High intake of added sweeteners is considered to have a causal role in the pathogenesis of cardiometabolic disorders. Especially, high-fructose intake is regarded as potentially harmful to cardiometabolic health. It may cause not only weight gain but also low-grade inflammation, which represents an independent risk factor for developing type 2 diabetes and cardiovascular disease. In particular, fructose has been suggested to induce plasminogen activator inhibitor-1 (PAI-1) expression in the liver and to increase circulating inflammatory cytokines. We therefore aimed to investigate, whether high-fructose diet has an impact on PAI-1, monocyte chemoattractant protein-1 (MCP-1), e-selectin and C-reactive protein (CRP) concentrations in healthy humans.
We studied 20 participants (12 males and 8 females) of the TUebingen FRuctose Or Glucose study. This is an exploratory, parallel, prospective, randomized, single-blinded, outpatient, hypercaloric, intervention study. The participants had a mean age of 30.9±2.1 years and a mean body mass index of 26.0±0.5 kg/m2 and they received 150 g of either fructose or glucose per day for 4 weeks.
There were neither significant changes of PAI-1, MCP-1, e-selectin and CRP after fructose (n=10) and glucose (n=10) intervention nor treatment effects (all P>0.2). Moreover, we did not observe longitudinal associations of the inflammatory parameters with triglycerides, liver fat, visceral fat and body weight in the fructose group.
Temporary high-fructose intake does not seem to cause inflammation in apparently healthy people in this secondary analysis of a small feeding trial.
There is a broad evidence that intake of large amounts of added sweeteners causes obesity,1, 2 which is strongly associated with the risk of developing type 2 diabetes and cardiovascular disease.3, 4, 5 Added sweeteners consist either of sucrose, a disaccharide composed of fructose and glucose, or of high-fructose corn sirup, a mixture of fructose and glucose in monosaccharide form.6 It is currently under debate, whether particularly high-fructose intake, for example, by causing visceral obesity and hyperuricaemia, may confer increased cardiometabolic risk.6, 7, 8, 9, 10, 11 Most of the evidence supporting an adverse effect of fructose on cardiometabolic health derives from animal studies or cross-sectional observations. Interventional controlled studies in humans are scarce.11 Moreover, many studies supporting an adverse effect of fructose on cardiometabolic risk tend not to adjust for energy intake, the important confounding factor and potentially main causal link in the understanding of this issue.11
Some recent studies have suggested that fructose rather than glucose may cause low-grade inflammation. An animal study has indicated that the possible adverse metabolic effects of fructose may in part be mediated by the induction of plasminogen activator inhibitor-1 (PAI-1).12 Moreover, high-dietary fructose intake has been suggested to induce PAI-1 expression in the liver13 and, possibly resulting from an accumulating visceral fat, to increase circulating markers of inflammation including PAI-1 and C-reactive protein (CRP) in humans.14, 15 On the other hand, low-fructose diet appeared to reduce inflammation in humans with chronic kidney disease.16 These findings seemed of relevance as chronic inflammation increases the risk of future diabetes and cardiovascular disease.17, 18 However, the authors have explicitly pointed out the need to replicate their findings in other cohorts with different patient characteristics.14
Thus, we aimed to investigate whether high-fructose intake is associated with an increase of PAI-1, monocyte chemoattractant protein-1 (MCP-1), e-selectin and CRP levels in an exploratory secondary analysis of the TUebingen FRuctose Or Glucose study (TUFROG) study.19, 20
Materials and methods
Study design and diet
TUFROG is an exploratory, parallel, prospective, randomized, single-blinded, outpatient, hypercaloric, intervention study.19 Inclusion criteria were: age 20.50 years, body mass index 20–35 kg/m2, physical health and not more than 1 h sports/week. Exclusion criteria were: pregnancy, any relevant illness (that is, diabetes, dyslipidaemia, endocrine disease, coronary artery disease, malignancy, gastrointestinal disease and psychological disease), fructose intolerance, medication, metal implants (for example, pacemaker and metal heart valve), regular alcohol consumption 10 g/d and claustrophobia.19 The dietary intervention consisted of either 150 g (2512 kJ (600 kcal)/d) of fructose or 150 g of glucose (2512 kJ (600 kcal)/d) for 4 weeks. This made up about additional 20–25% of the participants energy intake. The participants were instructed to consume the sugar, which had to be dissolved in water (50 g sugar in 250 ml water), in addition to a balanced weight-maintaining diet (50% carbohydrates, 35% fat and 15% protein). Specific recommendations with regard to dietary fiber intake were not made. Fructose or glucose was ingested three times daily with the main meals.19 The participants had to visit the study laboratory three times. There was a screening examination at the beginning of the study and there were two main examinations (blood withdrawal, magnetic resonance imaging and spectroscopy) at 2 and 6 weeks after the screening examination. Moreover, study participants were called after 2 weeks intervention to determine whether the fructose or glucose was well tolerated and regularly consumed.19 At the screening examination, the decision about inclusion of a participant was made by a physician based on the inclusion–exclusion criteria mentioned above. After the run-in phase of 2 weeks between screening examination and main examination 1, in which the subjects were instructed to keep an isoenergetic diet (50% carbohydrates, 35% fat and 15% protein), the participants were randomly allocated to the fructose or glucose intervention group.19 Dietary compliance was controlled via close telephone contact. Moreover, compliance was evaluated by interviews at main examinations 1 and 2 and the subjects were asked to fill out food intake records on 3 days weekly.19 The study was approved by the Ethics Committee of the University of Tübingen and was conducted in accordance with the Declaration of Helsinki. Informed written consent was obtained from all participants.
Blood samples were taken in the early morning after a 10-h overnight fast. All routine analytical procedures including CRP were performed immediately within 1 h after blood collection at the central laboratory facility of the University Hospital Tübingen,19 Germany. The laboratory has an accreditation according to DIN EN ISO 15189. Internal and external quality controls were always within the allowed ranges. Samples for the determination of inflammatory cytokines were centrifuged (2200 g, 7 min) directly after the blood sample was drawn. Plasma and serum aliquots were generated without delay, and they were directly frozen in cryotubes (Nunc, Thermo Scientific, Langenselbold, Germany) at −80 °C until further analysis. CRP concentrations were determined using a wide-range latex-enhanced turbidimetric assay on the ADVIA 1800 clinical chemistry analyzer (Siemens Healthcare Diagnostics, Eschborn, Germany). The coefficient of variation was 1.5% at 0.65 mg/dl and 1.0% at 4.7 mg/dl. All other inflammatory cytokines were measured by ELISA (R&D Systems, Wiesbaden, Germany) in a single batch only using non-re-frozen aliquots. We therefore do not have any information on the day-to-day variation of these tests. However, duplicate measurements revealed an intra-assay variation of 5.6% for PAI-1, 3.2% for MCP-1 and 5.3% for e-selectin.
Quantification of visceral and liver fat
The baseline characteristics are presented as means±standard errors of the means. Comparisons of the baseline characteristics between the fructose and glucose group were performed with the t-test.19 Changes in response to the 4-week high-hexose diets were studied with the paired samples t-test.19 All variables were transformed logarithmically before being used in parametric statistical models. Analysis of covariance (ANCOVA) was used to compare the changes of the inflammatory parameters between the fructose and glucose intervention groups, with study group as the main factor and the inflammatory parameter of interest at baseline as covariate (two-sided tests). To estimate the treatment effect, differences in least-squares means and the corresponding 95% confidence intervals were calculated based on the ANCOVA models.19 To test for longitudinal associations Pearson correlation coefficients among absolute changes were calculated. P-values <0.05 were considered significant. As this is an exploratory trial, we did not corrected for multiple testing. A per protocol analysis was performed. The JMP statistical software package 4.0 (SAS Institute, Cary, NC, USA) was used.
The detailed baseline characteristics of the TUFROG participants have been described.19 In short, the 10 participants (7 males and 3 females), who completed the fructose intervention, had a mean age of 32.8 years, body mass index of 25.5 kg/m2, visceral fat of 2.14 kg, liver fat of 1.32% and triglycerides of 80 mg/dl.19 The 10 participants (5 males and 5 females), who completed the glucose intervention, had a mean age of 28.2 years, body mass index of 26.2 kg/m2, visceral fat of 2.25 kg, liver fat of 1.59% and triglycerides of 98 mg/dl.19 The baseline concentration of circulating CRP, PAI-1, MCP-1 and E-selectin is shown in Table 1. There were no significant differences in the baseline characteristics between the fructose and the glucose groups (Table 1).
Changes in response to very-high-fructose and -glucose diets
A detailed description of the metabolic changes in response to very-high-fructose and very-high-glucose diets has been published.19 In short, weight significantly increased in the glucose group (+1.7 kg) but not in the fructose group (+0.2 kg). However, there was no statistically significant difference between interventions.19 Visceral fat and liver fat did neither significantly change in the fructose group (+0.07 kg and +0.45%, respectively), nor in the glucose group (+0.07 kg and +0.52%, respectively).19 Triglycerides significantly increased in the fructose (+35 mg/dl) but not in the glucose group (±0 mg/dl) with the difference between interventions reaching statistical significance.19 Circulating CRP, PAI-1, MCP-1 and E-selectin did not significantly change in response to 4-week very-high-fructose or -glucose diets and there were no treatment effects (Table 1 and Figure 1).
There was no significant longitudinal association of PAI-1 with triglycerides (r=0.28, P=0.438), intrahepatic lipids (r=−0.32, P=0.326), visceral fat (r=0.23, P=0.590) and body weight (r=0.25, P=0.948) during the fructose intervention. There were no significant longitudinal association of CRP, MCP-1 and E-selectin with triglycerides, intrahepatic lipids, visceral fat and body weight during fructose intervention either.
The present study did not reveal any evidence that a 4-week hypercaloric fructose diet causes inflammation as reflected by alterations of CRP, PAI-1, MCP-1 and E-selectin concentrations in healthy middle-aged persons.
The findings in relation to other studies
Three previous studies have investigated the impact of fructose intake on CRP concentrations. Aeberli et al.15 found that sugar sweetened beverages containing just 40 g of fructose increased CRP within 3 weeks in healthy young men when consumed daily. On the other hand, Brymora et al.16 reported a reduction of CRP in elderly patients with chronic kidney disease after receiving a diet low in fructose. In contrast, Cox et al.14 reported CRP concentrations to be unaffected by 10-week supplementation of fructose at 25% of energy requirements in a study including 16 overweight and obese middle-aged subjects. In agreement with the latter study, fructose supplementation did not increase CRP levels in the participants of the TUFROG cohort. Cox et al.14 however, reported that high-fructose intake may cause an increase of circulating PAI-1, MCP-1 and E-selectin. Such an observation was not made in a cross-sectional analysis presented by Thuy et al.13 In the TUFROG cohort, high-fructose diet had no significant effects on circulating PAI-1, MCP-1 and E-selectin. There were no significant longitudinal associations of inflammatory parameters with triglycerides, liver fat, visceral fat and body weight either. The difference between our findings compared with the study by Cox et al.14 may be due to the fact that the participants of the TUFROG study were younger and had lower baseline cardiometabolic risk. Moreover, they did not significantly gain body weight and visceral fat, which is regarded as crucial in mediating adverse metabolic effects of fructose.23 Finally, the duration of the intervention was longer in the study by Cox et al.14 than in the TUFROG cohort.
The fructose intervention, except from an increase in triglycerides, obviously did not cause any harm compared with the same amount of additional glucose. Consequently, there seems to be no clear indication that public health efforts should specifically aim at a decrease in fructose intake. Rather, energy intake in total and from added sugars should be reduced.24
Our study is small, the duration of the intervention was relatively short, and it was exploratory. Hence, the findings need confirmation from larger cohorts. We cannot exclude that long-term high-fructose intake may induce PAI-1, MCP-1, e-selectin and CRP, either.
As the TUFROG participants were relatively young and healthy, we cannot rule out adverse cardiometabolic effects in people with higher basal risk.
We did not use a cross-over design, which may be superior to a parallel design.
TUFROG is an outpatient study, which inevitably carries the risk of malcompliance. However, we aimed to prevent malcompliance by close telephone contact. Moreover, the participants were asked to fill out food intake records.19
Additional intake of 150 g of added sweeteners according to a report of NHANES data represents a very high dose.25 Thus, our results may not be generalizable to the general population.
Infections may have caused confounding. We did not directly ask for infections during the telephone visits or the final examination. However, we asked whether there were any problems during the 4 weeks of intervention. None of the participants reported on relevant infectious disease during the intervention period.
The major strength of the TUFROG study is the fact that we used a randomized controlled design. Thus, we were able to control for possible adverse cardiometabolic effects of a hypercaloric diet rich in monosacharides. Moreover, we want to highlight the detailed investigations including magnetic resonance imaging and magnetic resonance spectroscopy.
Temporary high-fructose intake does not seem to create relevant health risks in apparently healthy people in this secondary analysis of a small feeding trial.
We thank Anna Bury, Barbara Horrer, Ellen Kollmar, Andreas Vosseler and Heike Runge for expert technical assistance. The study was supported by a grant (Grant no. 4 AI) from the Zentrum Ernährungsmedizin Tübingen-Hohenheim. Funding did not include industrial sponsorship.
About this article
Scientific Reports (2016)