, V. 3rd (1999) Effect of wheat bran on serum lipids: influence of particle size and wheat protein. J. Am. Coll. Nutr. 18, 159–165. Jenkins, D.J., Kendall, C.W., Augustin, L.S., Martini, M.C., Axelsen, M., Faulkner, D., Vidgen, E., Parker, T., Lau, H., Connelly, P.W., Teitel, J., Singer, W., Vandenbroucke, A.C., Leiter, L.A. & Josse, R.G. (2002) Effect of wheat bran on glycemic control and risk factors for cardiovascular disease in type 2 diabetes. Diabetes Care 25, 1522–1528. Jones, J.M. (2008) Mining whole grains for functional components. In Food Science and Technology Bulletin: Functional Foods. ed. G.R.Gibson, pp. 67–83. Reading: IFIS Publishing. Kaur, N. & Gupta, A.K. (2002) Applications of inulin and oligofructose in health and nutrition. J. Biosci. 27, 703–714. Letexier, D., Diraison, F. & Beylot, M. (2003) Addition of inulin to a moderately high-carbohydrate diet reduces hepatic lipogenesis and plasma triacylglycerol concentrations in humans. Am. J. Clin. Nutr. 77, 559–564. Mattace-Raso, F.U., Van Der Cammen, T.J., Hofman, A., Van Popele, N.M., Bos, M.L., Schalekamp, M.A., Asmar, R., Reneman, R.S., Hoeks, A.P., Breteler, M.M. & Witteman, J.C. (2006) Arterial stiffness and risk of coronary heart disease and stroke: the Rotterdam Study. Circulation 113, 657–663. Mellen, P.B., Walsh, T.F. & Herrington, D.M. (2008) Whole grain intake and cardiovascular disease: a meta-analysis. Nutr. Metab. Cardiovasc. Dis. 18, 283–290. Roberfroid, M.B. (2005) Introducing inulin-type fructans. Br. J. Nutr. 93(Suppl. 1), S13–S25. Robertson, M.D., Bickerton, A.S., Dennis, A.L., Vidal, H. & Frayn, K.N. (2005) Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am. J. Clin. Nutr. 82, 559–567. Russo, F., Chimienti, G., Riezzo, G., Pepe, G., Petrosillo, G., Chiloiro, M. & Marconi, E. (2008) Inulin-enriched pasta affects lipid profile and Lp(a) concentrations in Italian young healthy male volunteers. Eur. J. Nutr. 47, 453–459. Safar, H., Mourad, J.J., Safar, M. & Blacher, J. (2002) Aortic pulse wave velocity, an independent marker of cardiovascular risk. Arch. Mal. Coeur Vaiss. 95, 1215–1218. Tarini, J. & Wolever, T.M. (2010) The fermentable fibre inulin increases postprandial serum short-chain fatty acids and reduces free-fatty acids and ghrelin in healthy subjects. Appl. Physiol. Nutr. Metab. 35, 9–16. Tovar, A.R., Caamano Mdel, C., Garcia-Padilla, S., Garcia, O.P., Duarte, M.A. & Rosado, J.L. (2012) The inclusion of a partial meal replacement with or without inulin to a calorie restricted diet contributes to reach recommended intakes of micronutrients and decrease plasma triglycerides: a randomized clinical trial in obese Mexican women. Nutr. J. 11, 44. Tripkovic, L., Hart, K., Frost, G. & Lodge, J.K. (2014) Inter- and intra-individual variation in pulse wave velocity
9
Inulin versus wheat germ and CVD risk
measurements in a male population. J. Blood Press. Monit., Epub ahead of print PMID: 24842490 May 14 2014. in press. Van Bortel, L.M., Duprez, D., Starmans-Kool, M.J., Safar, M.E., Giannattasio, C., Cockcroft, J., Kaiser, D.R. & Thuillez, C. (2002) Clinical applications of arterial stiffness, Task Force III: recommendations for user procedures. Am. J. Hypertens. 15, 445–452. Weickert, M.O., Mohlig, M., Koebnick, C., Holst, J.J., Namsolleck, P., Ristow, M., Osterhoff, M., Rochlitz, H., Rudovich, N., Spranger, J. & Pfeiffer, A.F. (2005) Impact of
10
L. Tripkovic et al.
cereal fibre on glucose-regulating factors. Diabetologia 48, 2343–2353. Weickert, M.O., Mohlig, M., Schofl, C., Arafat, A.M., Otto, B., Viehoff, H., Koebnick, C., Kohl, A., Spranger, J. & Pfeiffer, A.F. (2006) Cereal fiber improves whole-body insulin sensitivity in overweight and obese women. Diabetes Care 29, 775–780. Williams, C.M. & Jackson, K.G. (2002) Inulin and oligofructose: effects on lipid metabolism from human studies. Br. J. Nutr. 87(Suppl. 2), S261–S264.
ª 2014 The British Dietetic Association Ltd.
RESEARCH PAPER The effects of a diet rich in inulin or wheat fibre on markers of cardiovascular disease in overweight male subjects L. Tripkovic,1 N. C. Muirhead,1 K. H. Hart,1 G. S. Frost1,2 & J. K. Lodge1,3 1
Faculty of Health and Medical Sciences,, Department of Nutritional Sciences, University of Surrey, Guildford, UK Department of Medicine, Imperial College, University of London, London, UK 3 Faculty of Health and Life Sciences, Northumbria University, Ellison Building, Newcastle-Upon-Tyne, UK 2
Keywords arterial stiffness, cardiovascular disease, dietary intervention, fibre, glucose control, humans, inulin, lipid status, wheat germ, whole grain. Correspondence John K. Lodge, Faculty of Health & Life Sciences, Ellison Building, Northumbria University, Newcastle-Upon-Tyne, Tyne & Wear NE1 8ST, UK. Tel.: +44 (0)191 243 7710 Fax: +44 (0)191 227 3519 E-mail: [email protected] How to cite this article Tripkovic L., Muirhead N.C., Hart K.H., Frost G.S. & Lodge J.K. (2014) The effects of a diet rich in inulin or wheat fibre on markers of cardiovascular disease in overweight male subjects. J Hum Nutr Diet. doi: 10.1111/jhn.12251
Abstract Background: Previous studies suggest that the beneficial health effects of a diet rich in whole grains could be a result of the individual fibres found in the grain. The present study aimed to investigate the influence of a diet high in either wheat fibre (as an example of an insoluble fibre) or inulin (a nondigestible carbohydrate) on markers of cardiovascular disease. Methods: Ten male participants classified as at higher risk of cardiovascular disease [mean (SD) body mass index 30.2 (3) kg m 2, mean (SD) waist circumference 106.4 (7) cm, mean (SD) age 39.8 (9) years] were recruited to a randomised, controlled, cross-over study comparing the consumption of bespoke bread rolls containing either inulin, wheat germ or refined grain (control) (15 g day 1) for 4 weeks with a 4-week washout period between each regime. At the end of each regime, participants underwent an oral glucose tolerance test (OGTT), measures of pulse wave velocity (PWV), 24-h ambulatory blood pressure (AMBP), plasma lipid status and markers of glucose control. Results: There was no difference in measures of glucose control, lipid status, 24-h AMBP or PWV after the intervention periods and no changes compared to baseline. There was no significant difference between OGTT glucose and insulin time profiles; however, there was a significant difference in area under the curves between the wheat fibre and control interventions when comparing change from baseline (control +10.2%, inulin +4.3%, wheat fibre 2.5%; P = 0.03). Conclusions: Only limited differences between the interventions were identified, perhaps as a consequence of the amount of fibre used and intervention length. The wheat germ intervention resulted in a significant reduction in glucose area under the curve, suggesting that this fibre may aid glucose control.
Introduction Increasing whole grain intake is considered to be associated with decreasing cardiovascular disease (CVD) (Mellen et al., 2008; Aune et al., 2013; Cho et al., 2013), although intervention studies have not been convincing (Andersson et al., 2007; Brownlee et al., 2010). Most intervention studies implementing an increase in whole grain intake ª 2014 The British Dietetic Association Ltd.
focus on the grains as a whole entity. However, data are becoming available to suggest the beneficial health effects of the individual fibres found within the grain, such as inulin and wheat fibre. Both inulin and wheat fibre are found specifically within the wheat kernel, with inulin being an example of a fructan that is a nondigestible soluble fibre, whereas wheat fibre is an example of an insoluble fibre. The amount of inulin and wheat fibre in 1
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
wholegrain wheat flour is very different, with approximately 0.6–1 g/100 g wheat for inulin but approximately 12 g/100 g of wheat fibre (Jones, 2008; Biesiekierski et al., 2011). Inulin is a generic term that covers a range of linear fructans with degrees of polymerisation, although, typically, a hydrolysed form obtained from Chicory is used in health studies (Flamm et al., 2001; Kaur & Gupta, 2002; Roberfroid, 2005). Evidence is available to suggest that these constituents of wholegrain may exert beneficial effects on cardiovascular health, thus lowering the risk of mortality and morbidity over time (Flamm et al., 2001; Kaur & Gupta, 2002). For example, wheat fibre may play an important role in improving peripheral insulin sensitivity because Weickert et al. (2006) demonstrated that whole-body glucose disposal was significantly improved in a group of overweight individuals who consumed greater quantities of insoluble fibre in an enriched bread compared to a control bread. By contrast, the health benefits of inulin appear to be broader but have been concentrated on their lipid-lowering effects (Davidson & Maki, 1999; Williams & Jackson, 2002; Letexier et al., 2003; Beylot, 2005; Russo et al., 2008), with studies showing significant reductions in plasma triacyglycerides and moderate reductions in total and low-density lipoprotein cholesterol (Jackson et al., 1999; Beylot, 2005). More recently, inulin has been shown to have a prebiotic effect, increasing markers of colonic fermentation (Fernandes et al., 2011) and affecting appetite (Tarini & Wolever, 2010). However, overall, the current body of evidence regarding inulin and wheat fibre and their respective potentials in exerting beneficial metabolic changes remains mixed and requires clarification. One question that remains is whether the separate fibres found within a whole grain kernel impart separate effects or whether they have an additive effect and therefore must be consumed together as an intact grain to be most effective. The present study aimed to investigate the effects of inulin or wheat fibre supplementation (15 g day 1; the mean intake of fibre in the UK and representing an intake that was shown to be well tolerated; Williams & Jackson, 2002) versus a refined grain control, on markers of CVD in men at an increased risk of the disease. Materials and methods Study design The study comprised a three-way, cross-over, controlled, randomised dietary intervention study in men aged 30– 55 years. Participants consumed (in a randomised order) bread rolls that contained either wheat fibre or inulin or refined wheat grain (control) on top of their habitual diet. The rolls were matched for nutrient profiles [per 100 g: wheat fibre 932 kJ, 43.4 g of carbohydrate (CHO), 2
8.5 g of protein, 1.3 g of fat, 7.7 g of fibre; inulin 1015 kJ, 47.4 g of CHO, 9.3 g of protein, 1.4 g of fat, 7.6 g of fibre; control 1088 kJ, 50.7 g of CHO, 9.9 g of protein, 1.5 g of fat, 2.5 g of fibre]. Participants were requested to consume three rolls per day, with each roll containing 5 g of the ‘active’ intervention ingredient (i.e. 15 g day 1); either wheat fibre, inulin or refined grain. The rolls were consumed for 28 days, followed by a 28day washout period immediately after. The participants attended four study sessions: one at the beginning (Day 0 – baseline) and then one at the end of each intervention stage (Study Day 1 to Day 28; Study Day 2 to Day 84; Study Day 3 to Day 140). The study received a favourable ethical opinion from the University of Surrey Ethics Committee (EC/2007/73/FHMS). Participants The changes in plasma insulin and glucose concentrations were the primary end-points of the study with sample size calculations based on the change in insulin. A sample size of 10 would have at least 80% power to detect a significant drop in insulin of 20 pM (a difference to detect of 25%) between the groups assuming that the common standard deviation is 16 pM or 20% (Robertson et al., 2005). Participants suitable for inclusion were adult males, aged 30–55 years, waist circumference >94 cm (37 inches), body mass index (BMI) of 25–35 kg m 2, not taking any prescription medicines or supplements within the past 6 months, not drinking more than 21 units of alcohol per week, and not regularly undertaking vigorous exercise or fitness training (no more than 3 9 30-min aerobic sessions per week). Participants were recruited from University staff and the surrounding area through e-mail and advertisement. Participants were randomly assigned to their starting intervention group and to the order in which they would receive the three intervention products using web-based randomisation software (http://www.randomization.com). Study protocol The week before commencing the study, participants completed a 3-day diet diary to record their dietary intake at baseline. The night before the study, the participants consumed a standard meal (chosen from a set list matched for macronutrient intake) in addition to refraining from strenuous exercise, alcohol or caffeine. Study session: all sessions All study sessions were held at the Clinical Investigation Unit, Faculty of Health and Medical Sciences, University ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
of Surrey and followed the same protocol. Participants arrived early in the morning (session start times between 07.30 h and 09.00 h) and in a fasted state (12-h fasted). Anthropometric measurements were taken first using standard methodologies. Height was measured via a wall-mounted Harpenden Stadiometer (Holtain Ltd, Crymych, UK) to the nearest 0.1 cm. Weight and body fat percentage were measured, and body mass index calculated, using a Tanita TBF-300 Body Composition Analyser (Tanita Europe BV, Amsterdam, The Netherlands). Waist and hip circumferences were measured using a flexible tape measure, to the nearest 0.1 cm, around the navel and widest part of the hips, respectively. Participants were then requested to rest in a quiet room with dimmed lighting for 10 min in preparation for the pulse wave velocity (PWV) measurements. After the resting phase, triplicate recordings of static, supine blood pressure were taken and the process of recording PWV commenced. The participant was then immediately cannulated and an initial 30-mL fasting baseline blood sample taken. An oral glucose tolerance test (OGTT) was then conducted. The participants consumed 75 g of anhydrous glucose dissolved in 400 mL of water, within the space of 10 min. With fasted blood samples already taken prior to consuming the glucose, further blood sampling of 5 mL continued every 15 min for a further 2 h. Once the blood sampling had been completed, participants were offered refreshments and were then fitted with the ambulatory blood pressure monitors (ABPM) to be worn for the next 24 h. Prior to leaving the Clinical Investigation Unit, participants were provided with an allocated quantity of bread rolls for the entire intervention period. Participants were required to consume their allocated rolls per day (containing 15 g of test ingredient) and were blinded to the contents of the roll. Participants were encouraged to incorporate the rolls into their habitual diet (e.g. by adapting their lunch by substituting their routine bread for the study rolls) as opposed to consuming them in addition to their habitual diet. There were no restrictions on what study rolls could be consumed with and the rolls could be warmed before eating, if preferred. Compliance was encouraged and monitored via regular telephone and face-to-face contact with participants during the course of the intervention periods and participants were encouraged to return any uneaten rolls. A 3-day diary was completed during the final 3 days of each intervention period. At the end of each intervention period, participants attended for a study session identical to that described above. The intervention period would then be followed by a washout phase of 28 days before the participant rotated on to another intervention stage. ª 2014 The British Dietetic Association Ltd.
Inulin versus wheat germ and CVD risk
Measurements Pulse wave velocity PWV is a measure of arterial stiffness and a marker for CVD (Safar et al., 2002). We have recently demonstrated good inter and intra-individual variation in PWV recordings that validate its use in intervention studies such as the present study (Tripkovic et al., 2014). PWV measurements ran to a standardised protocol taking into account the recommendations of the First International Consensus Conference on the Clinical Applications of Arterial Stiffness (Van Bortel et al., 2002). Blood pressure was recorded on the nondominant arm of the participant with calculation of the mean, as well as mean arterial pressure and pulse pressure. The route of the arteries was then traced and measured: carotid pulse to suprasternal notch (proximal) and suprasternal notch to femoral or radial pulse (distal). A three-lead electrocardiogram was attached, followed by the mean blood pressure recording and the proximal and distal measurements entered into the PWV computer software (SPHYGMOCOR 2000, version 7.1; AtCor Medical Inc., Itasca, IL, USA). The pulse points were then recorded sequentially with the sensor placed flat against the pulse point. For the radial–carotid recording, the radial pulse was assessed first followed by the carotid; for the carotid–femoral recording, the carotid pulse was measured first then the femoral pulse. The recording was saved once a steady, consistent pattern of pulse waveforms was achieved and maintained for at least 10 cardiac cycles. Duplicate recordings were made for all PWV outcomes. Twenty-four hour ambulatory blood pressure measurements An ABPM-04 ambulatory blood pressure recorder was used (PWS Instruments, Meditech Ltd, Budapest, Hungary). Automatic recordings of blood pressure were taken by the equipment every 30 min between 06.00 h and 22.00 h and every 1 h between 22.00 h and 06.00 h (as recommended by the British Hypertension Society). After the 24-h monitoring period, data were analysed using CARDIOVISIONS software (Meditech Ltd). Biochemistry Blood samples were taken into sodium oxalate tubes for glucose analysis and dipotassium ethylenediaminetetraacetic acid tubes for insulin and lipid profile analysis. All blood samples were centrifuged for 10 min at 3000 g and 4 °C, with the respective aliquots stored at 80 °C until required. Plasma glucose and lipid profiles [total cholesterol, high-density lipoprotein (HDL)-cholesterol, non-esterified fatty acids (NEFA) and triacylglycerides] were measured using enzymatic colourimetric methods via commercially available kits for the ILab650 (Instrumentation Laboratory, Milan, Italy). Quality control variation for all 3
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
markers was ≤3.1%. Plasma insulin was measured by radioimmunoassay using a commercially available kit (Millipore Corporation, Billerica, MA, USA). The minimum detection level for insulin was 2 lU mL 1 (12 pM), QCs did not fall beyond two SDs. The mean inter-assay CV was 8.6%, with a mean intra-assay CV of 8.5%. Statistical analysis Food diaries were analysed using WINDIETS RESEARCH (2005 version; Robert Gordon University, Aberdeen, UK) to derive mean daily intakes of energy and macronutrients (including fibre) at baseline and the end of each intervention period. Statistical analysis was conducted using SPSS, version 16 (SPSS Inc., Chicago, IL, USA) with P < 0.05 considered statistical significanct unless stated otherwise. The data were explored for normal distribution and descriptive statistics were performed. Detailed statistical analysis focussed on the effects of the interventions (i.e. treatment effect, within-subject, using repeated measures analysis of variance or Friedmans dependant on distribution of data) in addition to analysing a time-treatment effect (percentage change from baseline per intervention; analysed with repeated measures or Friedman’s test).
Results Ten overweight males participated in the present study. Table 1 shows the anthropometric characteristics of participants. During the course of the study, no participants reported ill-effects (changes in bowel habits were documented but not considered serious by the participants) and no-one withdrew from the study. Anthropometrics and dietary analysis There were no significant changes in any anthropometric parameters when comparing intervention stages. No significant differences in energy or macronutrient intake were detected when comparing the intervention phases, nor when comparing each intervention from baseline. Glucose control and lipid status Table 2 shows the effect of the intervention on plasma markers of glucose control and lipid status. Plasma glucose concentration increased by 3.6% following the control diet but only by 0.4% and 1.5% following the wheat
Table 1 Anthropometric and demographic characteristics of the subjects during the intervention Baseline Mean (SD) Age (years) Weight (kg) BMI (kg m 2) WC (cm) Body fat (%)
39.80 (9.59) 101.62 (13.40) 30.20 (3.02) 106.45 (7.62) 28.54 (4.14)
Wheat fibre
Inulin
Control
Mean (SD)
%ΔBS
Mean (SD)
103.06 30.70 106.98 28.56
1.62 1.79 0.54 0.34
101.99 30.39 105.12 29.41
(12.23) (2.63) (7.90) (4.21)
(12.15) (2.71) (6.28) (4.48)
%ΔBS
0.57 0.74 1.14 3.09
Mean (SD)
101.69 30.23 105.89 28.45
(12.86) (2.88) (7.40) (4.32)
%ΔBS
0.17 0.13 0.49 0.17
Data are the mean (SD). No significant differences found.BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; WC, waist circumference; %ΔBS, the percentage change between baseline and the intervention. Table 2 Influence of the intervention on plasma markers of glucose control and lipid status Baseline Mean (SD Glucose control HOMA-IR HOMA-% b Insulin (pM) Glucose (mM) Lipid status TG (mM) TC (mM) NEFA (mM) HDL-C (mM)
3.40 120.65 83.91 5.45 1.65 5.17 0.45 1.22
(1.77) (52.16) (41.15) (0.31) (0.42) (0.67) (0.12) (0.27)
Wheat fibre
Inulin
Control
Mean (SD)
%ΔBS
Mean (SD)
%ΔBS
Mean (SD)
%ΔBS
3.85 138.88 95.04 5.45
(1.37) (48.28) (31.95) (0.37)
21.39 30.58 21.34 0.41
3.68 118.92 88.52 5.53
(1.68) (42.52) (37.60) (0.28)
14.1 8.71 12.47 1.55
3.63 113.84 86.37 5.64
(1.35) (43.56) (31.35) (0.37)
14.93 5.09 11.61 3.61
(0.94) (0.78) (0.08) (0.17)
37.27 5.24 1.61 4.59
2.08 5.15 0.48 1.24
(0.86) (0.77) (0.14) (0.29)
29.8 0.57 9.49 2.26
1.79 4.97 0.48 1.17
(0.61) (0.63) (0.15) (0.25)
14.77 3.42 12.93 3.32
2.23 4.88 0.42 1.14
Data are the mean (SD). No significant differences found.HDL-C, HDL-cholesterol; HOMA, homeostasis assessment model; IR, insulin resistance; NEFA, non-esterified fatty acids; TC, total cholesterol; TG, triacylglycerides; % b, b-cell function (insulin sensitivity); %ΔBS, the percentage change between baseline and the intervention.
4
ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
fibre and inulin treatments. Plasma insulin concentration, however, increased by 21% following wheat fibre compared to 12% in the control and inulin groups. A similar pattern was observed with the homeostasis assessment model (HOMA) results, with larger percentage changes in the wheat fibre group and similar changes in the inulin and control groups (Table 2). However, when comparing the absolute values between the post-intervention results, these ‘between-intervention’ differences were not significant for either treatment or time-treatment effects. Table 2 also shows markers of lipid status at baseline and post intervention. Even though no significant treatment effect was identified for any of the lipid parameters, increases in triacylglyceride levels were observed in all treatment groups: 14.8% for the control, 29.8% for inulin and 37.3% for the wheat fibre. By contrast, total cholesterol showed more moderate fluctuations of 5.24 to 0.57% (also not significant) across the different treatment groups. This effect appears to have been largely driven by (a) 15
changes in HDL-cholesterol, which also showed mean decreases for the wheat fibre and control groups and a mean increase for the inulin group. Changes in NEFA were not significantly different across interventions. They were also not different when comparing percentage changes between intervention, although, in this case, relatively large mean increases were seen after the control and inulin interventions (approximately 10%) compared to a mean decrease in the wheat fibre intervention of 1.6%. Oral glucose tolerance test measurements The OGTT time-course data (Fig. 1) show changes between the interventions for both glucose and insulin. The OGTT glucose curve for the wheat fibre (‘WF’) was lower than that of the other intervention phases and baseline but, overall, no significant changes were found. This is also illustrated by the area under the curve (AUC) data BS WF IN
Glucose (mM)
10
CON
5
0 FAST
15
30
45
60
75
90
105
120
75
90
105
120
Time point
(b) 1500
BS WF IN
1000
Insulin (pM)
CON
Figure 1 Influence of the intervention on (a) glucose and (b) insulin profiles during an oral glucose tolerance test. Data are the mean (SD). No significant differences between treatments. BS, baseline; WF, wheat fibre ; IN, inulin; CON, control.
ª 2014 The British Dietetic Association Ltd.
500
0 FAST
15
30
45
60
Time point –500
5
L. Tripkovic et al.
Glucose AUC (mM per 120 min)
(a) 1500
BS
(a)
WF
Percentage change from baseline for insulin AUC
Inulin versus wheat germ and CVD risk
IN CON 1000
500
100
50
0
–50
–100
0 BS
WF
IN
WF
CON
Intervention
(b) BS
Percentage change from baseline for glucose AUC
Insulin AUC (pM per 120 min)
(b) 150 000
WF IN CON 100 000
50 000
0 BS
WF
IN
CON
Intervention
IN
40
20
*
0
–20
–40 WF
CON
IN
CON
Intervention
Intervention
Figure 2 Influence of the intervention on (a) glucose and (b) insulin area under the curve profiles from the oral glucose tolerance test data. Data are the mean (SD). AUC, area under the curve; BS, baseline; WF, wheat fibre; IN, inulin; CON, control.
Figure 3 Influence of the intervention on individual (a) glucose and (b) insulin area under the curve% from baseline values. *Significant difference between wheat fibre and control, P = 0.03. AUC, area under the curve; WF, wheat fibre; IN, inulin; CON, control.
shown in Fig. 2. The mean (SD) glucose AUC was 868 (138) mM 1 120 min for wheat fibre, 930 (159) mM 1 120 min for inulin and 991 (200) mM 1 120 min for control (trend for significance at P = 0.08). However, when calculating the difference between the interventions based on percentage change from baseline, as shown in Fig. 3, there was a significant difference between wheat fibre and control (P = 0.03). Although the insulin AUC data appear to mimic the glucose AUC in terms of visual differences between interventions, the error bars indicate a far wider spread of data and are highly nonsignificant (P = 0.89). Similarly, the triacylglyceride and NEFA AUC results also showed substantial variance and no significant differences between groups (data not shown).
For both radial–carotid and carotid–femoral PWV, there were no significant differences between the three intervention stages. This is despite (compared to baseline) the inulin intervention inducing a 12.84% reduction in radial–carotid PWV compared to a 1.34% increase for the wheat fibre intervention and a 4.24% increase for the control intervention.
Whole body markers of vascular status The effect of the interventions on arterial stiffness (PWV) and blood pressure (24-h AMBP) are shown in Table 3. With blood pressure, no significant differences were found between the intervention groups for blood pressure averaged over the day (Table 3) and for measurements averaged during the day or the night (data not shown). 6
Discussion Overall, the present study does not support the hypothesis that the increased consumption of a nondigestible carbohydrate (inulin) or an insoluble fibre (wheat fibre) will beneficially impact on metabolic dysfunction associated with CVD and specifically insulin sensitivity and dyslipidaemia, as previously demonstrated with these types of carbohydrate. The current literature is concentrated on data from intervention studies investigating glycaemic regulation and lipid profile responses to fibre consumption. Within the present study, although there were no significant intervention effects, there were still a number of changes in key outcomes. All three interventions increased fasting insulin ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Inulin versus wheat germ and CVD risk
Table 3 Influence of the intervention on measurements of 24-h ambulatory blood pressure and arterial stiffness (pulse wave velocity) Baseline Mean (SD) Blood pressure SBP (mmHg) DBP (mmHg) Pulse wave velocity RC-PWV CF-PWV PP
Wheat fibre Mean (SD)
Inulin %ΔBS
Mean (SD)
Control %ΔBS
Mean (SD)
%ΔBS
124.8 (9.5) 72.9 (9.9)
126.8 (8.0) 76.5 (8.7)
1.77 6.84
125.2 (11.6) 73.3 (12.2)
0.28 1.88
121.5 (11.2) 70.9 (10.8)
2.65 1.37
8.39 (1.31) 7.79 (1.78) 52.90 (8.01)
8.46 (1.25) 7.66 (1.42) 50.30 (8.04)
1.34 0.17 4.61
7.84 (1.25) 6.91 (0.68) 51.90 (9.02)
12.84 1.40 1.67
8.72 (1.76) 7.36 (1.21) 50.60 (8.40)
4.24 3.79 4.09
Data are the mean (SD). No significant differences found.CF, carotid–femoral PWV; DBP, diastolic blood pressure; PP, pulse pressure; RC, radial– carotid PWV; SBP, systolic blood pressure; %ΔBS, the percentage change between baseline and the intervention.
levels and HOMA, although the effects were more dramatic following the wheat fibre intervention; however, it is not known whether these changes to glucose control are indicative of any longer-term changes to glucose processing. Weickert et al. (2006) found a significant improvement in whole body insulin insensitivity via the gold standard euglycaemic hyperinsulinaemic clamp methodology in response to an insoluble fibre (31 g day 1 for 3 days) intervention in 17 overweight/obese subjects (with normal glucose metabolism). However, although positive effects of fibre consumption were demonstrated, this was at a level of intake beyond ‘normal’ eating habits and at double the intake of the present study. Conversely, Jenkins et al. (1999, 2002) showed that a high-fibre (wheat bran) diet had no significant effects on any markers of CVD in diabetic (type 2 diabetes mellitus) subjects. This included markers of the lipid profile, inflammation and glycaemic control. However, when analysing the response to the oral glucose tolerance test in the present study, there was a significant difference between the response for the wheat fibre and control interventions from baseline. The control results showed a 10.2% increase from baseline, whereas the wheat fibre intervention reduced glucose AUC by 2.5%. This is a relatively small reduction in glucose AUC but is magnified when compared with the control, with a 13% difference between the two interventions. This result may indicate a benefit of consuming wheat fibre compared to refined grain, and, in the absence of changes to any other markers measured, also appears to support the current evidence base suggesting that the effects of wheat fibre are largely centred on glycaemic control (Weickert et al., 2005, 2006). The available literature on inulin tends to focus on an effect on the lipid profile (Davidson & Maki, 1999; Jackson et al., 1999; Williams & Jackson, 2002; Letexier et al., 2003; Beylot, 2005; Russo et al., 2008). In the present study, there were no statistically significant effects on lipid status; however, a potentially detrimental effect of ª 2014 The British Dietetic Association Ltd.
wheat fibre was indicated with a 4% reduction in HDLcholesterol and a 37% increase in triacylglycerides. Following the inulin intervention, there was a modest increase in HDL-cholesterol (2%), whereas the total cholesterol level was increased by a marginal 0.5%. The triacylglyceride levels following each intervention are of concern because, at baseline, all values were in the ‘normal’ range of 12.3 m s 1) (Mattace-Raso et al., 2006) and the relatively good baseline levels of vascular health in these study participants may explain the lack of effects seen following the dietary intervention. This is despite recruiting participants on the basis of recognised CVD risk factors (see below) and highlights the limitations of such non-invasive screening techniques. Similarly, there were no significant changes to 24-h monitoring of blood pressure. In some studies, high-fibre diets were shown to have a positive effect on blood pressure; for example, Behall et al. (2006) conducted a study assessing the impact of a diet containing barley, brown rice and whole wheat in normotensive, hypercholesterolaemic men and women and, over the course of the intervention, the diets lowered systolic and diastolic blood pressure (and therefore mean arterial pressure) compared to baseline. Despite design similarities between the study by Behall et al. (2006) and the present studies (i.e. comparing insoluble fibre versus soluble fibre versus a combination of the two), the findings are contradictory, with no beneficial blood pressure effects of soluble or insoluble fibre being observed in the present study. We recognise that the present study has several limitations. This baseline health status of the participants was despite a recruitment strategy designed to select those individuals considered to be at greater CVD risk as a result of a large waist circumferences and a high BMI (Guh et al., 2009). The present study was sufficiently powered in terms of insulin sensitivity; however, it remains relatively small in terms of absolute numbers of participants and no a priori specification of potential covariates was included in the study design, although subsequent analysis did not identify any significant differences in anthropometric or dietary variables. Compliance is an issue that could confound the dietary data and, indeed, the intervention because we are relying on participants to consume the required bread rolls and complete dietary records. The level of fibre intake per day used in the present study (15 g day 1) was found to be tolerable in previous studies (Williams & Jackson, 2002) and is fairly reflective of the amount of fibre recommended in the UK diet (18 g day 1) and that consumed within the recommended amount of whole grains (48 g day 1). Therefore, the levels are physiological and a realistic representation of possible intakes. However, 8
higher amounts have been used in previous studies and demonstrated positive effects. Because the amounts used in the present study did not exert a dramatic effect, it is possible that synergistic effects of a number of nutrients (e.g. consuming the whole grain) are more likely to force a positive metabolic change. Future studies could address this by comparing various types of fibre to whole grain, perhaps with higher levels of intake and extended intervention periods in higher-risk subjects. The beneficial effects of fibres and whole grain could also be assessed with respect to preventing metabolic changes rather than correcting them. Conclusions Overall, we found no significant effect of increasing consumption of either inulin or wheat fibre on clinical and whole body markers of CVD risk in overweight male participants. However, an interesting reduction in the glucose AUC for wheat fibre may indicate a potential beneficial effect on glycaemic control. This effect does support previous findings and is worth pursuing in further dietary intervention trials with a focus on the potential mechanisms of action of individual types of fibre.
Conflict of interests, source of funding and authorship The authors declare that they have no conflict of interests. The SLOWCARB consortium provided funding via PhD studentships to LT and NCM. JKL, KHH and GSF designed the study. LT and NCM co-ordinated the intervention, collected and analysed samples, and processed the data. All authors critically reviewed the manuscript and approved the final version submitted for publication.
References Andersson, A., Tengblad, S., Karlstrom, B., Kamal-Eldin, A., Landberg, R., Basu, S., Aman, P. & Vessby, B. (2007) Whole-grain foods do not affect insulin sensitivity or markers of lipid peroxidation and inflammation in healthy, moderately overweight subjects. J. Nutr. 137, 1401–1407. Aune, D., Norat, T., Romundstad, P. & Vatten, L.J. (2013) Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Eur. J. Epidemiol. 28, 845–858. ª 2014 The British Dietetic Association Ltd.
L. Tripkovic et al.
Behall, K.M., Scholfield, D.J. & Hallfrisch, J. (2006) Whole-grain diets reduce blood pressure in mildly hypercholesterolemic men and women. J. Am. Diet. Assoc. 106, 1445–1449. Beylot, M. (2005) Effects of inulin-type fructans on lipid metabolism in man and in animal models. Br. J. Nutr. 93 (Suppl. 1), S163–S168. Biesiekierski, J.R., Rosella, O., Rose, R., Liels, K., Barrett, J.S., Shepherd, S.J., Gibson, P.R. & Muir, J.G. (2011) Quantification of fructans, galacto-oligosacharides and other short-chain carbohydrates in processed grains and cereals. J. Hum. Nutr. Diet. 24, 154–176. Brighenti, F., Casiraghi, M.C., Canzi, E. & Ferrari, A. (1999) Effect of consumption of a ready-to-eat breakfast cereal containing inulin on the intestinal milieu and blood lipids in healthy male volunteers. Eur. J. Clin. Nutr. 53, 726–733. Brock, D.W., Davis, C.K., Irving, B.A., Rodriguez, J., Barrett, E.J., Weltman, A., Taylor, A.G. & Gaesser, G.A. (2006) A high-carbohydrate, high-fiber meal improves endothelial function in adults with the metabolic syndrome. Diabetes Care 29, 2313–2315. Brownlee, I.A., Moore, C., Chatfield, M., Richardson, D.P., Ashby, P., Kuznesof, S.A., Jebb, S.A. & Seal, C.J. (2010) Markers of cardiovascular risk are not changed by increased whole-grain intake: the WHOLEheart study, a randomised, controlled dietary intervention. Br. J. Nutr. 104, 125–134. Cho, S.S., Qi, L., Fahey, G.C. Jr & Klurfeld, D.M. (2013) Consumption of cereal fiber, mixtures of whole grains and bran, and whole grains and risk reduction in type 2 diabetes, obesity, and cardiovascular disease. Am. J. Clin. Nutr. 98, 594–619. Davidson, M.H. & Maki, K.C. (1999) Effects of dietary inulin on serum lipids. J. Nutr. 129, 1474S–1477S. Fernandes, J., Vogt, J. & Wolever, T.M. (2011) Inulin increases short-term markers for colonic fermentation similarly in healthy and hyperinsulinaemic humans. Eur. J. Clin. Nutr. 65, 1279–1286. Flamm, G., Glinsmann, W., Kritchevsky, D., Prosky, L. & Roberfroid, M. (2001) Inulin and oligofructose as dietary fiber: a review of the evidence. Crit. Rev. Food Sci. Nutr. 41, 353–362. Forcheron, F. & Beylot, M. (2007) Long-term administration of inulin-type fructans has no significant lipid-lowering effect in normolipidemic humans. Metabolism 56, 1093– 1098. Guh, D.P., Zhang, W., Bansback, N., Amarsi, Z., Birmingham, C.L. & Anis, A.H. (2009) The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health 9, 88. Jackson, K.G., Taylor, G.R., Clohessy, A.M. & Williams, C.M. (1999) The effect of the daily intake of inulin on fasting lipid, insulin and glucose concentrations in middle-aged men and women. Br. J. Nutr. 82, 23–30. Jenkins, D.J., Kendall, C.W., Vuksan, V., Augustin, L.S., Mehling, C., Parker, T., Vidgen, E., Lee, B., Faulkner, D.,
ª 2014 The British Dietetic Association Ltd.
Inulin versus wheat germ and CVD risk
Seyler, H., Josse, R., Leiter, L.A., Connelly, P.W. & Fulgoni, V. 3rd (1999) Effect of wheat bran on serum lipids: influence of particle size and wheat protein. J. Am. Coll. Nutr. 18, 159–165. Jenkins, D.J., Kendall, C.W., Augustin, L.S., Martini, M.C., Axelsen, M., Faulkner, D., Vidgen, E., Parker, T., Lau, H., Connelly, P.W., Teitel, J., Singer, W., Vandenbroucke, A.C., Leiter, L.A. & Josse, R.G. (2002) Effect of wheat bran on glycemic control and risk factors for cardiovascular disease in type 2 diabetes. Diabetes Care 25, 1522–1528. Jones, J.M. (2008) Mining whole grains for functional components. In Food Science and Technology Bulletin: Functional Foods. ed. G.R.Gibson, pp. 67–83. Reading: IFIS Publishing. Kaur, N. & Gupta, A.K. (2002) Applications of inulin and oligofructose in health and nutrition. J. Biosci. 27, 703–714. Letexier, D., Diraison, F. & Beylot, M. (2003) Addition of inulin to a moderately high-carbohydrate diet reduces hepatic lipogenesis and plasma triacylglycerol concentrations in humans. Am. J. Clin. Nutr. 77, 559–564. Mattace-Raso, F.U., Van Der Cammen, T.J., Hofman, A., Van Popele, N.M., Bos, M.L., Schalekamp, M.A., Asmar, R., Reneman, R.S., Hoeks, A.P., Breteler, M.M. & Witteman, J.C. (2006) Arterial stiffness and risk of coronary heart disease and stroke: the Rotterdam Study. Circulation 113, 657–663. Mellen, P.B., Walsh, T.F. & Herrington, D.M. (2008) Whole grain intake and cardiovascular disease: a meta-analysis. Nutr. Metab. Cardiovasc. Dis. 18, 283–290. Roberfroid, M.B. (2005) Introducing inulin-type fructans. Br. J. Nutr. 93(Suppl. 1), S13–S25. Robertson, M.D., Bickerton, A.S., Dennis, A.L., Vidal, H. & Frayn, K.N. (2005) Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am. J. Clin. Nutr. 82, 559–567. Russo, F., Chimienti, G., Riezzo, G., Pepe, G., Petrosillo, G., Chiloiro, M. & Marconi, E. (2008) Inulin-enriched pasta affects lipid profile and Lp(a) concentrations in Italian young healthy male volunteers. Eur. J. Nutr. 47, 453–459. Safar, H., Mourad, J.J., Safar, M. & Blacher, J. (2002) Aortic pulse wave velocity, an independent marker of cardiovascular risk. Arch. Mal. Coeur Vaiss. 95, 1215–1218. Tarini, J. & Wolever, T.M. (2010) The fermentable fibre inulin increases postprandial serum short-chain fatty acids and reduces free-fatty acids and ghrelin in healthy subjects. Appl. Physiol. Nutr. Metab. 35, 9–16. Tovar, A.R., Caamano Mdel, C., Garcia-Padilla, S., Garcia, O.P., Duarte, M.A. & Rosado, J.L. (2012) The inclusion of a partial meal replacement with or without inulin to a calorie restricted diet contributes to reach recommended intakes of micronutrients and decrease plasma triglycerides: a randomized clinical trial in obese Mexican women. Nutr. J. 11, 44. Tripkovic, L., Hart, K., Frost, G. & Lodge, J.K. (2014) Inter- and intra-individual variation in pulse wave velocity
9
Inulin versus wheat germ and CVD risk
measurements in a male population. J. Blood Press. Monit., Epub ahead of print PMID: 24842490 May 14 2014. in press. Van Bortel, L.M., Duprez, D., Starmans-Kool, M.J., Safar, M.E., Giannattasio, C., Cockcroft, J., Kaiser, D.R. & Thuillez, C. (2002) Clinical applications of arterial stiffness, Task Force III: recommendations for user procedures. Am. J. Hypertens. 15, 445–452. Weickert, M.O., Mohlig, M., Koebnick, C., Holst, J.J., Namsolleck, P., Ristow, M., Osterhoff, M., Rochlitz, H., Rudovich, N., Spranger, J. & Pfeiffer, A.F. (2005) Impact of
10
L. Tripkovic et al.
cereal fibre on glucose-regulating factors. Diabetologia 48, 2343–2353. Weickert, M.O., Mohlig, M., Schofl, C., Arafat, A.M., Otto, B., Viehoff, H., Koebnick, C., Kohl, A., Spranger, J. & Pfeiffer, A.F. (2006) Cereal fiber improves whole-body insulin sensitivity in overweight and obese women. Diabetes Care 29, 775–780. Williams, C.M. & Jackson, K.G. (2002) Inulin and oligofructose: effects on lipid metabolism from human studies. Br. J. Nutr. 87(Suppl. 2), S261–S264.
ª 2014 The British Dietetic Association Ltd.
