Fructose Intake at Current Levels in the United States May Cause Gastrointestinal Distress in Normal Adults

RESEARCH Current Research

Fructose Intake at Current Levels in the United States May Cause Gastrointestinal Distress in Normal Adults PETER L. BEYER, MS, RD; ELENA M. CAVIAR, MS, RD; RICHARD W. MCCALLUM, MD

ABSTRACT Objective Fructose intake has increased considerably in the United States, primarily as a result of increased consumption of high-fructose corn syrup, fruits and juices, and crystalline fructose. The purpose was to determine how often fructose, in amounts commonly consumed, would result in malabsorption and/or symptoms in healthy persons. Design Fructose absorption was measured using 3-hour breath hydrogen tests and symptom scores were used to rate subjective responses for gas, borborygmus, abdominal pain, and loose stools. Subjects/setting The study included 15 normal, free-living volunteers from a medical center community and was performed in a gastrointestinal specialty clinic. Intervention Subjects consumed 25- and 50-g doses of crystalline fructose with water after an overnight fast on separate test days. Main outcome measures Mean peak breath hydrogen, time of peak, area under the curve (AUC) for breath hydrogen and gastrointestinal symptoms were measured during a 3-hour period after subjects consumed both 25- and 50-g doses of fructose. Statistical analyses Differences in mean breath hydrogen, AUC, and symptom scores between doses were analyzed using paired t tests. Correlations among peak breath hydrogen, AUC, and symptoms were also evaluated. Results More than half of the 15 adults tested showed evidence of fructose malabsorption after 25 g fructose and greater than two thirds showed malabsorption after 50 g P. L. Beyer is an associate professor, Department of Dietetics and Nutrition, and R. W. McCallum is director, Center for Gastrointestinal Nerve and Muscle Function and the Division of GI Motility, University of Kansas Medical Center, Kansas City. E. M. Caviar is an infant nutrition representative with Nestle Nutritional Products, Overland Park, KS; at the time of the study, she was with the Department of Dietetics and Nutrition, University of Kansas Medical Center, Kansas City. Address correspondence to: Peter L. Beyer, MS, RD, Associate Professor, Dietetics and Nutrition, Mailstop 4013, University of Kansas Medical Center, 3901 Rainbow, Kansas City, KS 66160. E-mail: [email protected] Copyright © 2005 by the American Dietetic Association. 0002-8223/05/10510-0005$30.00/0 doi: 10.1016/j.jada.2005.07.002

© 2005 by the American Dietetic Association

fructose. AUC, representing overall breath hydrogen response, was significantly greater after the 50-g dose. Overall symptom scores were significantly greater than baseline after each dose, but scores were only marginally greater after 50 g than 25 g. Peak hydrogen levels and AUC were highly correlated, but neither was significantly related to symptoms. Conclusions Fructose, in amounts commonly consumed, may result in mild gastrointestinal distress in normal people. Additional study is warranted to evaluate the response to fructose-glucose mixtures (as in high-fructose corn syrup) and fructose taken with food in both normal people and those with gastrointestinal dysfunction. Because breath hydrogen peaks occurred at 90 to 114 minutes and were highly correlated with 180-minute breath hydrogen AUC, the use of peak hydrogen measures may be considered to shorten the duration of the exam. J Am Diet Assoc. 2005;105:1559-1566.

F

ructose, also known as D-fructose or levulose, is found naturally in foods as a monosaccharide, as part of the disaccharide sucrose, and as a component of plant oligosaccharides. Fructose is considerably sweeter than sucrose and its use enhances the flavors and physical appeal (eg, color stability, humectancy, and freezing point depression) of many foods and beverages (1,2). Because of its intense sweetness, fructose may be used in place of sucrose and other carbohydrates to reduce overall carbohydrate and energy content of dietetic products (2,3). The consumption of fructose has increased greatly in the United States, primarily as a result of increased use of high-fructose corn syrup in soft drinks and various confections. Portion sizes in general have increased in the United States (4,5), and consumption of fruit juices and drinks has also contributed to the increase in fructose intake (5-9). Early estimates of the daily intake of fructose, based on the 1977 to 1978 US Department of Agriculture Nationwide Food Consumption Survey, showed averages ranged from 15 g for infants to 54 g for men aged 15 to 18 years, and the average intake was 37 g for the total population (6). Fructose intake from high-fructose corn syrup and to a lesser extent fruit has continued to increase (6,9), but current estimates of the total fructose intake are not available. A potential consequence of increased consumption of fructose is gastrointestinal (GI) distress. Fructose empties from the stomach more rapidly than other sugars, fructose is more slowly absorbed than glucose (10-12),

Journal of the AMERICAN DIETETIC ASSOCIATION

1559

and less water and fewer electrolytes are absorbed during its transport across the intestine than during absorption of the same amount of glucose (10,11). When consuming foods and beverages with fructose as the dominant sugar, the capacity for absorption of fructose in the small intestine can easily be exceeded (10-12). When sufficient amounts of fructose are made available to the colonic flora, organic acids and gases are produced and symptoms such as bloating, rumbling, flatulence, and diarrhea can occur (12,13). Hydrogen produced during fermentation is absorbed into the bloodstream and excreted through the lungs. The amount of fructose malabsorbed can be estimated by measuring the concentration of hydrogen in the breath after consumption of a known amount of fructose. The same test is used to measure maldigestion/malabsorption of lactose and other carbohydrates (8,14-16).

The consumption of fructose has increased greatly in the United States, primarily as a result of increased use of high-fructose corn syrup in soft drinks and various confections. The amount of fructose that would produce significant malabsorption and GI symptoms, at least in some adults, has been reported to range from ⬍25 g to 50 g (1,12). A 16-oz bottle of apple juice may contain ⬎30 g fructose and a 22-oz soft drink could contain approximately 30 to 40 g depending on the percent fructose in the corn syrup sweetener (17). Juices and soft drinks with portion sizes ranging from 8 to 64 oz can now be easily found in grocery stores, convenience stores, and vending units. According to Perman (8), heavy consumers of fructose-sweetened beverages may consume 60 to 100 g fructose daily. The fact that ingestion of a significant amount of fructose has the potential to cause GI distress has been reported before. What has changed in recent years is the increase in portion sizes and frequency of consumption of high-fructose foods. Nutrition and GI practitioners made it a practice to warn persons with altered GI function (eg, short bowel syndrome, inflammatory bowel disease, irritable bowel) to limit the amount of fructose consumed, but there were little data regarding the amount of fructose that might result in malabsorption and/or elicit symptoms in both normal people or people with GI disorders. The goals of our study were to determine the frequency of fructose malabsorption and GI symptoms in normal, healthy people using two doses of fructose (25 g and 50 g) that were within amounts consumed in common beverages/foods today; to measure the highest value in breath hydrogen concentration (peak); to identify when the peaks occurred and describe the overall breath hydrogen response (ie, onset, height, and duration) by calculating the area under the curve (AUC); and to determine whether the hydrogen peaks, AUC, and/or the occurrence of symptoms were interrelated for either dose of fructose. These results would provide some additional normative

1560

October 2005 Volume 105 Number 10

data to those already available in the literature plus provide a basis for subsequent studies in both normal people and people with GI disorders using different forms of fructose and fructose in combination with other foods. SUBJECTS AND METHODS Fifteen subjects were selected for the study and met the following inclusion and exclusion criteria: the subject must have been older than age 18 years; nonsmoker; free of GI disorders or diseases, hereditary fructose intolerance, diabetes, and pulmonary disease; and had not recently used laxatives, antibiotics, or medications that would likely alter GI motility. Subjects were recruited from the medical center faculty, staff, and students. Subjects were provided an explanation of the nature of the study and each volunteer signed consent forms approved by the university medical center’s compliance boards. Each subject consumed both the 25-g and 50-g doses of fructose and served as their own controls. A total of 30 doses of fructose were given and 30 breath hydrogen tests were performed. A washout period of at least 3 days occurred between studies. At the recommendation of the human subject committee, subjects consumed the 25-g dose first in case of adverse symptoms and subjects were informed of the potential symptoms of fructose consumption that had already been reported in the literature. Subjects fasted a minimum of 8 hours before the test. A baseline breath sample was collected just before the consumption of the fructose and subsequently at each 30minute interval after the fructose dose up to 3 hours. After subjects exhaled into plastic collection bags, empty syringes were attached to a one-way valve and two 20-mL samples of expired air were removed from the bag. Fructose solutions were prepared just before consumption with pure crystalline fructose (Estee Fructose, Hain Food Group, Inc, Uniondale, NY) and 200 mL tap water at approximately 20°C to 22°C. A total of seven breath samples were collected from each subject using a doublechambered collection bag designed to ensure that alveolar rather than tidal breath samples were collected (GaSampler collection unit, Quintron Instrument Co Inc, Milwaukee, WI). Samples were analyzed using a Quintron Model DP MicroLyzer (Quintron Instrument Co Inc, Milwaukee, WI) that was calibrated before each test. Each sample was flushed through the cartridge of the MicroLyzer. Breath hydrogen concentration ⬎20 ppm was used as the criteria for fructose malabsorption to reduce the likelihood of false positives and is commonly used for evaluation of carbohydrate malabsorption (14). Measures Baseline and each 30-minute breath hydrogen measure were recorded for both doses of fructose until 180 minutes had elapsed. Mean hydrogen peak was computed and when peaks occurred was recorded. AUC for breath hydrogen was computed for both doses using breath hydrogen measures from baseline to final collections. Subjective symptoms, including flatus (rectal gas), abdominal pain, borborygmi (referred to as rumbling), and loose stools were rated by the study subjects according to se-

Figure 1. Peak breath hydrogen (H2) after 25 g and 50 g fructose in normal, free-living volunteers (N⫽15). aCutpoint for upper limit of normal is 20 ppm. ***P⬍.009.

Figure 2. Breath hydrogen (H2) area under the curve: 25 g fructose, baseline to 180 minutes in normal, free-living volunteers (N⫽15).

verity (none⫽0; mild⫽1; moderate⫽2; or severe⫽3) using a symptom score sheet. To obtain a sense of normal symptoms and to test subjects’ understanding of the score sheet, subjects completed the form on random days after fasting for 8 hours but without consuming fructose. Subsequently, each subject completed symptom score sheets before and after each fructose dose. Statistics Paired t tests were used to compare mean peak breath hydrogen, AUCs, and symptom scores after consumption of the 25-g and 50-g fructose solutions. Pearson correlation coefficients were calculated to test for relationships among peak breath hydrogen concentrations, breath hydrogen excursion (AUC), and gastrointestinal symptom scores. RESULTS Subjects All 15 subjects consumed the 25-g and 50-g fructose doses and completed the breath hydrogen studies without complications. Six subjects were men with a mean weight of 166⫾18.4 lb and a mean age of 40⫾15.8 years. The nine female subjects averaged 140⫾119 lb and age 37⫾13.9 years. All volunteers were white. Peak Hydrogen Levels and Mean Time of Peaks Eight of the 15 subjects (53%) exceeded the minimum level (20 ppm) considered clinical evidence of fructose malabsorption after the 25-g fructose dose, and 11 of the 15 subjects (73%) were above the cutpoint for malabsorption after the 50-g dose. The mean peak hydrogen after

Figure 3. Breath hydrogen (H2) area under the curve: 50 g fructose, baseline to 180 minutes in normal, free-living volunteers (N⫽15). ***P⫽.009.

the 25-g dose was 25.6 ppm and it was 51.3 ppm after the 50-g dose of fructose. The means exceeded the 20 ppm cutpoint for fructose malabsorption despite the fact that some subjects never exceeded the normal level. Mean breath hydrogen peaks were significantly higher after the 50-g dose than after the 25-g dose (P⫽.009) (Figure 1). Baseline hydrogen concentrations before the 25-g (8.7 ppm) and 50-g doses (8.4 ppm) were not significantly

October 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1561

baseline for both the 25-g (P⬍.05) and 50-g dose (P⬍.01). After 50 g fructose, more subjects reported symptoms and mean symptom scores were higher than at the 25-g dose, but the difference was not statistically significant (P⫽.055). Individual symptom scores after 25 g and 50 g fructose are shown in Tables 1 and 2. Relationships (Correlations) among Breath Hydrogen Measures and Symptoms Breath hydrogen peaks and AUC were highly correlated with both the 25-g and 50-g doses; that is, if a subject demonstrated a high peak breath hydrogen (which normally occurred 60 to 90 minutes into the 3-hour test), they also showed a large AUC for the 3-hour breath hydrogen test. Alternatively, if the peak response was low, the AUC tended to be small in comparison. The correlation coefficients between breath hydrogen peaks and AUC were 0.954 for the 25-g dose and 0.975 for the 50-g dose (P⬍.0001 for both).

Figure 4. Symptom scores in normal, free-living volunteers before and after 25 g and 50 g fructose (N⫽15). Symptom scores: normal or usual score, symptoms before and after 25 and 50 g fructose. Total score possible with four symptoms: 0 to 12. aSymptoms ⬎ baseline after 25 g fructose (P⬍.02). bSymptoms ⬎ baseline after 50 g fructose (P⬍.001). cSymptoms marginally greater after 50 g fructose than 25 g fructose (P⫽.055).

different from one another. The average time at which the highest (peak) breath hydrogen was recorded after the 25-g dose of fructose was at 1.5 hours, about halfway into the 3-hour test. After the 50-g dose, the average time of the peak occurred at 114 minutes. AUC at 25-g and 50-g Doses The AUC constructed from the breath hydrogen readings taken each 30 minutes are shown in Figures 2 and 3. The AUC was significantly greater after the 50-g dose than the 25-g dose (P⬍.009). After the 50-g fructose dose the mean breath hydrogen level was still above the malabsorption cutpoint at 180 minutes. Symptom Score Figure 4 includes the symptom scores taken under usual or normal conditions, at baseline before each dose, and after consumption of 25 g and 50 g fructose. Symptoms included gas, abdominal pain, borborygmi, and loose stools. A score of 0 indicated no symptoms, 1⫽mild symptoms, 2⫽moderate symptoms, and 3⫽severe symptoms. Usual score averaged 0.67, baseline symptoms averaged 0.4 before the 25-g dose and 0.4 before the 50-g dose. After 25 g fructose, gas and/or rumbling were the only symptom(s) reported by six of the 15 subjects. Abdominal pain was reported by only one subject after 25 g fructose, but was reported by seven subjects after the 50-g dose. Mean posttest symptom scores were significantly greater than

1562

October 2005 Volume 105 Number 10

Peak Hydrogen Levels, AUC, and Symptom Scores It was expected that if people had high peak breath hydrogen or produced large amounts of hydrogen over time, that symptoms might also be greater. However, no significant correlation was apparent for either measure of breath hydrogen at either dose of fructose. Individual symptom scores and peak hydrogen levels are shown in Tables 1 and 2. DISCUSSION Malabsorption The percent of fructose malabsorbers is consistent with past studies (1,12,13,18), in which about half of adult subjects malabsorbed a 25-g load of fructose and 58% to 80% malabsorbed a 50-g fructose load using the same breath hydrogen cutpoint for malabsorption (20 ppm). Mishkin and colleagues (19) retrospectively reviewed the hydrogen breath tests of 520 patients with dyspepsia and irritable bowel syndrome who had received 25 g fructose. They reported that 53% of their patients demonstrated fructose intolerance but their group used 10 ppm hydrogen as the cutpoint for malabsorption instead of the more rigorous 20-ppm cutpoint more commonly used (9). The actual number of people malabsorbing fructose as measured by hydrogen breath tests may actually be somewhat underestimated because some people may have produced methane rather than hydrogen as the predominant gas during the colonic fermentation of fructose (20,21). Kajs and colleagues (21) demonstrated that about one third of their 37 subjects were methane producers and those subjects showed lower breath hydrogen response after ingestion of fermentable substrates than nonmethane producers. Methane production was attributed to the presence of methanogenic bacteria (Methanobrevibacter smithii) in the colon (21). From the current sample and the work of others, a 25-g dose of crystalline fructose taken in water may indicate a threshold for fructose absorption for at least half of normal, healthy adults.

Table 1. Individual symptom scoresa and peak hydrogen levels in normal, free-living volunteers after 25 g oral fructose (N⫽15) Subject No. Symptom

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Gas Abdominal pain Rumbling Loose stool Peak hydrogen level (ppm)

0 0 0 0 40.5

0 0 1 0 14.5

0 0 0 0 7

1 0 2 0 15

0 0 0 0 48

1 1.5 1 0 5

1 0 1 0 45.5

1 0 1 0 21.6

0 0 0 0 7.5

1 0 1 0 2

1 0 0 0 30

1 0 0 0 66

1 0 1 0 29

1 0 1 0 50

0 0 0 0 3

a

Symptom scores range from 0-3 for each of the symptoms reported; 0⫽no symptoms, 3⫽severe.

Table 2. Individual symptom scoresa and peak hydrogen levels in normal, free-living volunteers after 50 g oral fructose (N⫽15) Subject No. Symptom

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Gas Abdominal pain Rumbling Loose stool Peak hydrogen level (ppm)

0 1 0 2 120

2 1 1 0 82

1 1 1 0 48

0 0 1 0 8.5

0 0 0 0 40.2

1 0 0 0 16

2 3 3 0 26.5

1 0 1 0 43.5

0 0 0 0 8.5

1 1 0 0 84

2 0 0 0 70

0 0 2 0 62.5

1 2 1 0 74.25

1 1 1 0 83

0 0 0 0 2

a

Symptom scores range from 0-3 for each of the symptoms reported; 0⫽no symptoms, 3⫽severe.

Mean Peaks The mean breath hydrogen peaks for all subjects for both the 25-g and 50-g doses were greater than the 20-ppm cutpoint for malabsorption. Mean peaks for the eight fructose malabsorbers after the 25-g dose was 41 ppm and the values ranged from 21 to 66 ppm. After the 50-g dose of fructose, the average peak breath hydrogen for those who malabsorbed fructose was 67 ppm with a range from 26 ppm to 120 ppm. The peaks were similar to those recorded in past studies (1,11,12), although mean peaks were not always reported in other studies. Consistently, breath hydrogen peaks after the 50-g dose were significantly higher than after the 25-g dose, indicating that the higher the fructose dose, the greater chance for malabsorption. The time of the breath hydrogen peaks ranged from about 40 to 140 minutes, with the majority of the peaks occurring at the 60- to 120-minute time periods. In a review of the effects of fructose intake, Rumessen (1) showed peaks ranging from about 50 minutes to about 110 minutes. The mean time of the peaks was reported in few other studies of adult fructose malabsorption. Two of our subjects experienced early breath hydrogen peaks (ie, peaks occurring within the first ½ hour after baseline) that were greater than 20 ppm. It was possibile that the early peaks did not represent entry of fructose into the colon but may have resulted from bacterial overgrowth in the stomach or small intestine (19), emptying of carbohydrate remnants from meals eaten the previous day into the colon, or lack of compliance with the fasting protocol (22,23). Another possibility was that the subjects had earlier peaks because they either had inherently faster gastric emptying or they may have been more active than others, resulting in faster transit of fructose (13,22-25).

None of the 15 subjects demonstrated elevated baseline breath hydrogen levels in either of their tests that would increase doubt that significant consumption of fermentable carbohydrates occurred before the tests. Both early peaks occurred after the 25-g dose. One possibility for earlier peaks for the 25-g dose might be that the lower osmolarity and lower energy density of the lower dose may have allowed more rapid emptying of the fructose from the stomach (10,13,26,27). AUC AUC reflects the overall breath hydrogen excursion over time as a result of the malabsorption and fermentation of the fructose dose. AUC for the 25-g fructose dose captured essentially all of the breath hydrogen response; but after 50 g, the mean breath hydrogen was still above the normal range after 3 hours. In no other studies in adults were statistical comparisons of breath hydrogen areas made after 25 g and 50 g fructose. Mean breath hydrogen levels at any one time (plotted in Figures 2 and 3) for both doses were lower than the respective mean hydrogen peaks (Figure 1) because the mean hydrogen peak was derived from each subject’s highest hydrogen level regardless of when it occurred. Symptom Scores Although symptom scores recorded after the 50-g dose were higher than after the 25-g dose (Figure 4), the difference was only marginally statistically significant (P⫽.055). Ten out of 15 subjects (66%) reported symptoms after the 25-g dose and 12 out of 15 subjects (80%) experienced symptoms after the 50-g dose. It is possible that with a larger sample size differences would be statistically significant.

October 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1563

In general, most of the symptoms reported were mild to modest and loose stools occurred only in one subject after the 50-g dose. A few subjects did report the highest score in some categories (Tables 1 and 2). Gas and rumbling were the predominant symptoms experienced by 10 and 12 subjects after the 25-g and 50-g dose, respectively. Abdominal pain, which was never recorded on pretest scores, was reported by one subject after the 25-g dose and seven subjects after the 50-g dose. Subjects all consumed the 25-g dose first because reviewers of the proposed study were concerned that subjects may have significant adverse effects with the larger dose and may not want to continue. Subjects were also made aware of the potential effects of fructose malabsorption according to standard institutional protocol. As a result, we were concerned that some subjects could have been sensitized or had anticipated symptoms because of the explanation in the consent form. We do not know how much the advanced knowledge or lack of crossover influenced the symptom reporting, but the results were similar to that reported in normal adults previously (1,8,1113). Additional symptoms noted by our subjects were nausea (reported by subject 1 shortly after ingestion of the 50-g dose and by subject 6 after ingestion of both the 25-g and 50-g amounts) and dizziness (reported by subject 13 after 25 g fructose). Because the symptoms occurred only a few minutes after consuming the sugars, the nausea and/or dizziness may have been more related to the osmotic effect of the sugar solution than later effects of fermentation by intestinal bacteria (28). We believe the results reported here and previously in the literature support the contention that at least some forms of fructose in the quantities easily available in today’s marketplace could be responsible for mild GI distress in some people. Young children tend to have more significant clinical symptoms (11,15,16). Doses of 1 to 2 g/kg body weight in infants and small children may cause significant abdominal cramping and diarrhea. Correlations, Peaks, AUC, and Symptoms There were no significant correlations between breath hydrogen peaks and mean symptom scores or AUC and symptom scores. Clinically relevant, however, might be the timing of symptom onset in subjects who expressed symptoms. Several subjects verbalized GI complaints (mainly rumbling and gas) around the time when their hydrogen peak occurred. After symptoms occurred, they tended to subside toward the end of the 3-hour test period. We did not however, record exact times prospectively. Several factors may explain the lack of correlation between breath hydrogen measures and symptoms. First, not all subjects malabsorbed fructose and all subjects were included in the analysis. Second, some may have malabsorbed fructose but produced primarily methane instead of hydrogen as the principal gas (20,21). Third, some may have been habitual consumers of fructose. Chronic fructose intake may have served as a prebiotic material resulting over time in a colonic flora more capable of producing and disposing of the short-chain fatty acids and gases produced (11,13). Little has been reported regarding the adaptive response from fructose. Another explanation may have simply been that some people ex-

1564

October 2005 Volume 105 Number 10

aggerated or attenuated the symptoms they recorded because they knew symptoms could occur. We had asked each subject in advance to note and rate any symptoms produced during and after each test. Finally, we did not control for menstrual stage in female participants and it is possible that those symptoms could have influenced symptom scores in some of the women. There were no differences between men and women in symptom scores or prevalence of malabsorption but the numbers in the sex groups were probably too small to allow conclusive statements. Sex differences in fructose tolerances for normal people have not been adequately studied, but in a study of patients with dyspepsia (19), a slightly higher prevalence of malabsorption was noted in women (56%) than men (45%). Some studies suggest the luteal phase of the menstrual cycle can slow GI transit but overall effects on GI motility or emptying are controversial (29,30). Additional study is warranted to better establish relationships between peaks, AUC, and symptoms.

The form in which fructose is consumed apparently matters. Peak hydrogen levels were highly correlated with AUC for each dose and may therefore predict overall AUC and serve as a simpler, shorter test for fructose malabsorption. Peaks in breath hydrogen normally occurred between 60 and 120 minutes and the overall elevation of breath hydrogen could last more than 180 minutes. From a practical basis, it may require less time and expenditure of resources to use the peak breath hydrogen as the terminal event rather than waiting for the last breath hydrogen to come into normal range. None of the subjects showed second peaks after the highest breath hydrogen reading. CONCLUSIONS Our study and those reported herein support that it is likely that the amounts of fructose consumed today, at least in some forms, could be responsible for adverse GI complaints in normal, healthy people. Approximately half of normal people are likely to malabsorb a significant portion of a 25-g dose of crystalline fructose in water and about three fourths will malabsorb a significant portion of a 50-g dose. The form in which fructose is consumed apparently matters. Absorption of fructose is enhanced in the presence of glucose and some amino acids but may be even more incompletely absorbed in the presence of sugar alcohols (10-13,15,16). When juices that contain more fructose than glucose (eg, apple, pear, watermelon, and some honey and grape juices) or contain alcohol sugars in addition to fructose (eg, pear, apple, plum, and cherry) are consumed, malabsorption and symptoms occur more readily (1,8,12,15,17). On the other hand, if fructose is consumed with equal amounts of glucose, as in most citrus juices, sucrose, and most forms of high fructose corn syrup, fructose absorption is increased and malabsorption is less likely. Little research has been published with fructose absorption from popular beverages containing 42% to 55% or greater concentrations of fructose, and

little has been published regarding the absorption or symptoms associated with consumption of fructose with other foods. The threshold for absorption of fructose appears to vary among people and it may reflect frequency of consumption or genetic differences in absorption and transport mechanisms (11). How frequently one consumes fructose may alter the tolerance or threshold for symptoms of fructose malabsorption, either by changing the colonic flora or enhancing the mechanisms in the small bowel for absorption (11). Logic dictates that increased fructose consumption may be more poorly tolerated by persons with altered GI function or malabsorption such as gastrectomy, short bowel syndrome, inflammatory bowel disease, irritable bowel syndrome, or celiac disease. Fructose malabsorption has been implicated in at least some cases of irritable bowel syndrome but whether malabsorption or frequency or severity of expression is worse than the normal population is controversial (18,19,31). Recent studies demonstrating small bowel bacterial overgrowth in irritable bowel syndrome in the small intestine and resolution with antibiotic therapy could be another explanation for symptom provocation with fructose exposure (32). From a methodology viewpoint, the hydrogen breath tests are commonly done over a 3-hour period to capture the overall breath hydrogen excursion. Fructose tolerance tests might be shortened by terminating the exam after the breath hydrogen has reached its highest peak. A minimum of at least 1 hour might be required to rule out an early peak caused by microbial overgrowth in the proximal GI tract. In addition, the test for fructose intolerance should include measures of methane production to capture those who may produce methane as the predominant sugar. For all the issues mentioned above, additional studies are appropriate to evaluate how fructose ingestion in various forms and in different conditions affects GI function, absorption, and symptoms. It appears that some persons may experience malabsorption and GI distress after consumption of common quantities of fructose-containing foods. Funds for this project were provided by the Dietetics and Nutrition research endowment, University of Kansas Medical Center, Kansas City. References 1. Rumessen JJ. Fructose and related food carbohydrates. Scan J Gastroenterol. 1992:27:819-828. 2. Hanover LM, White JS. Manufacturing, composition and applications of fructose. Am J Clin Nutr. 1993; 58(suppl):S724-S732. 3. Vuilleumier S. Worldwide production of high fructose syrup and crystalline fructose. Am J Clin Nutr. 1993; 58(suppl):S733-S736. 4. Nielsen SJ, Popkin BM. Patterns and trends in food portion sizes, 1977-1998. JAMA. 2003:289:450-459. 5. McCanaby KL, Smiciklas-Wright H, Birch LL, Mitchell DC, Picciano MF. Food portions are positively related to energy intake and body weight in children. J Pediatr. 2002;140:340-347.

6. Park YM, Yetley E. Health effects of dietary fructose. Am J Clin Nutr. 1993;58(suppl):S737-S747. 7. Dennison BA. Fruit juice consumption by infants and children: A review. J Am Coll Nutr. 1996;15(suppl 5):S4-S11. 8. Perman JA. Digestion and absorption of fruit juice carbohydrates. J Am Coll Nutr. 1996;15(suppl 5):S12S17. 9. Kantor LS. A dietary assessment of the US food supply. Washington, DC: Agricultural Economic Report No. 772, US Department of Agriculture, Economic Research Service. Available at: http://www.ers.usda. gov/publications/aer772/. Accessed August 17, 2005. 10. Moukarzel AA, Sabri MT. Gastric physiology and function effects of fruit juices. J Am Coll Nutr. 1996; 15(suppl 5):S18-S25. 11. Corpe CP, Burant CF, Hoekstra JH. Intestinal fructose absorption: Clinical and molecular aspects. J Pediatr Gastroenterol Nutr. 1999:364-374. 12. Riby JE, Takuji F, Kretchmer N. Fructose absorption. Am J Clin Nutr. 1993;58(suppl):S748-S753. 13. Southgate DAT. Digestion and metabolism of sugars. Am J Clin Nutr. 1995;62(suppl):S203-S211. 14. Solomons NW. The use of the H2 breath-analysis tests in gastrointestinal diagnosis. Curr Concepts Gastroenterol. 1983;8:30-40. 15. Hoekstra JH, Kneepkens CMF, Van Kempen AAMW. Apple juice malabsorption: Fructose or sorbitol? J Pediatr Gastroenterol Nutr. 1993;16:39-42. 16. Hyams JS, Leichtner AM. Apple juice: An unappreciated cause of chronic diarrhea. Am J Dis Child. 1985;139:503-505. 17. Matthews RH, Pehrsson PR, Farhat-Sabat M. Sugar content of selected foods: Individual and total sugars. Washington, DC: US Department of Agriculture, Human Nutrition Information Service; 1987. Publication No. HER 48. 18. Rumessen JJ, Gudmand-Hoyer E. Functional bowel disease: Malabsorption and abdominal distress after ingestion of fructose, sorbitol and fructose-sorbitol mixtures. Gastroenterology. 1988;95:694-700. 19. Mishkin D, Mishkin S, Sablauskas L, Yalovsky M. Fructose and sorbitol malabsorption in ambulatory patients with functional dyspepsia. Comparison with lactose maldigestion/malabsorption. Dig Dis Sci. 1997; 42:2591-2598. 20. Levitt MD, Strocchi A. Factors affecting hydrogen production and consumption by human fecal flora. The critical roles of hydrogen tension methanogenesis. J Clin Invest. 1992;89:1304-1311. 21. Kajs TM, Fitzgerald JA, Buckner RY, Coyle GA, Stinson BS, Morel JG, Levitt MD. Influence of a methanogenic flora on the breath H2 and symptom response to ingestion of sorbitol or oat fiber. Am J Gastroenterol. 1997;92:89-94. 22. Avanzini P, Chezzi C, Corazza GR, Gasbarrini G, Lecchini R, Menozzi MG, Rasciti L, Strocchi A, Vaira D. The diagnosis of small bowel bacterial overgrowth. Reliability of jejunal culture and inadequacy of breath hydrogen testing. Gastroenterology. 1990;98: 302-309. 23. Holt PR, Kotler DP, Rosensweig NS. Modification of the breath hydrogen test: Increased sensitivity for

October 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1565

24. 25.

26. 27. 28. 29.

1566

the detection of carbohydrate malabsorption. J Lab Clin Med. 1982;100:798-804. Sole CC, Noakes TD. Faster gastric emptying for glucose polymer and fructose solutions than for glucose in humans. Eur J Appl Physiol. 1989;58:605-612. Edelbroek MA, Horowitz M, Jones K, Read NW, Smout AJ. The effect of posture and intragastric distribution of oil and aqueous meal components and appetite. Gastroenterology. 1993;105:382-390. Velchik MG, Reynolds JC, Alavi A. The effect of meal energy content on gastric emptying. J Nucl Med. 1989;30:1106-1110. Vist GE, Maughan RJ. The effect of osmolality and carbohydrate content on the rate of gastric emptying of liquids in man. J Physiol. 1995;486:523-531. Karcher RE, Truding RE, Stawick LE. Using a cutoff of ⬍10 ppm for breath hydrogen testing: A review of five years’ experience. Ann Clin Lab Sci. 1999;29:1-8. Horowitz M, Maddern GJ, Chatterton BE, Collins PJ,

October 2005 Volume 105 Number 10

Petrucco OM, Seamark R, Shearman DJ. The normal menstrual cycle has no effect on gastric emptying. Br J Obstet Gynecol. 1985;92:743-746. 30. Knight LC, Parkmen HP, Brown KL, Miller MA, Trate DM, Maurer AH, Fisher RS. Delayed gastric emptying and decreased antral contractility in normal premenopausal women compared with men. Am J Gastroenterol. 1997;92:968-975. 31. Evans PR, Piesse C, Bak YT, Kellow JE. Fructosesorbitol malabsorption and symptom provocation in irritable bowel syndrome: Relationship to enteric hypersensitivity and dysmotility. Scan J Gastroenterol. 1998;33:1158-1163. 32. Pimentel M, Chow EJ, Lin HC. Normalization of lactulose breath testing correlates with symptom improvement in irritable bowel syndrome: A double blind, randomized, placebo-controlled study. Am J Gastroenterol. 2003;98:412-419.