GASTROENTEROLOGY 1992:102:787-792
Sustained Slowing Effect of Lentils on Gastric Emptying of Solids in Humans and Dogs H. C. LIN, N. A. MOLLER, M. M. WOLINSKY, B. H. KIM, J. E. DOTY, and J. H. MEYER Departments of Medicine and Surgery, Cedars-Sinai Medical Center, Los Angeles, California; Veterans Administration Medical Center, Sepulveda, California; and School of Medicine, UCLA, Los Angeles, California
The distal small intestine is an especially potent site for carbohydrate-triggered intestinal inhibition of gastric emptying of solids. Poorly digestible carbohydrates, such as lentils, may escape proximal absorption, travel over time to reach these inhibitory mechanisms, and slow the gastric emptying of a later meal. A slowing effect on gastric emptying may be associated with a lowering effect on postprandial glucose. The aims of this study were to determine (a) whether lentils (a poorly digestible carbohydrate) vs. bread (an easily digestible carbohydrate) eaten as a premeal (with equal amounts of carbohydrates) slow the gastric emptying of a second solid meal taken 4.0-4.5 hours later and (b) whether a slowing effect on the gastric emptying of the second meal is associated with a lower postprandial glucose response. We found that in 7 dogs and 10 humans, gastric emptying of the second meal was delayed after a lentil premeal compared with a bread premeal. However, there was no difference in the glucose response to the second meal under the two conditions. hether a starch reaches the distal small intestine may have a significant effect on gastric emptying. The ileum has been reported to be an especially potent region for carbohydrate triggered intestinal inhibition.’ Recently, we have extended this observation and found this distal potency particularly important for the inhibition of solid emptying.’ Although 10%~20% of ingested carbohydrates has been estimated to be unabsorbed by the terminal ileum, much greater malabsorption is possible with certain poorly digestible starches.3s4 The slower rate of intestinal digestion characteristic of these starches has been correlated with glucose lowering effects in both the immediate and later postprandial periods.5 We hypothesized that eating a slowly digestible starch, such as lentils, may trigger these potent distal mechanisms to result in a sustained delaying effect on gastric emptying. This effect may also contribute to a lower blood glucose response after a later meal.
W
We tested this idea by comparing the gastric emptying and glucose response of a later meal following a premeal containing either a rapidly (bread) or a slowly (lentils) digestible starch. Materials and Methods Similar experiments were carried out in dogs and subjects. On 2 separate days, we studied the gastric emptying of a standardized meal eaten 4.0-4.5 hours after a premeal containing either bread or boiled lentils. The two premeals contained the same amounts of digestible carbohydrates, protein, and fat. In the human studies, blood glucose response to the second meal was also measured. The procedures used in this study were approved by the Review Committees on the Use of Animal or Human Subjects at the Veterans Administration Medical Center, Sepulveda, CA. human
Dog Experiments Seven mongrel dogs were tested in a randomized schedule that balanced dogs across the two premeal conditions. The animals were deprived of food but not water for 18 hours before each experimental day when the dogs were fed two different meals separated by 4 hours. The first meal contained 75 g of either white bread or lentils (dry wt). After soaking over night in XI g of water, the lentils were first cooked in a beaker with 150 g of boiling water. After cooking, any excess water (usually minimal) was decanted. The dogs were then fed a mixture consisting of one of the two starches, 75 g of cooked steak, and 15 g of margarine (Table 1). In the case of the bread premeal, 200 g of water was also added. The second meal was eaten 4 hours later and consisted of 20 g of bread mixed with 75 g of cooked steak, 15 g of margarine, 200 g of water, and 25 g of chicken liver radiolabeled with ‘%Tc (Table I). The dogs were imaged from the left side.2 Each experiment was performed twice in each dog. Each set of the two-part experiment required 2 separate days to complete. The 2 experimental days were at least 2 days apart. Gastric emptying of the second radiolabeled meal was tracked with images taken every 5 minutes for the first 20 minutes and then every 10 minutes for the remainder of the 180 minute 0 1992 by the American Gastroenterological Association 0016-5085/92/$3.00
788 LIN ET AL.
GASTROENTEROLOGYVol. IO&No.3
Table 1. Meal Compositions
of Dog Studies
Premeal (g) Contents Lentils Bread Steak Liver Margarine Water
Lentils
Bread
Second meal (g)
75O
-
-
-
75
20
75
75
-
-
75 25 15 200
15 200b
15 200
NOTE. Lentil or bread premeal followed in 4.5 hours by second meal. “Dry weight. bUsed to cook lentils.
study. Results from the two pairs of tests in each dog were averaged to give one value for the lentil and one value for the bread premeal experiments. Human
Experiments
Ten healthy volunteer subjects were each studied on 2 separate days to complete this protocol. On one day, the subject ate the lentil premeal, and on the other day, the bread premeal, with the order alternated among the subjects. This group consisted of 5 men and 5 women with ages ranging from 30 to 67 years. The subjects were specifically screened to avoid diabetes, significant obesity (>20% over ideal weight), and gastrointestinal diseases. All but 1 subject were free of all medications. The exception was on long-term therapy using Dyazide and Synthroid. Results of thyroid function studies were within normal values 1 week before his participation in this study. Three of the 5 women were postmenopausal. None were on hormone replacement therapy. The other 2 were tested within 2 weeks of their last menstrual period. Of the whole group, all but 2 subjects completed their &day set of experiments within a period of 1 week. Because of scheduling difficulties, the 2 exceptions were tested with 1 and 3 months separating their two studies. The first meal (Table 2) was served at 12 noon and contained either 140 g of white bread or 125 g of lentils (dry wt). After soaking overnight in 150 g of water, the lentils were first cooked a beaker with 150 g of boiling water. The subjects then ate a mixture consisting of one of the two starches, 75 g of cooked steak, and 15 g of margarine (Table 2). The order of the premeal was randomized so that the premeal eaten on the first day consisted of either lentils or bread. In the case of the bread premeal, 300 g of water was also added. The second meal (Table 2) was eaten 4.5 hours later (430 to 5 PM) and consisted of 70 g of bread mixed with 75 g of cooked steak, 15 g of margarine, 200 g of water and 25 g of chicken liver radiolabeled with 99mTc. Gastric emptying was tracked with both anterior and the posterior images taken every 5 minutes for the first 30 minutes and then every 10 minutes for the remainder of the 180-minute study. The images were taken with subjects in the standing position. Between imaging, the subjects remained seated.
Blood samples were drawn at 20-minute intervals from a 19F butterfly needle inserted into a forearm vein and kept open with a slow infusion of 0.15 mol/L NaCl. This continued until the completion of the gastric emptying study (180 minutes). The first sample was drawn at the start of the meal (time zero). The samples were immediately centrifuged with the serum removed. The samples were then frozen until blood glucose levels were assayed using Hexokinase Assay (Beckman Scientific, Fullerton, CA). Preparation
of Radiolabeled
Liver
For each experiment, 40 g of liver was labeled by injecting subcapsularly -1.0 mCi of ““Tc-sulfur colloid into the uncooked meats6 This was then sealed in a pouch and cooked by boiling. Twenty-five grams of cooked liver was used for each study. All meat used (liver and steak) was first cut into IO-mm cubes. Data Analyses In the human experiments, a small external marker with -0.005 mCi ““‘Tc sulfur colloid was taped over the lower left anterior chest to serve as a reference point by which gastric regions of interest could be aligned from frame to frame. In the dog experiments, the positioning of the animal next to the camera was kept constant so a marker was not needed. The gastric region of interest was then selected with visual inspection of each image. Gastric emptying data was corrected for radioactive decay and septal penetration as previously described.’ The results were expressed as the percentage of the highest total abdominal counts retained in the stomach during the first 20 minutes. Each experiment was then represented by the area under the 180-minute gastric emptying course. The area was estimated by the trapezoidal ruleeB Gastric emptying is characterized by wide day-to-day and subject-tosubject variability.g The distribution of areas under the curve (AUC) is skewed to the right with larger variances under conditions with higher measures (this is typical of area measurements). Because standard statistical tests assume the data are normal in distribution and variances are
Table 2. Meal Compositions
of Human
Studies
Premeal (g) Contents Lentils Bread Steak Liver Margarine Water Carbohydrate”
Lentils 125’
75
Bread
Second meal (g)
-
-
140 75
70 75 25 15 200 38
15 300b 64
15 300 57
NOTE. Lentil or bread premeal followed in 4.5 hours by second meal. “Dry weight. bUsed to cook lentils. ‘Amyloglucosidase and hexokinase assays.
SUSTAINED
March 199’2
homogeneous across groups with different means, the skewness of gastric emptying AUC was reduced by taking the square root of the areas (sqrt AUC). Sqrt AUC values vary between 0 (complete emptying at time 0) and 41 (no emptying at 180 minutes). In addition, the time course of percentage in stomach vs. minutes after 15% emptied could be closely fitted to a linear regression. The slopes of these lines (percent in the stomach per minute) were calculated. The time to 50% remaining (T-50%) were calculated from the slopes and intercepts of these regressions. For statistical analysis, T-50% that exceeded 190 minutes were calculated as 190 minutes. Using the paired t test, sqrt AUC, T-50%, and slope values representing the gastric emptying of the second meal following lentils vs. bread premeals were compared. The area under the plasma glucose response time course over 180 minutes was also calculated by the trapezoidal rule.’ The glucose response after the lentil premeal was then compared with the bread premeal results using the paired t test. Results Compared with bread, lentils eaten as a premeal slowed the gastric emptying of a second meal taken 4.5 hours later. In both dogs and humans, the difference was highly significant (P < 0.025 by comparing T-50% or sqrt AUC, and P < 0.05 in humans and P < 0.005 in dogs by slope comparisons, Tables 3 and 4, Figures 1 and 2). The slowing effect of the lentil premeal was very profound in 5 of 10 human subjects. In this group of 5, the T-50% following lentils greatly exceeded 190 minutes, whereas the T-50% following bread ranged between 87 and 124 minutes. Similar effect was also noted in 3 of 7 dogs. Slope analysis further emphasized the sustained slowing effects of lentils. In humans, slope values of c-0.35 (percent in stomach per minute) were not observed at all following bread but were found in 5 of the 10 subjects after the lentil premeal. In 2 of the 7
SLOWING
OF SOLID EMPTYING
789
dogs and 3 of the 10 humans, emptying was slightly faster after the lentil meal. In humans, 4 of the 5 subjects who showed profound slowing by lentil premeal (T-50% > 190 minutes) were female (Table 3). Two of these 4 subjects were postmenopausal. This gender pattern was not obvious in dogs (Table 4). The blood glucose responses after the two premeals were not different (P > 0.25; Figure 3). There was also no difference in the postprandial plasma glucose between the subjects who showed slowing of emptying by lentil premeal when compared with those who did not. Discussion Carbohydrates vary considerably in their effects on the gastrointestinal tract. In this study, lentils eaten 4.0-4.5 hours earlier slowed the gastric emptying of a second meal significantly more than the bread premeal. In both the human and the dog studies, slowing by lentils was shown (Figures 1 and 2). In the human study, 7 subjects emptied slower following the lentil premeal (Table 3). By calculating T-50% and slopes of the emptying courses, the potent inhibitory effects of a lentil premeal could be easily appreciated. Five of these 7 subjects showed profound slowing by lentils with T-50% > 190 minutes (Table 3). Similar results were also found in the dogs (Table 4). The calculated T-50% values greatly exceeded 190 minutes. By using 190 minutes for the statistical analysis, our estimation of the difference between the two groups was probably conservative. Although the source of carbohydrates has been correlated acutely with substantial differences in the gastric emptying rate of a meal,‘“~‘l this is the first demonstration of the prolonged effect of a starch on the gastric emptying of a later meal.
Table 3. Gastric Emptying of a Meal 4.5 Hours After a Premeal of Bread Versus Lentils in 10 Human Subjects No./Sex l/F 2/F 3/F 4/F 5/M 6/M 7/M 8/F 9/M 10/M Mean + SE
Sqrt AUC bread
Sqrt AUC lentils
T-50% bread
T-50% lentils
Slope bread
Slope lentils
27.4 28.6 32.3 33.0 33.4 23.3 31.7 34.4 31.2 31.9
39.7 38.6 36.6 39.2 38.4 29.6 33.5 33.2 28.9 30.1
87 95 119 122 124 69 111 141 110 117
>190 >190 >190 >190 >190 97 125 124 96 100
-0.58 -0.65 -0.45 -0.51 -0.52 -0.59 -0.59 -0.47 -0.49 -0.46
-0.0 -0.34 -0.27 -0.23 -0.21 -0.53 -0.52 -0.52 -0.52 -0.64
30.7 + 1.1
34.8b + 1.3
109 k 6
149b k 14
“P -c 0.05 vs. bread (paired t test). bP i 0.025 vs. bread (paired t test).
-0.53
k 0.02
-0.38” k 0.06
790
LIN ET AL.
GASTROENTEROLOGY
Vol. 102, No. 3
Table 4. Average Gastric Emptying of Meals 4.0 Hours After Premeals of Bread Versus Lentils in 7 Dogs No./Sex
Sqrt AUC
Sqrt AUC
T-50%
T-50%
Slope
Slope
bread
lentils
bread
lentils
bread
lentils
32.3 35.1 33.2 28.1 30.2 38.4 31.6
38.1 39.6 36.1 36.4 32.8 38.2 31.5
124 163 136 92 103 ,190 115
>190 >190 >190 188 129 >190 115
-0.42 -0.34 -0.37 -0.47 -0.52 -0.15 -0.47
-0.23 -0.14 -0.16 -0.30 -0.40 -0.13 -0.47
32.7 k 1.3
36.1’ zk 1.1
l/F 2/M 3/M 6/M 4/F 5/F 7/M Mean + SE
NOTE. Each experiment was performed “P < 0.025 vs. bread (paired t test). bP < 0.005 vs. bread [paired t test).
132 f
3
170° + 13
-0.39
f 0.05
-0.26b + 0.05
twice. Duplicate values were averaged.
In the group of 5 human subjects with T-50% >190 minutes following lentil premeal, 4 were female. Two of these 4 were postmenopausal and not on hormone replacement therapy. Because postmenopausal women off sex steroid therapy have been reported to empty solids at the same rate as men,” the greater representation by women may not indicate a significant gender difference in the slowing effect by lentils. Furthermore, neither age nor gender distinguished the 4 subjects who failed to respond. By randomizing the order of the premeals and completing the Z-day studies in these subjects within 2 weeks of the start of their menstrual period, the effect of their menstrual cycle on the emptying result for the premenopausal group was also minimized.
Recent attention has been focused on differences in the digestibility of starch. Jenkins et al. noted that the glucose lowering effects were inversely correlated with the rate of in vitro liberation of maltose and maltotriose.13 A significant relationship was found between slower rates of sugar release and improved glucose tolerance in humans. Lentils are digested slowly compared with bread. Many factors may delay the rate of digestion of lentils and other low glycemic starches. *4-2*With lentils, the protein coat seems to delay digestion.17 By decreasing starch hydrolysis, these variables can be expected to increase the amount of unabsorbed starch and to continue the release of digestive products from starch in distal bowel, even hours after a meal.
loo-
100-l go-
go-
80-
80-
$
70-
3 a
70-
g
60-
5
60-
:
50-
:
50-
t
40-
E
40-
Y
30-
!?
30-
* 0
20-
s
2010-
10-
0:
I,
0
I,
I,
30
60
1,
90 120 MINUTES
I,
-
150
1
180
Figure 1. Gastric emptying courses of the second meal eaten 4.5 hours after premeal of either lentils (W)or bread (0) in humans (data are means).
of 0
I,
I,
30
I,
60
I,
90 120 MINUTES
1,
0
150
1
180
Figure 2. Gastric emptying courses of the second meal eaten 4.0 hours after premeal of either lentils [B) or bread (0) in dogs (data are means).
SUSTAINED SLOWING OF SOLID EMPTYING
March 1992
g
60
8 3 (3
40
20 1
1
00 0
30
60
90
120
150
180
MINUTES Figure 3. Glucose time course following the second meal with premeal of lentils (B) or bread (0) (data are means).
Several studies have confirmed these ideas Using a multilumen tube positioned near the ileocecal valve for recovery, 2%-20% of dietary starch was found to escape absorption by the small intestine in normal subjects.4 A similar conclusion was reached indirectly with the use of hydrogen test3 Greater malabsorption was found with macaroni (less digestible) than with rice bread (easily digestible). These observations emphasized the idea that enzymatic digestion of starch depends on the accessibility of the starch substrate to the luminal enzymes.” Slowly digestible starches such as lentils, by escaping complete digestion and absorption in the proximal small intestine, may then reach the distal ileum to release maltose hours after a meal. Because digestive products of starch in the distal ileum are two to three times as potent at inhibiting gastric emptying of solids’ as the same products in the jejunum, we postulated that a premeal of lentils would reduce the rate of gastric emptying of a test meal taken 4.5 hours later. By triggering these potent distal inhibitory mechanisms and slowing the entry of nutrients into the small intestine, blood glucoses may be lowered after a second meal. The addition of a poorly digestible starch to a meal slows the gastric emptying of that meal.‘“~” In our study, we found that the inhibitory effect was sustained for more than 4.0-4.5 hours so that the gastric emptying of a second meal was also slowed. By escaping digestion proximally, the lentils from the first meal could have slowed the gastric emptying of both meals by reaching the potent distal inhibitory
791
mechanism. At the time of the second meal, all the lentils from the premeal may not have reached the distal ileum; some may still be found in the stomach and the proximal small intestine. The inhibitory effect can then be potentially sustained for an extended period of time as more lentils travel to reach the distal gut. Although we observed an effect of lentils on gastric emptying in this group of normal volunteers, we were not able to show a glucose lowering effect. Because no postprandial glucose peak was seen after either premeal, the carbohydrate load eaten during the second meal (38 g) was probably insufficient to show any glucose-lowering effect from a meal eaten 4.5 hours earlier. Ileal feedback may also affect the gastric emptying of the starches differently than the liver marker tracked with the gamma camera. Because up to 60% of bread is already digested in the stomach by salivary amylase,23 whereas only lO%15% of meat is hydrolyzed by pepsin during gastric residence,24 our failure to observe an effect on glucose tolerance may have resulted from the ileal inhibition slowing the gastric emptying of insoluble liver particles while not affecting the rate of emptying of soluble digestive products of bread. Furthermore, because the complete digestion of lentils is more time consuming than that of bread, 4-7 hours after the lentil premeal, a reduction in the entry of glucose from the second meal might have been compensated by the continued entry of glucose from the lentils eaten earlier. This compensatory effect would then negate any plasma glucose lowering effect of the lentil premeal on the second meal. Alternatively, because plasma glucose is tightly regulated by multifactorial regulations in our nondiabetic subjects, the differences in the gastric emptying of a meal may not be reflected in the plasma glucose levels. References 1. Jain NK, Boivin M, Zinsmeister
AR, Brown ML, Malagelada JR, DiMagno EP. Effect of ileal perfusion of carbohydrates and amylase inhibitor on gastrointestinal hormones and emptying. Gastroenterology 1989;96:377-387. 2. Lin HC, Kim BH, Elashoff JD, Doty JE, Gu Y-G, Meyer JH. Gastric emptying of solid food is most potently inhibited by carbohydrate in the canine distal ileum. Gastroenterology 1992;102:430-438. 3. Anderson IH, Levine AS, Levitt MD. Incomplete absorption of the carbohydrate in all-purpose wheat flour. N Engl J Med 1981;304:891-892. 4. Stephen AM, Haddad AC, Phillips SF. Passage of carbohydrate into the colon. Gastroenterology 1983;85:589-595. 5. Jenkins DJA, Wolever TMS, Taylor RH, Griffiths C, Krzeminska K, Lawrie JA, Bennett CM, Goff DV, Sarson DL, Bloom SR. Slow release dietary carbohydrate improves second meal tolerance. Am J Clin Nutr 1982;35:1339-1346. 6. Wright RA, Thompson D, Syed I. Simultaneous markers for fluid and solid gastric emptying. New variations on an old theme: Concise communication. J Nucl Med 1981;22:772-776.
792
LIN ET AL.
GASTROENTEROLOGY Vol. 102, No. 3
7. Meyer JH, Dressman J, Fink AS, Amidon G. Effect of size and
8.
9.
10.
11.
12.
13.
14.
density on canine gastric emptying of non-digestible solids. Gastroenterology 1989;89:805-813. Lin HC, Doty JE, Reedy TJ, Meyer JH. Inhibition of gastric emptying by glucose depends on the length of the intestine exposed to the nutrient. Am J Physiol 1989;256:G204-G411. Collins PJ, Horowite M, Cook DJ, Harding PE, Shearman DJC. Gastric emptying in normal subjects-A reproducible technique using a single scintillation camera and computer system. Gut 1983;24:1117-1125. Mourot J, Thouvenot P, Couet C, Antoine JM, Krobicka A, Debry G. Relationship between the rate of gastric emptying and glucose and insulin responses to starchy foods in young healthy adults. Am J Clin Nutr 1988;48:1035-1040. Torsdottir I, Alpsten M, Andersson D, Brummer RJM, Andersson H. Effect of different starchy foods in composite meals on gastric emptying rate and glucose metabolism. Hum Nutr Clin Nutr 1984;38C:329-338. Hutson WR, Roehrkasse RL, Wald A. Influence of gender and menopause on gastric emptying and motility. Gastroenterology 1989;96:11-17. Jenkins DJA, Ghafari H, Wolever TMS, Taylor RH, Jenkins AL, Barker HM, Fielden H, Bowling AC. Relationship between rate of digestion of foods and post-prandial glycaemia. Diabetologia 1982;22:450-455. Englyst HN, Trowel1 H, Southgate DAT, Cummings JH. Dietary fiber and resistant starch. Am J Clin Nutr 1987;46:873-
874. 15. Holm J, Lundquist
I, Bjorck I, Eliasson AC, George N. Degree of starch gelatinization, digestion rate of starch in vitro, and metabolic response in rats. Am J Clin Nutr 1988;47:1010-
1016. 16. Jaffe WIG, Lette CLV. Heat-labile
growth inhibitory factors in beans (Phaseolus Vulgaris). J Nutr 1986;94:203-210. 17. Jenkins DJA, Thorne MJ, Wolever TMS, Jenkins AL, Rao AV, Thompson LU. The effect of starch-protein interaction in
18.
19.
20.
21.
22.
23.
24.
wheat on the glycemic response and rate of in vitro digestion. Am J Clin Nutr 1987;45:946-951. Jenkins DJA, Thorne MJ, Camelon K, Jenkins A, Rao AV, Taylor RH, Thompson LLJ, Kalmusky J, Reichert R, Francis T. Effect of processing on digestibility and the blood glucose response: a study of lentils. Am J Clin Nutr 1982;36:1093-1101. Layer P, Carlson GL, DiMagno EP. Partially purified white bean Amylase inhibitor reduces starch digestion in vitro and inactivates intraduodenal amylase in humans. Gastroenterology 1985;88:1895-1902. Wolever TMS, Jenkins DJA, Kalmusky J, Giordano C, Giudio S, Jenkins AL, Thompson LU, Wong GS, Jesse RG. Glycemic response to pasta: effect of surface area, degree of cooking, and protein enrichment. Diabetes Care 1986;9:401-403. Yoon JH, Thompson LU, Jenkins DJA. The effect of phytic acid on in vitro rate of starch digestibility and blood-glucose response. Am J Clin Nutr 1983;38:835-842. Wong S, O’Dea K. Importance of physical form rather than viscosity in determining the rate of starch hydrolysis in legumes. Am J Clin Nutr 1983;37:66-70. James AH. The nature of the gastric contents in man. in: James AH. The physiology of gastric digestion. London: Edward Arnold, 1957:1-24. Meyer JH, Kelly KA. Canine pancreatic responses to intestinally perfused proteins and protein digest. Am J Physiol 1976;231:G682-G691.
Received November 30, 1990. Accepted August 15, 1991. Address requests for reprints to: Henry C. Lin, M.D., CedarsSinai Medical Center, Suite 7511, 8700 Beverly Boulevard, Los Angeles, California 90048-1869. Supported by Veterans Administration Research Funds. Dr. Lin was a Veterans Affairs Associate Investigator when these studies were performed. The authors thank Don Jehn for expert technical assistance.
Sustained Slowing Effect of Lentils on Gastric Emptying of Solids in Humans and Dogs H. C. LIN, N. A. MOLLER, M. M. WOLINSKY, B. H. KIM, J. E. DOTY, and J. H. MEYER Departments of Medicine and Surgery, Cedars-Sinai Medical Center, Los Angeles, California; Veterans Administration Medical Center, Sepulveda, California; and School of Medicine, UCLA, Los Angeles, California
The distal small intestine is an especially potent site for carbohydrate-triggered intestinal inhibition of gastric emptying of solids. Poorly digestible carbohydrates, such as lentils, may escape proximal absorption, travel over time to reach these inhibitory mechanisms, and slow the gastric emptying of a later meal. A slowing effect on gastric emptying may be associated with a lowering effect on postprandial glucose. The aims of this study were to determine (a) whether lentils (a poorly digestible carbohydrate) vs. bread (an easily digestible carbohydrate) eaten as a premeal (with equal amounts of carbohydrates) slow the gastric emptying of a second solid meal taken 4.0-4.5 hours later and (b) whether a slowing effect on the gastric emptying of the second meal is associated with a lower postprandial glucose response. We found that in 7 dogs and 10 humans, gastric emptying of the second meal was delayed after a lentil premeal compared with a bread premeal. However, there was no difference in the glucose response to the second meal under the two conditions. hether a starch reaches the distal small intestine may have a significant effect on gastric emptying. The ileum has been reported to be an especially potent region for carbohydrate triggered intestinal inhibition.’ Recently, we have extended this observation and found this distal potency particularly important for the inhibition of solid emptying.’ Although 10%~20% of ingested carbohydrates has been estimated to be unabsorbed by the terminal ileum, much greater malabsorption is possible with certain poorly digestible starches.3s4 The slower rate of intestinal digestion characteristic of these starches has been correlated with glucose lowering effects in both the immediate and later postprandial periods.5 We hypothesized that eating a slowly digestible starch, such as lentils, may trigger these potent distal mechanisms to result in a sustained delaying effect on gastric emptying. This effect may also contribute to a lower blood glucose response after a later meal.
W
We tested this idea by comparing the gastric emptying and glucose response of a later meal following a premeal containing either a rapidly (bread) or a slowly (lentils) digestible starch. Materials and Methods Similar experiments were carried out in dogs and subjects. On 2 separate days, we studied the gastric emptying of a standardized meal eaten 4.0-4.5 hours after a premeal containing either bread or boiled lentils. The two premeals contained the same amounts of digestible carbohydrates, protein, and fat. In the human studies, blood glucose response to the second meal was also measured. The procedures used in this study were approved by the Review Committees on the Use of Animal or Human Subjects at the Veterans Administration Medical Center, Sepulveda, CA. human
Dog Experiments Seven mongrel dogs were tested in a randomized schedule that balanced dogs across the two premeal conditions. The animals were deprived of food but not water for 18 hours before each experimental day when the dogs were fed two different meals separated by 4 hours. The first meal contained 75 g of either white bread or lentils (dry wt). After soaking over night in XI g of water, the lentils were first cooked in a beaker with 150 g of boiling water. After cooking, any excess water (usually minimal) was decanted. The dogs were then fed a mixture consisting of one of the two starches, 75 g of cooked steak, and 15 g of margarine (Table 1). In the case of the bread premeal, 200 g of water was also added. The second meal was eaten 4 hours later and consisted of 20 g of bread mixed with 75 g of cooked steak, 15 g of margarine, 200 g of water, and 25 g of chicken liver radiolabeled with ‘%Tc (Table I). The dogs were imaged from the left side.2 Each experiment was performed twice in each dog. Each set of the two-part experiment required 2 separate days to complete. The 2 experimental days were at least 2 days apart. Gastric emptying of the second radiolabeled meal was tracked with images taken every 5 minutes for the first 20 minutes and then every 10 minutes for the remainder of the 180 minute 0 1992 by the American Gastroenterological Association 0016-5085/92/$3.00
788 LIN ET AL.
GASTROENTEROLOGYVol. IO&No.3
Table 1. Meal Compositions
of Dog Studies
Premeal (g) Contents Lentils Bread Steak Liver Margarine Water
Lentils
Bread
Second meal (g)
75O
-
-
-
75
20
75
75
-
-
75 25 15 200
15 200b
15 200
NOTE. Lentil or bread premeal followed in 4.5 hours by second meal. “Dry weight. bUsed to cook lentils.
study. Results from the two pairs of tests in each dog were averaged to give one value for the lentil and one value for the bread premeal experiments. Human
Experiments
Ten healthy volunteer subjects were each studied on 2 separate days to complete this protocol. On one day, the subject ate the lentil premeal, and on the other day, the bread premeal, with the order alternated among the subjects. This group consisted of 5 men and 5 women with ages ranging from 30 to 67 years. The subjects were specifically screened to avoid diabetes, significant obesity (>20% over ideal weight), and gastrointestinal diseases. All but 1 subject were free of all medications. The exception was on long-term therapy using Dyazide and Synthroid. Results of thyroid function studies were within normal values 1 week before his participation in this study. Three of the 5 women were postmenopausal. None were on hormone replacement therapy. The other 2 were tested within 2 weeks of their last menstrual period. Of the whole group, all but 2 subjects completed their &day set of experiments within a period of 1 week. Because of scheduling difficulties, the 2 exceptions were tested with 1 and 3 months separating their two studies. The first meal (Table 2) was served at 12 noon and contained either 140 g of white bread or 125 g of lentils (dry wt). After soaking overnight in 150 g of water, the lentils were first cooked a beaker with 150 g of boiling water. The subjects then ate a mixture consisting of one of the two starches, 75 g of cooked steak, and 15 g of margarine (Table 2). The order of the premeal was randomized so that the premeal eaten on the first day consisted of either lentils or bread. In the case of the bread premeal, 300 g of water was also added. The second meal (Table 2) was eaten 4.5 hours later (430 to 5 PM) and consisted of 70 g of bread mixed with 75 g of cooked steak, 15 g of margarine, 200 g of water and 25 g of chicken liver radiolabeled with 99mTc. Gastric emptying was tracked with both anterior and the posterior images taken every 5 minutes for the first 30 minutes and then every 10 minutes for the remainder of the 180-minute study. The images were taken with subjects in the standing position. Between imaging, the subjects remained seated.
Blood samples were drawn at 20-minute intervals from a 19F butterfly needle inserted into a forearm vein and kept open with a slow infusion of 0.15 mol/L NaCl. This continued until the completion of the gastric emptying study (180 minutes). The first sample was drawn at the start of the meal (time zero). The samples were immediately centrifuged with the serum removed. The samples were then frozen until blood glucose levels were assayed using Hexokinase Assay (Beckman Scientific, Fullerton, CA). Preparation
of Radiolabeled
Liver
For each experiment, 40 g of liver was labeled by injecting subcapsularly -1.0 mCi of ““Tc-sulfur colloid into the uncooked meats6 This was then sealed in a pouch and cooked by boiling. Twenty-five grams of cooked liver was used for each study. All meat used (liver and steak) was first cut into IO-mm cubes. Data Analyses In the human experiments, a small external marker with -0.005 mCi ““‘Tc sulfur colloid was taped over the lower left anterior chest to serve as a reference point by which gastric regions of interest could be aligned from frame to frame. In the dog experiments, the positioning of the animal next to the camera was kept constant so a marker was not needed. The gastric region of interest was then selected with visual inspection of each image. Gastric emptying data was corrected for radioactive decay and septal penetration as previously described.’ The results were expressed as the percentage of the highest total abdominal counts retained in the stomach during the first 20 minutes. Each experiment was then represented by the area under the 180-minute gastric emptying course. The area was estimated by the trapezoidal ruleeB Gastric emptying is characterized by wide day-to-day and subject-tosubject variability.g The distribution of areas under the curve (AUC) is skewed to the right with larger variances under conditions with higher measures (this is typical of area measurements). Because standard statistical tests assume the data are normal in distribution and variances are
Table 2. Meal Compositions
of Human
Studies
Premeal (g) Contents Lentils Bread Steak Liver Margarine Water Carbohydrate”
Lentils 125’
75
Bread
Second meal (g)
-
-
140 75
70 75 25 15 200 38
15 300b 64
15 300 57
NOTE. Lentil or bread premeal followed in 4.5 hours by second meal. “Dry weight. bUsed to cook lentils. ‘Amyloglucosidase and hexokinase assays.
SUSTAINED
March 199’2
homogeneous across groups with different means, the skewness of gastric emptying AUC was reduced by taking the square root of the areas (sqrt AUC). Sqrt AUC values vary between 0 (complete emptying at time 0) and 41 (no emptying at 180 minutes). In addition, the time course of percentage in stomach vs. minutes after 15% emptied could be closely fitted to a linear regression. The slopes of these lines (percent in the stomach per minute) were calculated. The time to 50% remaining (T-50%) were calculated from the slopes and intercepts of these regressions. For statistical analysis, T-50% that exceeded 190 minutes were calculated as 190 minutes. Using the paired t test, sqrt AUC, T-50%, and slope values representing the gastric emptying of the second meal following lentils vs. bread premeals were compared. The area under the plasma glucose response time course over 180 minutes was also calculated by the trapezoidal rule.’ The glucose response after the lentil premeal was then compared with the bread premeal results using the paired t test. Results Compared with bread, lentils eaten as a premeal slowed the gastric emptying of a second meal taken 4.5 hours later. In both dogs and humans, the difference was highly significant (P < 0.025 by comparing T-50% or sqrt AUC, and P < 0.05 in humans and P < 0.005 in dogs by slope comparisons, Tables 3 and 4, Figures 1 and 2). The slowing effect of the lentil premeal was very profound in 5 of 10 human subjects. In this group of 5, the T-50% following lentils greatly exceeded 190 minutes, whereas the T-50% following bread ranged between 87 and 124 minutes. Similar effect was also noted in 3 of 7 dogs. Slope analysis further emphasized the sustained slowing effects of lentils. In humans, slope values of c-0.35 (percent in stomach per minute) were not observed at all following bread but were found in 5 of the 10 subjects after the lentil premeal. In 2 of the 7
SLOWING
OF SOLID EMPTYING
789
dogs and 3 of the 10 humans, emptying was slightly faster after the lentil meal. In humans, 4 of the 5 subjects who showed profound slowing by lentil premeal (T-50% > 190 minutes) were female (Table 3). Two of these 4 subjects were postmenopausal. This gender pattern was not obvious in dogs (Table 4). The blood glucose responses after the two premeals were not different (P > 0.25; Figure 3). There was also no difference in the postprandial plasma glucose between the subjects who showed slowing of emptying by lentil premeal when compared with those who did not. Discussion Carbohydrates vary considerably in their effects on the gastrointestinal tract. In this study, lentils eaten 4.0-4.5 hours earlier slowed the gastric emptying of a second meal significantly more than the bread premeal. In both the human and the dog studies, slowing by lentils was shown (Figures 1 and 2). In the human study, 7 subjects emptied slower following the lentil premeal (Table 3). By calculating T-50% and slopes of the emptying courses, the potent inhibitory effects of a lentil premeal could be easily appreciated. Five of these 7 subjects showed profound slowing by lentils with T-50% > 190 minutes (Table 3). Similar results were also found in the dogs (Table 4). The calculated T-50% values greatly exceeded 190 minutes. By using 190 minutes for the statistical analysis, our estimation of the difference between the two groups was probably conservative. Although the source of carbohydrates has been correlated acutely with substantial differences in the gastric emptying rate of a meal,‘“~‘l this is the first demonstration of the prolonged effect of a starch on the gastric emptying of a later meal.
Table 3. Gastric Emptying of a Meal 4.5 Hours After a Premeal of Bread Versus Lentils in 10 Human Subjects No./Sex l/F 2/F 3/F 4/F 5/M 6/M 7/M 8/F 9/M 10/M Mean + SE
Sqrt AUC bread
Sqrt AUC lentils
T-50% bread
T-50% lentils
Slope bread
Slope lentils
27.4 28.6 32.3 33.0 33.4 23.3 31.7 34.4 31.2 31.9
39.7 38.6 36.6 39.2 38.4 29.6 33.5 33.2 28.9 30.1
87 95 119 122 124 69 111 141 110 117
>190 >190 >190 >190 >190 97 125 124 96 100
-0.58 -0.65 -0.45 -0.51 -0.52 -0.59 -0.59 -0.47 -0.49 -0.46
-0.0 -0.34 -0.27 -0.23 -0.21 -0.53 -0.52 -0.52 -0.52 -0.64
30.7 + 1.1
34.8b + 1.3
109 k 6
149b k 14
“P -c 0.05 vs. bread (paired t test). bP i 0.025 vs. bread (paired t test).
-0.53
k 0.02
-0.38” k 0.06
790
LIN ET AL.
GASTROENTEROLOGY
Vol. 102, No. 3
Table 4. Average Gastric Emptying of Meals 4.0 Hours After Premeals of Bread Versus Lentils in 7 Dogs No./Sex
Sqrt AUC
Sqrt AUC
T-50%
T-50%
Slope
Slope
bread
lentils
bread
lentils
bread
lentils
32.3 35.1 33.2 28.1 30.2 38.4 31.6
38.1 39.6 36.1 36.4 32.8 38.2 31.5
124 163 136 92 103 ,190 115
>190 >190 >190 188 129 >190 115
-0.42 -0.34 -0.37 -0.47 -0.52 -0.15 -0.47
-0.23 -0.14 -0.16 -0.30 -0.40 -0.13 -0.47
32.7 k 1.3
36.1’ zk 1.1
l/F 2/M 3/M 6/M 4/F 5/F 7/M Mean + SE
NOTE. Each experiment was performed “P < 0.025 vs. bread (paired t test). bP < 0.005 vs. bread [paired t test).
132 f
3
170° + 13
-0.39
f 0.05
-0.26b + 0.05
twice. Duplicate values were averaged.
In the group of 5 human subjects with T-50% >190 minutes following lentil premeal, 4 were female. Two of these 4 were postmenopausal and not on hormone replacement therapy. Because postmenopausal women off sex steroid therapy have been reported to empty solids at the same rate as men,” the greater representation by women may not indicate a significant gender difference in the slowing effect by lentils. Furthermore, neither age nor gender distinguished the 4 subjects who failed to respond. By randomizing the order of the premeals and completing the Z-day studies in these subjects within 2 weeks of the start of their menstrual period, the effect of their menstrual cycle on the emptying result for the premenopausal group was also minimized.
Recent attention has been focused on differences in the digestibility of starch. Jenkins et al. noted that the glucose lowering effects were inversely correlated with the rate of in vitro liberation of maltose and maltotriose.13 A significant relationship was found between slower rates of sugar release and improved glucose tolerance in humans. Lentils are digested slowly compared with bread. Many factors may delay the rate of digestion of lentils and other low glycemic starches. *4-2*With lentils, the protein coat seems to delay digestion.17 By decreasing starch hydrolysis, these variables can be expected to increase the amount of unabsorbed starch and to continue the release of digestive products from starch in distal bowel, even hours after a meal.
loo-
100-l go-
go-
80-
80-
$
70-
3 a
70-
g
60-
5
60-
:
50-
:
50-
t
40-
E
40-
Y
30-
!?
30-
* 0
20-
s
2010-
10-
0:
I,
0
I,
I,
30
60
1,
90 120 MINUTES
I,
-
150
1
180
Figure 1. Gastric emptying courses of the second meal eaten 4.5 hours after premeal of either lentils (W)or bread (0) in humans (data are means).
of 0
I,
I,
30
I,
60
I,
90 120 MINUTES
1,
0
150
1
180
Figure 2. Gastric emptying courses of the second meal eaten 4.0 hours after premeal of either lentils [B) or bread (0) in dogs (data are means).
SUSTAINED SLOWING OF SOLID EMPTYING
March 1992
g
60
8 3 (3
40
20 1
1
00 0
30
60
90
120
150
180
MINUTES Figure 3. Glucose time course following the second meal with premeal of lentils (B) or bread (0) (data are means).
Several studies have confirmed these ideas Using a multilumen tube positioned near the ileocecal valve for recovery, 2%-20% of dietary starch was found to escape absorption by the small intestine in normal subjects.4 A similar conclusion was reached indirectly with the use of hydrogen test3 Greater malabsorption was found with macaroni (less digestible) than with rice bread (easily digestible). These observations emphasized the idea that enzymatic digestion of starch depends on the accessibility of the starch substrate to the luminal enzymes.” Slowly digestible starches such as lentils, by escaping complete digestion and absorption in the proximal small intestine, may then reach the distal ileum to release maltose hours after a meal. Because digestive products of starch in the distal ileum are two to three times as potent at inhibiting gastric emptying of solids’ as the same products in the jejunum, we postulated that a premeal of lentils would reduce the rate of gastric emptying of a test meal taken 4.5 hours later. By triggering these potent distal inhibitory mechanisms and slowing the entry of nutrients into the small intestine, blood glucoses may be lowered after a second meal. The addition of a poorly digestible starch to a meal slows the gastric emptying of that meal.‘“~” In our study, we found that the inhibitory effect was sustained for more than 4.0-4.5 hours so that the gastric emptying of a second meal was also slowed. By escaping digestion proximally, the lentils from the first meal could have slowed the gastric emptying of both meals by reaching the potent distal inhibitory
791
mechanism. At the time of the second meal, all the lentils from the premeal may not have reached the distal ileum; some may still be found in the stomach and the proximal small intestine. The inhibitory effect can then be potentially sustained for an extended period of time as more lentils travel to reach the distal gut. Although we observed an effect of lentils on gastric emptying in this group of normal volunteers, we were not able to show a glucose lowering effect. Because no postprandial glucose peak was seen after either premeal, the carbohydrate load eaten during the second meal (38 g) was probably insufficient to show any glucose-lowering effect from a meal eaten 4.5 hours earlier. Ileal feedback may also affect the gastric emptying of the starches differently than the liver marker tracked with the gamma camera. Because up to 60% of bread is already digested in the stomach by salivary amylase,23 whereas only lO%15% of meat is hydrolyzed by pepsin during gastric residence,24 our failure to observe an effect on glucose tolerance may have resulted from the ileal inhibition slowing the gastric emptying of insoluble liver particles while not affecting the rate of emptying of soluble digestive products of bread. Furthermore, because the complete digestion of lentils is more time consuming than that of bread, 4-7 hours after the lentil premeal, a reduction in the entry of glucose from the second meal might have been compensated by the continued entry of glucose from the lentils eaten earlier. This compensatory effect would then negate any plasma glucose lowering effect of the lentil premeal on the second meal. Alternatively, because plasma glucose is tightly regulated by multifactorial regulations in our nondiabetic subjects, the differences in the gastric emptying of a meal may not be reflected in the plasma glucose levels. References 1. Jain NK, Boivin M, Zinsmeister
AR, Brown ML, Malagelada JR, DiMagno EP. Effect of ileal perfusion of carbohydrates and amylase inhibitor on gastrointestinal hormones and emptying. Gastroenterology 1989;96:377-387. 2. Lin HC, Kim BH, Elashoff JD, Doty JE, Gu Y-G, Meyer JH. Gastric emptying of solid food is most potently inhibited by carbohydrate in the canine distal ileum. Gastroenterology 1992;102:430-438. 3. Anderson IH, Levine AS, Levitt MD. Incomplete absorption of the carbohydrate in all-purpose wheat flour. N Engl J Med 1981;304:891-892. 4. Stephen AM, Haddad AC, Phillips SF. Passage of carbohydrate into the colon. Gastroenterology 1983;85:589-595. 5. Jenkins DJA, Wolever TMS, Taylor RH, Griffiths C, Krzeminska K, Lawrie JA, Bennett CM, Goff DV, Sarson DL, Bloom SR. Slow release dietary carbohydrate improves second meal tolerance. Am J Clin Nutr 1982;35:1339-1346. 6. Wright RA, Thompson D, Syed I. Simultaneous markers for fluid and solid gastric emptying. New variations on an old theme: Concise communication. J Nucl Med 1981;22:772-776.
792
LIN ET AL.
GASTROENTEROLOGY Vol. 102, No. 3
7. Meyer JH, Dressman J, Fink AS, Amidon G. Effect of size and
8.
9.
10.
11.
12.
13.
14.
density on canine gastric emptying of non-digestible solids. Gastroenterology 1989;89:805-813. Lin HC, Doty JE, Reedy TJ, Meyer JH. Inhibition of gastric emptying by glucose depends on the length of the intestine exposed to the nutrient. Am J Physiol 1989;256:G204-G411. Collins PJ, Horowite M, Cook DJ, Harding PE, Shearman DJC. Gastric emptying in normal subjects-A reproducible technique using a single scintillation camera and computer system. Gut 1983;24:1117-1125. Mourot J, Thouvenot P, Couet C, Antoine JM, Krobicka A, Debry G. Relationship between the rate of gastric emptying and glucose and insulin responses to starchy foods in young healthy adults. Am J Clin Nutr 1988;48:1035-1040. Torsdottir I, Alpsten M, Andersson D, Brummer RJM, Andersson H. Effect of different starchy foods in composite meals on gastric emptying rate and glucose metabolism. Hum Nutr Clin Nutr 1984;38C:329-338. Hutson WR, Roehrkasse RL, Wald A. Influence of gender and menopause on gastric emptying and motility. Gastroenterology 1989;96:11-17. Jenkins DJA, Ghafari H, Wolever TMS, Taylor RH, Jenkins AL, Barker HM, Fielden H, Bowling AC. Relationship between rate of digestion of foods and post-prandial glycaemia. Diabetologia 1982;22:450-455. Englyst HN, Trowel1 H, Southgate DAT, Cummings JH. Dietary fiber and resistant starch. Am J Clin Nutr 1987;46:873-
874. 15. Holm J, Lundquist
I, Bjorck I, Eliasson AC, George N. Degree of starch gelatinization, digestion rate of starch in vitro, and metabolic response in rats. Am J Clin Nutr 1988;47:1010-
1016. 16. Jaffe WIG, Lette CLV. Heat-labile
growth inhibitory factors in beans (Phaseolus Vulgaris). J Nutr 1986;94:203-210. 17. Jenkins DJA, Thorne MJ, Wolever TMS, Jenkins AL, Rao AV, Thompson LU. The effect of starch-protein interaction in
18.
19.
20.
21.
22.
23.
24.
wheat on the glycemic response and rate of in vitro digestion. Am J Clin Nutr 1987;45:946-951. Jenkins DJA, Thorne MJ, Camelon K, Jenkins A, Rao AV, Taylor RH, Thompson LLJ, Kalmusky J, Reichert R, Francis T. Effect of processing on digestibility and the blood glucose response: a study of lentils. Am J Clin Nutr 1982;36:1093-1101. Layer P, Carlson GL, DiMagno EP. Partially purified white bean Amylase inhibitor reduces starch digestion in vitro and inactivates intraduodenal amylase in humans. Gastroenterology 1985;88:1895-1902. Wolever TMS, Jenkins DJA, Kalmusky J, Giordano C, Giudio S, Jenkins AL, Thompson LU, Wong GS, Jesse RG. Glycemic response to pasta: effect of surface area, degree of cooking, and protein enrichment. Diabetes Care 1986;9:401-403. Yoon JH, Thompson LU, Jenkins DJA. The effect of phytic acid on in vitro rate of starch digestibility and blood-glucose response. Am J Clin Nutr 1983;38:835-842. Wong S, O’Dea K. Importance of physical form rather than viscosity in determining the rate of starch hydrolysis in legumes. Am J Clin Nutr 1983;37:66-70. James AH. The nature of the gastric contents in man. in: James AH. The physiology of gastric digestion. London: Edward Arnold, 1957:1-24. Meyer JH, Kelly KA. Canine pancreatic responses to intestinally perfused proteins and protein digest. Am J Physiol 1976;231:G682-G691.
Received November 30, 1990. Accepted August 15, 1991. Address requests for reprints to: Henry C. Lin, M.D., CedarsSinai Medical Center, Suite 7511, 8700 Beverly Boulevard, Los Angeles, California 90048-1869. Supported by Veterans Administration Research Funds. Dr. Lin was a Veterans Affairs Associate Investigator when these studies were performed. The authors thank Don Jehn for expert technical assistance.
