Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity1,2

Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity1,2

Commentary

Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity1,2

George A Bray, Samara Joy Nielsen, and Barry M Popkin

ABSTRACT Obesity is a major epidemic, but its causes are still unclear. In this article, we investigate the relation between the intake of high- fructose corn syrup (HFCS) and the development of obesity. We analyzed food consumption patterns by using US Department of Agriculture food consumption tables from 1967 to 2000. The con- sumption of HFCS increased � 1000% between 1970 and 1990, far exceeding the changes in intake of any other food or food group. HFCS now represents � 40% of caloric sweeteners added to foods and beverages and is the sole caloric sweetener in soft drinks in the United States. Our most conservative estimate of the consumption of HFCS indicates a daily average of 132 kcal for all Americans aged � 2 y, and the top 20% of consumers of caloric sweeteners ingest 316 kcal from HFCS/d. The increased use of HFCS in the United States mirrors the rapid increase in obesity. The digestion, absorption, and metabolism of fructose differ from those of glucose. Hepatic metabo- lism of fructose favors de novo lipogenesis. In addition, unlike glucose, fructose does not stimulate insulin secretion or enhance leptin produc- tion. Because insulin and leptin act as key afferent signals in the regu- lation of food intake and body weight, this suggests that dietary fructose may contribute to increased energy intake and weight gain. Further- more, calorically sweetened beverages may enhance caloric overcon- sumption. Thus, the increase in consumption of HFCS has a temporal relation to the epidemic of obesity, and the overconsumption of HFCS in calorically sweetened beverages may play a role in the epidemic of obesity. Am J Clin Nutr 2004;79:537–43.

KEY WORDS Epidemiology, food intake, obesity, artificial sweeteners, fructose

INTRODUCTION

As obesity has escalated to epidemic proportions around the world, many causes, including dietary components, have been suggested. Excessive caloric intake has been related to high-fat foods, increased portion sizes, and diets high both in simple sugars such as sucrose and in high-fructose corn syrup (HFCS) as a source of fructose (1–3). In this article, we discuss the evidence that a marked increase in the use of HFCS, and therefore in total fructose consumption, preceded the obesity epidemic and may be an important contributor to this epidemic in the United States.

To provide a common frame of reference for the terms used in this paper, the following definitions should be understood. Sugar is any free monosaccharide or disaccharide present in a food. Sugars includes at least one sugar; composite sugars refers to the aggregate of all forms of sugars in a food and is thus distinguish- able from specific types of sugar, such as fructose, glucose, or

sucrose. Added sugar is sugar added to a food and includes sweeteners such as sucrose, HFCS, honey, molasses, and other syrups. Naturally occurring sugar is sugar occurring in food and not added in processing, preparation, or at the table. Total sugars represents the total amount of sugars present in a food and in- cludes both naturally occurring and added sugars. Free fructose is fructose that exists in food as the monosaccharide. Fructose refers to both the free and bound forms of fructose (4).

Added sweeteners are important components of our diet, rep- resenting 318 kcal of dietary intake for the average American aged � 2 y, or 16% of all caloric intake as measured by a na- tionally representative survey in 1994–1996 (5). Sweet corn- based syrups were developed during the past 3 decades and now represent close to one-half of the caloric sweeteners consumed by Americans (6, 7). HFCS made by enzymatic isomerization of glucose to fructose was introduced as HFCS-42 (42% fructose) and HFCS-55 (55% fructose) in 1967 and 1977, respectively, and opened a new frontier for the sweetener and soft drink industries. Using a glucose isomerase, the starch in corn can be efficiently converted to glucose and then to various amounts of fructose. The hydrolysis of sucrose produces a 50:50 molar mixture of fructose and glucose. The development of these inexpensive, sweet corn- based syrups made it profitable to replace sucrose (sugar) and simple sugars with HFCS in our diet, and they now represent 40% of all added caloric sweeteners (8). Fructose is sweeter than sucrose. In comparative studies of sweetness, in which the sweet- ness of sucrose was set at 100, fructose had a sweetness of 173 and glucose had a sweetness of 74 (9). If the values noted above are applied, HFCS-42 would be 1.16 times as sweet as sucrose, and HFCS-55 would be 1.28 times as sweet as sucrose. This contrasts with the estimates reported by Hanover and White (10). In their study, the sweetness of sucrose was set at 100 as in reference 8. Fructose, however, had a sweetness of only 117, whereas a 50:50 mixture of fructose and sucrose had a sweetness of 128. It is difficult to see why fructose and sucrose combined would be sweeter than either one alone and as sweet as HFCS-55. On the basis of data in Agriculture Handbook no. 8 from the US Department of Agriculture (USDA) (11), a cola beverage in 1963

1 From the Pennington Biomedical Research Center, Louisiana State Uni- versity, Baton Rouge, LA (GAB), and the Department of Nutrition, Univer- sity of North Carolina, Chapel Hill (SJN and BMP).

2 Reprints not available. Address correspondence to GA Bray, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808. E-mail: brayga@pbrc.edu.

Received October 3, 2003. Accepted for publication December 15, 2003.

537Am J Clin Nutr 2004;79:537–43. Printed in USA. © 2004 American Society for Clinical Nutrition

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

had 39 kcal/100 g, whereas a cola beverage in 2003 had 41 kcal/ 100 g. Because the number of calories per 100 g has not changed substantially over the past 40 y, current beverages are probably sweeter, depending on the temperature at which they are served.

HFCS has become a favorite substitute for sucrose in carbon- ated beverages, baked goods, canned fruits, jams and jellies, and dairy products (10). The major user of HFCS in the world is the United States; however, HFCS is now manufactured and used in many countries throughout the world (7). In the United States, HFCS is the major source of caloric sweeteners in soft drinks and many other sweetened beverages and is also included in numer- ous other foods; therefore, HFCS constitutes a major source of dietary fructose. Few data are available on foods containing HFCS in countries other than the United States (7).

THE BIOLOGY

Absorption of fructose

The digestive and absorptive processes for glucose and fruc- tose are different. When disaccharides such as sucrose or maltose enter the intestine, they are cleaved by disaccharidases. A sodium- glucose cotransporter absorbs the glucose that is formed from cleav- age of sucrose. Fructose, in contrast, is absorbed further down in the duodenum and jejunum by a non-sodium-dependent process. After absorption, glucose and fructose enter the portal circulation and either are transported to the liver, where the fructose can be taken up and converted to glucose, or pass into the general circulation. The addition of small, catalytic amounts of fructose to orally ingested glucose increases hepatic glycogen synthesis in human subjects and reducesglycemic responses in subjectswith type2diabetesmellitus (12), which suggests the importance of fructose in modulating me- tabolism in the liver. However, when large amounts of fructose are ingested, they provide a relatively unregulated source of carbon precursors for hepatic lipogenesis.

Fructose and insulin release

Along with 2 peptides, glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 released from the gas- trointestinal tract, circulating glucose increases insulin release from the pancreas (13, 14). Fructose does not stimulate insulin secretion in vitro, probably because the � cells of the pancreas lack the fructose transporter Glut-5 (15, 16). Thus, when fructose is given in vivo as part of a mixed meal, the increase in glucose and insulin is much smaller than when a similar amount of glu- cose is given. However, fructose produces a much larger increase in lactate and a small (1.7%) increase in diet-induced thermo- genesis (17), which again suggests that glucose and fructose have different metabolic effects.

Insulin and leptin

Insulin release can modulate food intake by at least 2 mecha- nisms. First, Schwartz et al (18) have argued that insulin con- centrations in the central nervous system have a direct inhibitory effect on food intake. In addition, insulin may modify food intake by its effect on leptin secretion, which is mainly regulated by insulin-induced changes in glucose metabolism in fat cells (19, 20). Insulin increases leptin release (21) with a time delay of several hours. Thus, a low insulin concentration after ingestion of fructose would be associated with lower average leptin concen- trations than would be seen after ingestion of glucose. Because

leptin inhibits food intake, the lower leptin concentrations induced by fructose would tend to enhance food intake. This is most dramat- ically illustrated inhumanswho lack leptin (22,23).Persons lacking leptin (homozygotes) are massively obese (22), and heterozygotes with low but detectable serum leptin concentrations have increased adiposity (23), which indicates that low leptin concentrations are associated with increased hunger and gains in body fat. Adminis- tration of leptin to persons who lack it produces a dramatic decrease in food intake, as expected. Leptin also increases energy expendi- ture, and during reduced calorie intake, leptin attenuates the de- creases in thyroid hormones and 24-h energy expenditure (24). To the extent that fructose increases in the diet, one might expect less insulin secretion and thus less leptin release and a reduction in the inhibitory effect of leptin on food intake, ie, an increase in food intake. This was found in the preliminary studies reported by Teff et al (25). Consumption of high-fructose meals reduced 24-h plasma insulin and leptin concentrations and increased postprandial fasting triacylglycerol concentrations in women but did not suppress circu- lating ghrelin concentrations.

Fructose and metabolism

The metabolism of fructose differs from that of glucose in several other ways as well (3). Glucose enters cells by a transport mechanism (Glut-4) that is insulin dependent in most tissues. Insulin activates the insulin receptor, which in turn increases the densityofglucose transporterson thecell surfaceand thus facilitates the entry of glucose. Once inside the cell, glucose is phosphorylated by glucokinase to become glucose-6-phosphate, from which the intracellular metabolism of glucose begins. Intracellular enzymes can tightly control conversion of glucose-6-phosphate to the glyc- erol backbone of triacylglycerols through modulation by phospho- fructokinase. In contrast with glucose, fructose enters cells via a Glut-5 transporter that does not depend on insulin. This transporter is absent from pancreatic � cells and the brain, which indicates limited entry of fructose into these tissues. Glucose provides “sati- ety” signals to the brain that fructose cannot provide because it is not transported into the brain. Once inside the cell, fructose is phosphor- ylated to formfructose-1-phosphate (26). In thisconfiguration, fruc- tose is readily cleaved by aldolase to form trioses that are the back- bone for phospholipid and triacyglycerol synthesis. Fructose also provides carbon atoms for synthesis of long-chain fatty acids, al- though in humans, the quantity of these carbon atoms is small. Thus, fructose facilitates the biochemical formation of triacylglycerols more efficiently than does glucose (3). For example, when a diet containing 17% fructose was provided to healthy men and women, the men, but not the women, showed a highly significant increase of 32% in plasma triacylglycerol concentrations (27).

Overconsumption of sweetened beverages

One model for producing obesity in rodents is to provide sweet- ened (sucrose, maltose, etc) beverages for them to drink (28). In this setting, the desire for the calorically sweetened solution reduces the intake of solid food, but not by enough to prevent a positive caloric balance and the slow development of obesity. Adding the same amount of sucrose or maltose as of a solid in the diet does not produce the same response. Thus, in experimental animals, sweet- ened beverages appear to enhance caloric consumption.

Fructose and soft drinks

A similar argument about the role of overconsumption of calorically sweetened beverages may apply to humans (29–32).

538 BRAY ET AL

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

Mattes (29) reported that when humans ingest energy-containing beverages, energy compensation is less precise than when solid foods are ingested. In another study in humans, DiMeglio and Mattes (30) found that when 15 healthy men and women were given a carbohydrate load of 1880 kJ/d (450 kcal/d) as a calori- cally sweetened soda for 4 wk, they gained significantly more weight than when the same carbohydrate load was given in a solid form as jelly beans. Additional support for our hypothesis that calorically sweetened beverages may contribute to the epidemic of obesity comes from a longitudinal study in adolescents. Lud- wig et al (31) showed that in adolescents participating in the Planet Health project, the quantity of sugar-sweetened beverages ingested predicted initial body mass index (BMI; in kg/m2) and gain in BMI during the follow-up period. Raben et al (32) de- signed a randomized, double-blind study to compare the effect of calorically sweetened beverages with that of diet drinks on weight gain in moderately overweight men and women. This European study found that drinking calorically sweetened bev- erages resulted in greater weight gain over the 10-wk study than did drinking diet drinks. Compared with the subjects who con- sumed diet drinks, those who consumed calorically sweetened

beverages did not compensate for this consumption by reducing the intake of other beverages and foods and thus gained weight. The beverages in this study were sweetened with sucrose, whereas in the United States almost all calorically sweetened beverages are sweetened with HFCS. Thus, we need a second randomized controlled study that compares sucrose- and HFCS- sweetened beverages. This could establish whether the form of the caloric sweetener played a role in the weight gain observed in the study by Raben et al (32).

The results of the studies by Raben et al (32) and Ludwig et al (31) suggest that the rapid increase in the intake of calorically sweetened soft drinks could be a contributing factor to the epi- demic of weight gain. Between 1970, when HFCS was intro- duced into the marketplace, and 2000, the per capita consumption of HFCS in the United States increased from 0.292 kg · person�1

· y�1 (0.6 lb · person�1 · y�1) to 33.4 kg · person�1 · y�1 (73.5 lb · person�1 · y�1), an increase of � 100-fold (8) (Table 1). The total consumption of fructose increased nearly 30%. The con- sumption of free fructose showed a greater increase, which re- flected the increasing use of HFCS (Figure 1). During the same interval, the consumption of sucrose decreased nearly 50%, and

TABLE 1 Availability of high-fructose corn syrup (HFCS) in the US caloric sweetener supply1

Year HFCS Total caloric sweeteners HFCS as percentage of total

caloric sweeteners Percentage of HFCS

from HFCS-42 Percentage of HFCS from

HFCS-55

g � person�1 � d�1 g � person�1 � d�1 % % % 1966 0.0 165.9 0.0 — — 1970 0.8 175.1 0.4 100.0 0.0 1975 7.1 168.8 4.2 100.0 0.0 1980 27.3 176.0 15.5 71.2 28.8 1985 64.7 184.4 35.1 34.3 65.7 1990 71.0 195.7 36.3 41.0 59.0 1995 82.3 211.7 38.9 39.9 60.1 2000 91.6 218.0 42.0 38.8 61.2

1 Data from reference 8. HFCS-42 and HFCS-55, HFCS containing 42% and 55% fructose, respectively.

FIGURE 1. Estimated intakes of total fructose (F), free fructose (Œ), and high-fructose corn syrup (HFCS, �) in relation to trends in the prevalence of overweight (■) and obesity (x) in the United States. Data from references 7 and 35.

HIGH-FRUCTOSE CORN SYRUP AND OBESITY 539

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

the intakes of sucrose and HFCS are now nearly identical. Al- though this shift has clearly led to a major increase in free- fructose consumption, it is unclear how much of the increase in consumption of calorically sweetened soft drinks is a result of the shift to beverages in which one-half of the fructose is free rather than bound with glucose as in sucrose. A recent review described many facets of this issue (3).

HFCS USE AND INTAKE

Availability of HFCS in the food supply

In 1970 HFCS represented � 1% of all caloric sweeteners available for consumption in the United States, but the HFCS portion of the caloric sweetener market jumped rapidly in the 1980s and by 2000 represented 42.0% of all caloric sweeteners (Table 1) (8). HFCS-42 was initially the only HFCS component, but by the early 1980s, HFCS-55 had become the major source and constituted 61.2% of all HFCS in 2000. These data are based on per capita food disappearance data. In the absence of direct measures of HFCS intake, these data provide the best indirect measure of the HFCS available for consumption in the United States. The data are useful for studying trends but probably over- estimate intake patterns. Although it is useful to understand that HFCS intake represents more than two-fifths of the total intake of caloric sweeteners in the United States, it is also important to recognize that the proportion of HFCS in some foods is much higher than that in other foods.

Foods containing HFCS

In the United States, HFCS is found in almost all foods con- taining caloric sweeteners. These include most soft drinks and fruit drinks, candied fruits and canned fruits, dairy desserts and flavored yogurts, most baked goods, many cereals, and jellies. Over 60% of the calories in apple juice, which is used as the base for many of the fruit drinks, come from fructose, and thus apple juice is another source of fructose in the diet. Lists of HFCS- containing foods can be obtained from organizations concerned with HFCS-related allergies (33). It is clear that almost all caloric sweeteners used by manufacturers of soft drinks and fruit drinks are HFCS (4, 34). In fact, about two-thirds of all HFCS consumed in the United States are in beverages. Aside from beverages, there is no definitive literature on the proportion of caloric sweeteners that is HFCS in other processed foods. HFCS is found in most processed foods; however, the exact compositions are not avail- able from either the manufacturer or any publicly available food- composition table.

Trends in obesity and HFCS availability

There are important similarities between the trend in HFCS availability and the trends in the prevalence of obesity in the United States (Figure 1). Using age-standardized, nationally rep- resentative measures of obesity at 5 time points from 1960 to 1999 (35) and data on the availability of HFCS collected annually over this same period, we graphed both patterns. The data on obesity are from the National Center for Health Statistics for the following periods: 1960–1962 (National Health Examination Survey I), 1971–1975 [National Health and Nutrition Examination Survey (NHANES)], 1976–1980 (NHANES II), 1988–1994 (NHANES III), and 1999 (NHANES 1999–2000) (35). The HFCS data are those fromTable1.Theprevalenceofoverweight (BMIof25–29.9)

and the prevalence of obesity (BMI � 30) were fit with fourth-order polynomial curves so that the limited number of data points could be fitted into a curve to capture the US trends. We also included esti- matesof free-fructose intakeand total fructose intake.Total fructose is thesumof free fructoseandfructose that ispartof thedisaccharide sucrose. Free fructose is the monosaccharide in HFCS and is also obtained in small amounts from other sources. Free-fructose intake closely follows the intake of HFCS. Total fructose intake increased nearly 30% between 1970 and 2000.

Estimated HFCS consumption

The intake of caloric sweeteners in the United States has in- creased rapidly, and nationally representative data from 1994 to 1998 from the USDA allow us to estimate an intake of 318 kcal/d for the average US resident aged � 2 y. This value is one-sixth of the intake of all calories and close to one-third of the intake of all carbohydrates and represents a significant increase over the past 2 decades (Table 2). As the intake of caloric sweeteners increased, so did the fructose load, which increased from 158.5 to 228 kcal · person�1 · d�1 between 1977–1978 (36) and 1994–1998 (38, 39).

Furthermore, as shown in Table 2, the major increase in the intake of caloric sweeteners from 1977 to 1998 came from the intake of soft drinks and fruit drinks, and these intake values were 105 and 31 kcal · person�1 · d�1, respectively, of the total added sugar intake of 318 kcal · person�1 · d�1 in 1994–1998. More- over, more than one-half of the increased caloric sweetener in- take during this time period came from the intake of these bev- erages. The intakes of these 2 types of beverages and of desserts total about two-thirds of all caloric sweetener intake in the United States today (Table 2).

We have no way to directly measure total HFCS use. However, one Food and Drug Administration study used very conservative methods to estimate HFCS use for the nationally representative dietary intake sample from 1977 to 1978 from the USDA (35). Using measures of the proportion of HFCS in each food, Glins- mann et al (40) created food category–wide estimates of the proportion of caloric sweeteners that is HFCS. We applied those same proportions to a set of food groups to estimate the use of caloric sweeteners not only during 1977–1978 but also during later periods. Using our conversion technique applied to the initial 1977–1978 Nationwide Food Consumption Survey data, we obtained results that were only 4 kcal higher than the esti- mates of Glinsmann et al (39). This approach is based on HFCS composition in the early 1980s. With the use of the USDA value of 4.2 g HFCS/tsp (0.84 g/mL), the availability of HFCS in the food supply in the early 1980s was only one-half of the current availability on a gram per capita basis (as shown in Table 1). Thus, we feel confident that we can use this approach to provide a conservative lower limit of HFCS intake.

In Table 2, the 2 columns at the right contain our estimates of HFCS intake and total fructose intake. On the basis of the trend in intakes, our estimate of HFCS intake for the most recent period of measurement from 1994–1998 is 132 kcal · person�1 · d�1

(37). This represents a shift between 1977–1978 and 1994–1998 from 4.5% of total calories to 6.7% of total calories, or from 10.1% of carbohydrates to 13.1% of carbohydrates (41). The estimate of total fructose intake, which was obtained from the intakes of sucrose and HFCS, would be somewhat higher if we knew the fructose content of the fruit drinks.

540 BRAY ET AL

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

Distribution of HFCS intake matters!

We also explored the distribution of HFCS consumption by examining quintiles of caloric sweetener intake among Ameri- cans aged � 2 y (Table 3). Most of the increase in caloric sweetener intake from the middle quintile to the upper quintile came from increases in the intake of calorically sweetened bev- erages, particularly soft drinks. Consumers in the top quintile, which represents 20% of all Americans, consume � 11% of their calories from HFCS. Again, remember that this is a very conser-

vative estimate. This same group obtains almost one-half of its carbohydrates from caloric sweeteners and one-fifth of its car- bohydrates from HFCS.

DISCUSSION

Genetic factors play an important role in the development of obesity (42). However, the rapidity with which the current epi- demic of obesity has descended on the United States (35) and

TABLE 2 Trends in intakes of added sugar (sucrose) and high-fructose corn syrup (HFCS) by Americans aged �2 y1

Soft drinks Fruit drinks Desserts Total added

sugar Estimated HFCS

intake2 Estimated total fructose

intake3

Intake (kcal � person�1 � d�1) 1977–1978 52 18 54 235 80 158.5 1989–1991 744 19 474 242 95 170 1994–1998 1054,5 314,5 604,5 3184,5 132 228

Percentage of total energy (%) 1977–1978 2.9 1.0 3.0 13.1 4.5 8.8 1989–1991 4.14 1.1 2.64 13.5 5.3 9.4 1994–1998 5.34,5 1.64,5 3.04,5 16.04,5 6.7 11.5

Percentage of total carbohydrates (%) 1977–1978 6.5 2.3 6.8 29.5 10.1 19.8 1989–1991 8.54 2.2 5.44 27.8 10.9 19.5 1994–1998 10.44,5 3.14,5 5.94,5 31.54,5 13.1 22.8

1 Data from references 36–39. 2 Coefficients derived from Glinsmann et al (40) were used to estimate HFCS intake. 3 Estimated as 50% of sucrose � fructose in HFCS. 4 Significantly different from 1977–1978, P � 0.01 (t test). 5 Significantly different from 1989–1991, P � 0.01 (t test).

TABLE 3 Distribution of consumption of added caloric sweeteners and high-fructose corn syrup (HFCS) by quintile (Q) of added sugar intake in persons aged �2 y (1994–1996, 1998)1

Q1 (n � 3907) Q2 (n � 4259) Q3 (n � 4228) Q4 (n � 3645) Q5 (n � 2988) Total (n � 19 027) P for trend

Added caloric sweeteners Intake (kcal � person�1 � d�1)

Soft drinks 6 31 69 129 290 105 0.022 Fruit drinks 4 14 26 43 69 31 0.003 Total 69 169 263 388 703 318 0.001

Percentage of total energy in kcal (%) Soft drinks 0.4 1.9 3.7 6.0 10.2 5.3 0.004 Fruit drinks 0.3 0.8 1.4 2.0 2.4 1.6 0.0001 Total 4.9 10.2 14.0 18.1 24.7 16.0 0.0004

Percentage of carbohydrates (%) Soft drinks 0.9 3.8 7.3 11.6 18.8 10.4 0.003 Fruit drinks 0.6 1.7 2.8 3.9 4.5 3.1 0.0004 Total 10.5 20.8 27.9 35.0 45.6 31.4 0.0002

HFCS Intake (kcal � person�1 � d�1)

Soft drinks 4 22 49 91 205 74 0.022 Fruit drinks 3 10 18 30 49 22 0.004 Total 21 58 103 166 316 132 0.012

Percentage of total energy in kcal (%) Soft drinks 0.3 1.3 2.6 4.2 7.2 3.7 0.005 Fruit drinks 0.2 0.6 1.0 1.4 1.7 1.1 �0.0001 Total 1.5 3.5 5.5 7.7 11.1 6.7 0.0007

Percentage of carbohydrates (%) Soft drinks 0.6 2.7 5.2 8.2 13.3 7.3 0.003 Fruit drinks 0.4 1.2 1.9 2.7 3.2 2.2 0.0002 Total 3.2 7.2 10.9 14.9 20.5 13.1 0.0002

1 Data from references 38 and 39 (weighted to be nationally representative).

HIGH-FRUCTOSE CORN SYRUP AND OBESITY 541

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

many other countries (43) makes environmental factors the more likely explanation. From a public health perspective, the key question is whether there are modifiable environmental agents that could have triggered this epidemic and that might be altered. Several environmental agents, including reduced levels of phys- ical activity (44), a decrease in smoking, increased portion size (2), eating outside the home and at fast-food restaurants, and changes in the types of food that are ingested (45), have been suggested. In this article, we propose that the introduction of HFCS and the increased intakes of soft drinks and other sweet- ened beverages have led to increases in total caloric and fructose consumption that are important contributors to the current epi- demic of obesity.

In this article, we address an important potential hypothesis by which HFCS may have an environmental link with the epidemic of obesity. When total calorie intake is fixed, ie, if a person eats the same amount of fructose, glucose, or sucrose in a metaboli- cally controlled setting, the response should be the same, and this was shown by McDevitt et al (46). This situation is not one in which differences in taste and portion size are allowed to operate. However, many biological factors that we noted in this article suggest that calorically sweetened beverages are associated with overconsumption when the sweetener is in a liquid form (29–32). The collective data suggest that overconsumption of beverages sweetened with HFCS and containing � 50% free fructose and the increased intake of total fructose may play a role in the epidemic of obesity. Whether HFCS used in solid food produces the same overconsumption as it does in beverages is unknown, but we suspect that if the HFCS was entirely in the solid form, it would not pose the same problem (30). Total fructose, both free fructose and fructose combined in sucrose, in both beverages and solid food may be viewed as a precursor to fat because of the ease with which the carbon skeleton of fructose can form the back- bone for triacylglycerols and be used for the synthesis of long- chain fatty acids (25). Additional clinical trials are clearly needed to buttress the conclusions of Raben et al (32) that beverages containing sugar caused more weight gain over 10 wk than did diet beverages.

There is a distinct likelihood that the increased consumption of HFCS in beverages may be linked to the increase in obesity. One US study showed that beverages sweetened with HFCS are linked with increased energy intake and weight gain (31). Fur- thermore, we showed that the increase in the use of HFCS was concurrent with the increase in obesity rates in the United States. HFCS was introduced into the food supply just before 1970 and increased rapidly to constitute � 40% of the sweeteners used by 2000 (8). The increase in HFCS consumption just preceded the rapid increase in the prevalence of obesity that occurred between the National Center for Health Statistics survey in 1976–1980 and the next survey in 1988–1994 (35) (Figure 1).

HFCS is used in soft drinks, and HFCS and apple juice, which has 65% fructose, are used as the principal sweeteners for fruit drinks. The increasing consumption of calorically sweetened soft drinks has been associated with a decrease in the intake of milk (5, 8, 47, 48). This relation adds another mechanism by which HFCS consumption in beverages may be related to the epidemic of obesity. Per capita calcium intake decreased from 890 to 860 mg/d between 1970 and 1981 but has increased slowly since then (8). This was a mean 3% decrease that probably reflected a much larger decrease in calcium intake at the upper ends of the skewed distribution for intake. In US teenagers, total calcium

intake decreased by � 50 mg/d (� 10%) between 1977 and 1996 (49). A convincing set of epidemiologic and clinical studies suggests that dairy products may have a favorable effect on body weight and insulin resistance in both children and adults (50–52). During the interval from 1970 to 1990, the intake of whole milk decreased 58%. However, the intake of cheese, which is high in fat, increased and partially offset the decrease in calcium intake from milk (8). Because milk is a major source of dietary calcium for most humans, this decrease in milk intake may play an im- portant role in the decrease in calcium in the diet.

One possible public health option is to address the sweet taste preference of humans by reducing HFCS and, if sweetness is needed, relying on artificial sweeteners to make up any differ- ence in sweetness. It is becoming increasingly clear that soft drink consumption may be an important contributor to the epi- demic of obesity, in part through the larger portion sizes of these beverages and through the increased intake of fructose from HFCS and sucrose (53). If HFCS acts as an agent in the disease, then reducing exposure to this agent may help to reduce the epidemic (54).

Some will question our measures of HFCS intake and avail- ability. Our very conservative estimate of HFCS use and the Food and Drug Administration data showed high HFCS intakes for a large segment of the US population (39).

In conclusion, we believe that an argument can now be made that the use of HFCS in beverages should be reduced and that HFCS should be replaced with alternative noncaloric sweeten- ers. Sweetness is a preferred taste as well as an acquired one that may be enhanced by exposure to sweet foods. The hypothesis that providing sodas and juice drinks in which caloric sweeteners are partially or completely replaced with noncaloric sweeteners will help reduce the prevalence of obesity is worth testing. If the intake of calorically sweetened beverages is contributing to the current epidemic, then reducing the availability of these bever- ages by removing soda machines from schools would be a strat- egy worth considering, as would reducing the portion sizes of sodas that are commercially available (55).

We thank Dan Blanchette for programming assistance, Tom Swasey for graphics support, and Frances Dancy for support in administrative matters. We also thank Linda Adair for thoughtful comments on an early version of this article.

REFERENCES 1. Bray GA, Popkin BM. Dietary fat intake does affect obesity! Am J Clin

Nutr 1998;68:1157–73. 2. Young LR, Nestle M. The contribution of expanding portion sizes to the

US obesity epidemic. Am J Public Health 2002;92:246–9. 3. Elliott SS, Keim NL, Stern JS, Teff K, Havel PJ. Fructose, weight gain,

and the insulin resistance syndrome. Am J Clin Nutr 2002;76:911–22. 4. Smith SM. High fructose corn syrup replaces sugar in processed food.

Environ Nutr 1998;11:7–8. 5. Nielsen SJ, Siega-Riz AM, Popkin BM. Trends in energy intake in U.S.

between 1977 and 1996: similar shifts seen across age groups. Obes Res 2002;10:370–8.

6. Higley NA, White JS. Trends in fructose availability and consumption in the United States. Food Technol 1991;45:118–22.

7. Vuilleumier S. Worldwide production of high-fructose syrup and crys- talline fructose. Am J Clin Nutr 1993;58(suppl):733S–6S.

8. Putnam JJ, Allshouse JE. Food consumption, prices and expenditures, 1970–97. US Department of Agriculture Economic Research Service statistical bulletin no. 965, April 1999. Washington, DC: US Govern- ment Printing Office, 1999.

9. Krause MV, Mahan LK. Food, nutrition and diet therapy. 7th ed. Phil- adelphia: WB Saunders Company, 1984.

542 BRAY ET AL

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from

10. Hanover LM, White JS. Manufacturing, composition, and applications of fructose. Am J Clin Nutr 1993;58(suppl):724S–32S.

11. Watt BK, Merrill AL. Composition of foods: raw, processed, prepared. Agriculture handbook no. 8. Washington, DC: US Government Printing Office, 1963.

12. Petersen KF, Laurent D, Yu C, Cline GW, Shulman GI. Stimulating effects of low-dose fructose on insulin-stimulated hepatic glycogen syn- thesis in humans. Diabetes 2001;50:1263–8.

13. Edwards DM, Todd JF, Mahmoudi M, et al. Glucagon-like peptide 1 has a physiological role in the control of postprandial glucose in humans: studies with the antagonist exendin 9–39. Diabetes 1999;48:86–93.

14. Vilsboll T, Krarup T, Madsbad S, Holst JJ. Both GLP-1 and GIP are insulinotropic at basal and postprandial glucose levels and contribute nearly equally to the incretin effect of a meal in healthy subjects. Regul Pept 2003;114:115–21.

15. Curry DL. Effects of mannose and fructose on the synthesis and secretion of insulin. Pancreas 1989;4:2–9.

16. Sato Y, Ito T, Udaka U, et al. Immunohistochemical localization of facilitated-diffusion glucose transporters in rat pancreatic islets. Tissue Cell 1996;28:637–43.

17. Schwarz J-M, Schutz Y, Froidevaux F, et al. Thermogenesis in men and women induced by fructose vs glucose added to a meal. Am J Clin Nutr 1989;49:667–74.

18. Schwartz MW, Woods SC, Porte D Jr, Seeley RJ, Baskin DG. Central nervous system control of food intake. Nature 2000;404:661–71.

19. Muller WM, Gregoire FM, Stanhope KL, et al. Evidence that glucose metabolism regulated leptin secretion from cultured rat adipocytes. En- docrinology 1998;39:551–8.

20. Havel PJ. Control of energy homeostasis and insulin action by adipocyte hormones: leptin, acylation stimulating protein, and adiponectin. Curr Opin Lipidol 2002;13:51–9.

21. Saad MF, Khan A, Sharma A, et al. Physiological insulinemia acutely modulated plasma leptin. Diabetes 1998;47:544–9.

22. Farooqi IS, Matarese G, Lord GM, et al. Beneficial effects of leptin on obesity, T cell hyporesponsiveness, and neuroendocrine/metabolic dys- function of human congenital leptin deficiency. J Clin Invest 2002;110: 1093–1103.

23. Farooqi IS, Keogh JM, Kamath S, et al. Partial leptin deficiency and human adiposity. Nature 2001;414:34–5.

24. Rosenbaum M, Murphy EM, Heymsfield SB, Matthews DE, Leibel RL. Low dose leptin administration reverses effects of sustained weight- reduction on energy expenditure and circulating concentrations of thy- roid hormones. J Clin Endocrinol Metab 2002;87:2391–4.

25. Teff K, Elliot S, Tschoep MR, et al. Consuming high fructose meals reduces 24 hour plasma insulin and leptin concentrations, does not sup- press circulating ghrelin, and increases postprandial and fasting triglyc- erides in women. Diabetes 2002;52(suppl):A408(abstr).

26. Mayes PA. Intermediary metabolism of fructose. Am J Clin Nutr 1993; 58(suppl):754S–65S.

27. Bantle JP, Raatz SK, Thomas W, Georgopoulos A. Effects of dietary fructose on plasma lipids in healthy subjects. Am J Clin Nutr 2000;72: 1128–34.

28. Sclafani A. Starch and sugar tastes in rodents: an update. Brain Res Bull 1991;27:383–6.

29. Mattes RD. Dietary compensation by humans for supplemental energy provided as ethanol or carbohydrate in fluids. Physiol Behav 1996;59: 179–87.

30. DiMeglio DP, Mattes RD. Liquid versus solid carbohydrate: effects on food intake and body weight. Int J Obes Relat Metab Disord 2000;24: 794–800.

31. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consump- tion of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 2001;357:505–8.

32. Raben A, Vasilaras TH, Moller AC, Astrup A. Sucrose compared with artifical sweeteners: different effects on ad libitum food intake and body weight after 10 wk of supplementation in overweight subjects. Am J Clin Nutr 2002;76:721–9.

33. Hurt-Jones M. The allergy self-help cookbook: over 350 natural food

recipes, free of all common food allergens. Emmaus, PA: Rodale Press, 2002.

34. Park YK, Yetley E. Intakes and food sources of fructose in the United States. Am J Clin Nutr 1993;58(suppl):737S–47S.

35. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999–2000. JAMA 2002;288:1723–7.

36. Rizek RL. The 1977–78 Nationwide Food Consumption Survey. Fam Econ Rev 1978;2001;4:3–7.

37. Tippett KS, Mickle SJ, Goldman JD, Sykes KE, Cook DA, Sebastian RS. Food and nutrient intakes by individuals in the United States, 1 day, 1989–91. Continuing Survey of Food Intakes by Individuals 1989–91, Nationwide Food Surveys report no. 91-2. Beltsville, MD: US Depart- ment of Agriculture, Agriculture Research Service, 1995.

38. Tippett KS, Cypel YS. Design and operation: the Continuing Survey of Food Intakes by Individuals and the Diet and Health Knowledge Survey 1994–96. Continuing Survey of Food Intakes by Individuals 1994–96, Nationwide Food Surveys report no. 96-1. Beltsville, MD: US Depart- ment of Agriculture, Agriculture Research Service, 1998.

39. United States Department of Agriculture, Agricultural Research Ser- vice. Design and operation: the Continuing Survey of Food Intakes by Individuals and the Diet and Health Knowledge Survey, 1994–96 and 1998. Beltsville, MD: US Department of Agriculture, Agricultural Re- search Service, 2000.

40. Glinsmann WH, Irausquin H, Park YK. Evaluation of health aspects of sugars contained in carbohydrate sweeteners. Report of Sugars Task Force, 1986. J Nutr 1986;116:S1–216.

41. Popkin BM, Nielsen SJ. The sweetening of the world’s diet. Obes Res 2003;11:1325–32.

42. Rankinen T, Perusse L, Weisnagel SJ, Snyder EE, Chagnon YC, Bou- chard C. The human obesity gene map: the 2001 update. Obes Res 2002;10:196–243.

43. World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000;894:i–xii, 1–253.

44. Prentice AM, Jebb SA. Obesity in Britain: gluttony or sloth? Br Med J 1995;311:437–9.

45. Troiano RP, Flegal KM. Overweight children and adolescents: descrip- tion, epidemiology, and demographics. Pediatrics 1998;101:497–504.

46. McDevitt RM, Poppitt SD, Murgatroyd PR, Prentice AM. Macronutrient disposal during controlled overfeeding with glucose, fructose, sucrose, or fat in lean and obese women. Am J Clin Nutr 2000;72:369–77.

47. Cavadini C, Siega-Riz AM, Popkin BM. US adolescent food intake trends from 1965 to 1996. Arch Dis Child 2000;83:18–24.

48. Harnack L, Stang J, Story M. Soft drink consumption among US children and adolescents: nutritional consequences. J Am Diet Assoc 1999;99: 436–41.

49. Carruth BR, Skinner JD. The role of dietary calcium and other nutrients in moderating body fat in preschool children. Int J Obes Relat Metab Disord 2001;25:559–66.

50. Davies KM, Heaney RP, Recker RR, et al. Calcium intake and body weight. J Clin Endocrinol Metab 2000;85:4635–8.

51. Pereira MA, Jacobs DR Jr, Van Horn L, et al. Dairy consumption, obesity, and the insulin resistance syndrome in young adults: the CAR- DIA study. JAMA 2002;287:2081–9.

52. Zemel MB, Shi H, Greer B, Dirienzo D, Zemel PC. Regulation of adi- posity by dietary calcium. FASEB J 2000;14:1132–8.

53. Barr SI, McCarron DA, Heaney RP, et al. Effects of increased consump- tion of fluid milk on energy and nutrient intake, body weight, and car- diovascular risk factors in healthy older adults. J Am Diet Assoc 2000; 100:810–7.

54. Nielsen SJ, Popkin BM. Patterns and trends in portion sizes, 1977–1998. JAMA 2003;289:450–3.

55. Bray GA. The fluoride hypothesis and diobesity. How to prevent dia- betes by preventing obesity. In: Mederios-Neto G, Halpern A, Bouchard C, eds. Progress in obesity research: 9. Proceedings of the 9th Interna- tional Congress on Obesity. Surrey, United Kingdom: John Libbey Eu- rotext, 2003:26–8.

HIGH-FRUCTOSE CORN SYRUP AND OBESITY 543

at O H

IO S

T A

T E

U N

IV John A

P rior H

ealth S ciences Library (H

E A

) on June 30, 2011 w

w w

.ajcn.org D

ow nloaded from


Comments are closed.