COMMENTARY (SEE AEBERLI ET AL., P. 150)

Potential Health Risks From Beverages Containing Found in or High-Fructose Corn

ructose isasweet tastingsugar that is beverages. In one study there was an in- relative to and that free fatty Ffound naturally in fruits and some crease in body weight, blood pressure, were increased or showed a trend toward vegetables and has been part of the and inflammatory markers (14,15), and an increase in the fructose beverage groups. human diet—in modest amounts—for in a second study there was an increase This article has several strengths, one eons. The increasing consumption of in triglycerides levels (particularly at of which is that it is a randomized sugar has dramatically increased our ex- night), a stimulation of de novo lipogen- crossover comparison of four beverages posure to fructose (1). Sugar consumption esis, and an increase in visceral fat with two levels of fructose, glucose, and has risen more than 40-fold since the Dec- (16,17). In the third study, which com- (50% fructose). Another strength laration of Independence was signed 250 pared milk, diet , a sugar-sweetened is that the study used modest amounts of years ago, and more than 40% of the added cola, and , the sugar-sweetened bev- fructose and had a glucose control. One in our diet are in sugar-sweetened erage increased liver fat, visceral fat, and limitation is that it had only a small beverages and fruit (2,3). Thus, the triglycerides over the 6 months of bev- number of subjects and that they were principal sources of fructose in our diet are erage intake (18). The latter study sug- all male, so we cannot be absolutely sure now sugar and high-fructose , gests that consuming two 16-ounce that these results extrapolate to females. each of which has about 50% fructose. sugar-containing beverages per day for 6 The authors did not find any effect on The intake of soft drinks has risen fivefold months can mimic many of the features of fasting triglycerides. However, they did since 1950 (4,5) (Fig.1) and with it the the and nonalco- not design the study to look at post- intake of fructose. The rise in the con- holic fatty liver disease. prandial or nocturnal levels of triglycer- sumption of high-fructose corn syrup in ides where they might have detected beverages has paralleled the rise in the Brief overview —The article by differences. In the comparison of the prevalence of and the metabolic Aeberli et al. (19) in this issue of Diabetes effect of glucose, fructose, and sucrose syndrome and is associated with the ap- Care and their previous study (20) have on plasma triglycerides, Cohen and Schall pearance of nonalcoholic fatty liver dis- (21) found that both fructose in the – added important data on the responses ease (6 8). Although association does to fructose. They conducted a 4-week ran- amount found in sucrose AND sucrose not prove causation, it has stimulated re- domized crossover study with a 4-week increased triglycerides following a meal, search to understand whether current lev- wash-out between each diet in 9 healthy but that glucose did not—leading them to els of fructose intake in beverages pose a young men comparing 4 different soft conclude that the effects on lipids were due health risk. drinks with levels of fructose, glucose, to the fructose either alone or as part of and sucrose that are closer to “normal” sucrose (table sugar), and not glucose. Background —Over the past decade intake than some other studies. The low- This study adds to the information fructose from either sucrose or high- fructose beverage had 40 g per day of fruc- about the role of fructose either from sucrose fructose corn syrup has received growing tose, which was the same amount of (ordinary table sugar) or from high-fructose attention as it has been associated with a fructose as in the 80 g per day sucrose corn syrup in initiating liver dysfunction widening group of health-related prob- beverage (40 g). This is less fructose than and possibly leading to nonalcoholic fatty lems. Several meta-analyses have shown a is contained in two 16-ounce sugar-sweet- liver disease and the metabolic syndrome, relationship between the consumption of ened soft drinks with 10% sugar. There which have become increasingly prevalent. sugar-sweetened soft drinks and obesity was also a high-glucose beverage (80 g Figure 1 relates the findings from this study (9–11). The relation of these beverages to per day), which is twice what was in the to those of other studies (13,16–18,22). The obesity can be attributed to the increased sucrose beverage, and an 80 g per day increasing intake of soft drinks (4,5) is caloric intake and to the fact that bever- fructose beverage, which is also twice the viewedasthedriverfortheincreaseinen- ages do not suppress the intake of other amount in the sucrose and low-fructose ergy and fructose, which may play a part in foods to an appropriate degree—thus beverages. With the hyperinsulinemic- the development of obesity and the meta- beverage calories serve as “add-on” calo- euglycemic clamp, the authors examined bolic consequences depicted here (22). The ries enhancing the risk of obesity (12) insulin sensitivity of the liver and the present in these beverages is viewed (Fig. 1). Meta-analyses have also sug- whole body. Compared with the high- as a positive feedback signal because of gested that the consumption of sugar- glucose beverage, the low-fructose bever- its ability to stimulate the central nervous sweetened beverages is related to the age impaired hepatic insulin sensitivity, system. risk of diabetes, the metabolic syndrome, but not whole-body insulin sensitivity, Two other meta-analyses of crystalline and cardiovascular disease (13). pointing again to the pathophysiological fructose added to the diet appeared to reach Several short-term clinical trials have effects that fructose can have on the liver. different conclusions. Livesey and Taylor provided insights into the metabolic con- In addition, they found that total and LDL (23) and Sievenpiper et al. (24) examined sequences of ingesting sugar-sweetened cholesterol were increased by fructose the effects of replacing in the care.diabetesjournals.org DIABETES CARE, VOLUME 36, JANUARY 2013 11 Commentary

DOI: 10.2337/dc12-1631 young adults. Int J Obes (Lond) 2007;31: © 2013 by the American Diabetes Association. 1688–1695 Readers may use this article as long as the work is 13. Malik VS, Popkin BM, Bray GA, Després properly cited, the use is educational and not for fi JP, Willett WC, Hu FB. Sugar-sweetened pro t, and the work is not altered. See http:// beverages and risk of metabolic syndrome creativecommons.org/licenses/by-nc-nd/3.0/ for details. and : a meta-analysis. Di- abetes Care 2010;33:2477–2483 14. Raben A, Vasilaras TH, Møller AC, Astrup Acknowledgments—No potential conflicts of AA. Sucrose compared with artificial interest relevant to this article were reported. sweeteners: different effects on ad libitum food intake and body weight after 10 wk cccccccccccccccccccccccc of supplementation in overweight sub- jects. Am J Clin Nutr 2002;76:721–729 References 15. Sørensen LB, Raben A, Stender S, Astrup 1. Havel PJ. Dietary fructose: implications A. Effect of sucrose on inflammatory for dysregulation of energy homeostasis markers in overweight humans. Am J and lipid/ metabolism. Nutr Clin Nutr 2005;82:421–427 Rev 2005;63:133–157 16. Stanhope KL, Schwarz JM, Keim NL, et al. 2. Vos MB, Kimmons JE, Gillespie C, Welsh Consuming fructose-sweetened, not glucose- J, Blanck HM. Dietary fructose consump- sweetened, beverages increases visceral ad- tion among US children and adults: the iposity and lipids and decreases insulin Third National Health and Nutrition Ex- sensitivity in overweight/obese humans. amination Survey. Medscape J Med 2008; J Clin Invest 2009;119:1322–1334 10:160 17. Stanhope KL, Bremer AA, Medici V, et al. Figure 1—Model showing some potential 3. Marriott BP, Olsho L, Hadden L, Connor Consumption of fructose and high fructose consequences of increasing fructose and energy P. Intake of added sugars and selected nu- corn syrup increase postprandial triglycer- intake from sugar or high-fructose corn syrup trients in the United States, National ides, LDL-cholesterol, and apolipoprotein-B in beverages. VAT, visceral adipose tissue. Health and Nutrition Examination Survey in young men and women. J Clin (NHANES) 2003-2006. Crit Rev Food Sci Endocrinol Metab 2011;96:E1596–E1605 Nutr 2010;50:228–258 18. Maersk M, Belza A, Stødkilde-Jørgensen diet with crystalline fructose. Both ex- 4. Duffey KJ, Popkin BM. Shifts in patterns H, et al. Sucrose-sweetened beverages in- and consumption of beverages between cluded high-fructose corn syrup and crease fat storage in the liver, muscle, and 1965 and 2002. Obesity (Silver Spring) thus the beverage form of fructose, which visceral fat depot: a 6-mo randomized in- seems to play the central role in the re- 2007;15:2739–2747 Food Consump- tervention study. Am J Clin Nutr 2012;95: 5. Putnam JJ, Allshouse JE. – sponse to the fructose in beverages. Crys- tion, Prices and Expenditure, 1970-1997 283 289 talline fructose added to the food supply . U.S. Department of Agriculture Food 19. Aeberli I, Hochuli M, Gerber PA, et al. represents only a few percent of the total Moderate amounts of fructose consump- “ ” and Rural Economics Division, Economic added sugars and behaves differently Research Service Bulletin, No. 965, p. 34 tion impair insulin sensitivity in healthy from the fructose that is in beverages. young men: a randomized controlled trial. 6. Bray GA, Nielsen SJ, Popkin BM. Con- – The largest amount of dietary fructose sumption of high-fructose corn syrup in Diabetes Care 2013;36:150 156 comes from the fructose in sucrose or beverages may play a role in the epidemic 20. Aeberli I, Gerber PA, Hochuli M, et al. high-fructose corn syrup, both of which of obesity. Am J Clin Nutr 2004;79:537– Low to moderate sugar-sweetened bev- are the major components of calorie- 543 erage consumption impairs glucose and lipid metabolism and promotes inflam- sweetened beverages but were excluded 7. Bray GA. How bad is fructose? Am J Clin – mation in healthy young men: a random- from these meta-analyses. Nutr 2007;86:895 896 8. Dekker MJ, Su Q, Baker C, Rutledge AC, ized controlled trial. Am J Clin Nutr 2011; One key question which Aeberli et al. – Adeli K. Fructose: a highly lipogenic 94:479 485 begin to address is whether the detrimen- nutrient implicated in insulin resistance, 21. Cohen JC, Schall R. Reassessing the effects tal effects of fructose are simply the result hepatic steatosis, and the metabolic syn- of simple carbohydrates on the serum tri- of a linear dose-response to our increasing drome. Am J Physiol Endocrinol Metab glyceride responses to fat meals. Am J Clin – dietary intake of fructose or whether there 2010;299:E685–E694 Nutr 1988;48:1031 1034 is a threshold below which fructose is 9. Vartanian LR, Schwartz MB, Brownell KD. 22. Welsh JA, Sharma A, Abramson JL, without harm. The current data suggest Effects of soft consumption on nu- Vaccarino V, Gillespie C, Vos MB. Caloric “ ” trition and health: a systematic review and sweetener consumption and dyslipidemia that it is a linear response, and the rea- – son we are now detecting the pathophys- meta-analysis. Am J Public Health 2007; among US adults. JAMA 2010;303:1490 – 1497 iological consequences of fructose is that 97:667 675 23. Livesey G, Taylor R. Fructose consump- its dietary load has continued to increase, 10. Olsen NJ, Heitmann BL. Intake of calori- cally sweetened beverages and obesity. tion and consequences for glycation, largely as a consequence of increased soft Obes Rev 2009;10:68–75 plasma triacylglycerol, and body weight: drink and fruit drink consumption. 11. Malik VS, Willett WC, Hu FB. Sugar- meta-analyses and meta-regression mod- sweetened beverages and BMI in children els of intervention studies. Am J Clin Nutr – GEORGE A. BRAY, MD and adolescents: reanalyses of a meta- 2008;88:1419 1437 analysis. Am J Clin Nutr 2009;89:438– 24. Sievenpiper JL, de Souza RJ, Mirrahimi A, – From the Pennington Biomedical Research Center, 439; author reply 439 440 et al. Effect of fructose on body weight in Louisiana State University, Baton Rouge, Louisiana. 12. Mourao DM, Bressan J, Campbell WW, controlled feeding trials: a systematic re- Corresponding author: George A. Bray, brayga@ Mattes RD. Effects of food form on appe- view and meta-analysis. Ann Intern Med pbrc.edu. tite and energy intake in and obese 2012;156:291–304

12 DIABETES CARE, VOLUME 36, JANUARY 2013 care.diabetesjournals.org