GASTROENTEROLOGY
1990;98:191-196
Effect of Alcohol and Alcoholic Beverages on Meal-Stimulated Pancreatic Secretion in Humans FERENC HAJNAL, M. CARMEN JORGE E. VALENZUELA
FLORES, STEVEN
Gastroenterology Section, Los Angeles County-University Center, Los Angeles, California
Alcohol and alcoholic beverages may have different effects on pancreatic secretion and hormone release in humans. To test this hypothesis we studied the effects of an alcohol solution and a glucose solution and compared them with those of alcoholic beverages on postprandial pancreatic secretion and release of gastrin, trypsin, and cholecystokinin in 6 healthy nonalcoholic male volunteers. Pancreatic enzyme secretion was measured in duodenal aspirate, plasma trypsin, and gastrin by radioimmunoassay and cholecystokinin by bioassay. The meal plus glucose significantly stimulated pancreatic enzyme secretion, release of gastrin and cholecystokinin, and caused no changes in plasma trypsin. The alcohol solution and all beverages added to the meal caused similar increases in alcohol blood levels and significantly less pancreatic enzyme secretion compared with the meal plus glucose. Plasma trypsin levels remained unchanged. Compared with the meal plus glucose, wine and beer caused a significantly higher release of gastrin, and beer also released significantly more cholecystokinin. Inhibition of pancreatic enzyme secretion stimulated by a meal in nonalcoholics is a common effect of alcohol and alcoholic beverages despite some differences on release of gastrointestinal peptides. This effect may have some implications in the pathogenesis of alcoholic pancreatitis.
A
lthough chronic alcoholism is one of the most common factors related to pancreatitis (1) the effect of alcohol and alcoholic beverages on pancreatic secretion are poorly understood. Most studies of humans have investigated the effect of alcoholic solutions given intragastrically (2,3), intravenously (4-6), or into the intestine (7,8). Except for 1 recent study using beer (9), there is scarce information on the effect of
RADLEY,
and
of Southern California Medical
alcoholic beverages of common use such as beer, wine, and spirits on exocrine pancreatic secretion of humans. Furthermore, there are no data on the effect of alcoholic beverages on meal-stimulated pancreatic secretion. In this study we investigated the effects of an alcohol solution, beer, wine, and gin on pancreatic enzyme secretion stimulated by a meal and compared them with that of a glucose solution. Also, the effects of these beverages and alcohol solution on the mealstimulated release of gastrin, cholecystokinin (CCK), and trypsin were investigated.
Materials and Methods Subjects Six healthy nonalcoholic male volunteers, 19-52 yr old (mean 34 yr] (mean daily alcohol consumption less than 30 g) with no history of excessive drinking or drug abuse were studied. The design of the study was approved by the Los Angeles County-University of Southern California Medical Center Human Research Committee, and written informed consent was given by each subject.
Pancreatic
Secretory
Studies
The subjects were intubated with a modified doublelumen gastroduodenal rubber (Dreiling, Davol, Inc., Providence, R.I.) tube. The tube position was verified by fluoroscopy so that its metal end was located at the level of the ligament of Treitz. Two thin polyvinyl catheters (outer diameter, 1.5. mm] were attached to this tube. One of these catheters had a distal opening 5 cm beyond the pylorus, and the second ended 5 cm distal to the tip of the tube, in the upper jejunum. Through the first catheter a poorly absorbed
Abbreviations used in this paper: CCK, cholecystokinin; PEG, polyethylene glycol; RIA, radioimmunoassay. 0 1990 by the American Gastroenterological Association 0016-5065/90/$3.00
192
GASTROENTEROLOGY
HAJNAL ET AL.
marker, polyethylene glycol [PEG] 4000 (20 mg . ml-’ saline), was infused at 2 ml - min-‘. The second catheter served to reinfuse the duodenal aspirate distally. Fasting gastric content was aspirated manually and discarded. Duodenal content was collected by continuous suction at -5 mmHg by a mechanical pump interrupted by flushing of air every 5 min to ensure patency of the tube. Gastric and duodenal contents were aspirated for at least 30 min to allow clear separation of secretions. Preprandial gastric and duodenal samples were collected for 60 min. Then 265 ml of a liquid test meal (protein 8.8 g, fat 8.8 g, carbohydrate 34.3 g, and 250 kcal, Ensure, Ross Laboratories, Columbus, Ohio) was infused into the stomach, plus either the glucose solution, the alcohol solution, or alcoholic beverages. The gastric outlet was closed, and duodenal aspiration and reinfusion continued for 6 h. Duodenal aspirates were collected and separated in 15-min samples. Volume was measured to the closest 0.1 ml, an aliquot was saved on ice for enzyme determination, and the rest was reinfused into the intestine. Because gastric content was allowed to empty into the intestine, trypsin was chosen as the most suitable indicator of pancreatic enzyme secretion.
Assays Trypsin (10) and PEG 4000 (11) concentrations were determined in the duodenal aspirates usually within 24 h of collection. Output of trypsin in duodenal secretions was corrected by recovery of PEG 4000, which averaged 75%. Blood alcohol concentration was determined by an enzymatic method (121, and plasma trypsin-like immunoreactivity measurements were carried out by radioimmunoassay (RIA) (13) (Trypsik, CIS-kit Sorin Biomedica, Saluggia, Vercelli, Italy). The assay variability is 6%, and the sensitivity is about 1.5 ng . ml-’ (13). Cholecystokinin was measured by bioassay (14). Briefly, 2 ml of plasma samples were extracted with SEP-PAK cartridges and washed with 5 ml of methanol and 20 ml of water. The cartridges were washed again with water, and CCK was eluted in 1 ml of 80% ethanol and 0.2% trifluoroacetic acid. Extracts were incubated with 0.5 ml of a suspension of isolated dispersed rat pancreatic acini, and the amylase release in the incubation medium was determined. Dose-response curves were generated with sulphated CCK octapeptide (CCK-8). Cholecystokinin-like bioactivity, after comparison of the amylase releasing capability of the plasma samples with the dose-response curves, was expressed in pM CCK-8 equivalents. During the extraction procedure a fourfold concentration was achieved, and by this way the sensitivity of assay was as low as 0.5 pM CCK-8; intra-assay variability was 7%, and interassay variance, 12%. Radioimmunoassay of plasma gastrin was kindly performed by John H. Walsh (15). All measurements were performed in duplicate.
Study
Subjects were admitted to Center, and studies were conducted night fasting in 5 different days. On preprandial aspiration of gastric and
liquid meal plus 400 ml of a 12% glucose solution were infused into the stomach over a 20-min period. On day 2,400 ml of a 12% ethanol solution (vol/vol) was added to the meal, and on day 3,400 ml of white California wine; on day 4,850 ml of beer, and on day 5,400 ml water containing 120 ml of gin was infused with the meal. The osmolality, pH, and caloric content of each test meal are given in Table 1. Blood samples were drawn from an arm vein through an indwelling catheter at basal, 30, 60, 90, 120, 180, and 240 min following the meal. Blood samples for trypsin and hormones were collected into iced ethylenediaminetetraacetic acid tubes and immediately centrifuged to obtain plasma, which was stored at -20°C until assayed. Samples were usually assayed within 72 h of collection.
Statistical
Analysis
Mean (*SEM) trypsin outputs, plasma gastrin, CCK and trypsin levels and increments, and blood alcohol levels per each corresponding period and study days were calculated. Comparison between different periods of the same day and between similar periods on different study days were made by Student’s t-test for paired and unpaired samples, nonparametric Wilcoxon rank sum test, and analysis of variance, as appropriate. Also, the areas under the 90-min curve of plasma CCK following the meal were calculated, and comparison was carried out between study days. A p value to.05 was considered statistically significant.
Results Basal blood alcohol levels were undetectable and, as expected, increased after the administration of the alcohol solution and the different alcoholic beverages. Values were not significantly different between the study days (Figure 1).Alcohol blood levels returned to basal levels by the end of the fifth hour. Basal trypsin output in the duodenal aspirate between the different studies was similar, with a grand mean (*SEMI of 74.6 + 22.0 mg - h-‘. Trypsin output increased significantly (p < O.Ol] during the first hour after the meal plus the glucose solution to 378 f 77 mg . h-l. This ou t pu t represents approximately 80% of the enzyme secretion following submaximal doses of secretin and CCK in healthy volunteers in our Table 1. Osmolality, pH and Caloric Content of Each Test Meal
Ensure + glucose (12%) 400 ml Ensure + alcohol (12%) 400 ml
Design the Clinical Research at random after overday 1,at the end of the duodenal contents, the
Vol. 98, No. 1
Ensure + wine 400 ml Ensure + beer 850 ml Ensure + gin solution (30%) 400 ml
Caloric content
Osmolality (mosmoVkg)
pH
(kcal]
663 3708 1696 692
6.63 6.70 4.22 5.60
442 591 590 620
1809
6.32
590
ALCOHOL
January 1990
15 r
MEAL B 5 J
IO-
A--b
ETHANOL
a-0 0-o
WINE GIN
A-A
BEER
P 8 a 8
5-
id
0
30
180
120
60
240
300
MINUTES Figure 1. Mean (&EM) blood alcohol levels following the meal plus 400 ml 12% ethanol, 400 ml white wine, 650 ml beer, and 400 ml water containing 120 ml gin in 6 volunteers. In this and subsequent figures, bars represent SEM.
AND HUMAN
PANCREAS
193
laboratory [unpublished observations]. During the following hours, the trypsin activity decreased, but it remained significantly higher than basal (p < 0.02) until the fourth hour after the meal (Figure 2). When the ethanol solution was added to the test meal, there was a small and insignificant increase of the trypsin response to the meal compared with basal, and this was significantly less (p < 0.05) than that observed following the meal and glucose (Figure ZA). Similar effects were observed when beer (Figure 2BJ and gin (Figure 2C) were added to the meal. When wine was infused with the meal, there was an increase in trypsin output, but this was significantly less than the effect of glucose during the first postprandial hour (p < 0.05). However, no difference was observed during the second and third hours following the meal and wine compared with the same periods in day 1, when glucose was added to the meal. Cholecystokinin bioactivity in the plasma during fasting, 2.4 * 0.5 pM, increased after the meal and
400
d
=
p
1990;98:191-196
Effect of Alcohol and Alcoholic Beverages on Meal-Stimulated Pancreatic Secretion in Humans FERENC HAJNAL, M. CARMEN JORGE E. VALENZUELA
FLORES, STEVEN
Gastroenterology Section, Los Angeles County-University Center, Los Angeles, California
Alcohol and alcoholic beverages may have different effects on pancreatic secretion and hormone release in humans. To test this hypothesis we studied the effects of an alcohol solution and a glucose solution and compared them with those of alcoholic beverages on postprandial pancreatic secretion and release of gastrin, trypsin, and cholecystokinin in 6 healthy nonalcoholic male volunteers. Pancreatic enzyme secretion was measured in duodenal aspirate, plasma trypsin, and gastrin by radioimmunoassay and cholecystokinin by bioassay. The meal plus glucose significantly stimulated pancreatic enzyme secretion, release of gastrin and cholecystokinin, and caused no changes in plasma trypsin. The alcohol solution and all beverages added to the meal caused similar increases in alcohol blood levels and significantly less pancreatic enzyme secretion compared with the meal plus glucose. Plasma trypsin levels remained unchanged. Compared with the meal plus glucose, wine and beer caused a significantly higher release of gastrin, and beer also released significantly more cholecystokinin. Inhibition of pancreatic enzyme secretion stimulated by a meal in nonalcoholics is a common effect of alcohol and alcoholic beverages despite some differences on release of gastrointestinal peptides. This effect may have some implications in the pathogenesis of alcoholic pancreatitis.
A
lthough chronic alcoholism is one of the most common factors related to pancreatitis (1) the effect of alcohol and alcoholic beverages on pancreatic secretion are poorly understood. Most studies of humans have investigated the effect of alcoholic solutions given intragastrically (2,3), intravenously (4-6), or into the intestine (7,8). Except for 1 recent study using beer (9), there is scarce information on the effect of
RADLEY,
and
of Southern California Medical
alcoholic beverages of common use such as beer, wine, and spirits on exocrine pancreatic secretion of humans. Furthermore, there are no data on the effect of alcoholic beverages on meal-stimulated pancreatic secretion. In this study we investigated the effects of an alcohol solution, beer, wine, and gin on pancreatic enzyme secretion stimulated by a meal and compared them with that of a glucose solution. Also, the effects of these beverages and alcohol solution on the mealstimulated release of gastrin, cholecystokinin (CCK), and trypsin were investigated.
Materials and Methods Subjects Six healthy nonalcoholic male volunteers, 19-52 yr old (mean 34 yr] (mean daily alcohol consumption less than 30 g) with no history of excessive drinking or drug abuse were studied. The design of the study was approved by the Los Angeles County-University of Southern California Medical Center Human Research Committee, and written informed consent was given by each subject.
Pancreatic
Secretory
Studies
The subjects were intubated with a modified doublelumen gastroduodenal rubber (Dreiling, Davol, Inc., Providence, R.I.) tube. The tube position was verified by fluoroscopy so that its metal end was located at the level of the ligament of Treitz. Two thin polyvinyl catheters (outer diameter, 1.5. mm] were attached to this tube. One of these catheters had a distal opening 5 cm beyond the pylorus, and the second ended 5 cm distal to the tip of the tube, in the upper jejunum. Through the first catheter a poorly absorbed
Abbreviations used in this paper: CCK, cholecystokinin; PEG, polyethylene glycol; RIA, radioimmunoassay. 0 1990 by the American Gastroenterological Association 0016-5065/90/$3.00
192
GASTROENTEROLOGY
HAJNAL ET AL.
marker, polyethylene glycol [PEG] 4000 (20 mg . ml-’ saline), was infused at 2 ml - min-‘. The second catheter served to reinfuse the duodenal aspirate distally. Fasting gastric content was aspirated manually and discarded. Duodenal content was collected by continuous suction at -5 mmHg by a mechanical pump interrupted by flushing of air every 5 min to ensure patency of the tube. Gastric and duodenal contents were aspirated for at least 30 min to allow clear separation of secretions. Preprandial gastric and duodenal samples were collected for 60 min. Then 265 ml of a liquid test meal (protein 8.8 g, fat 8.8 g, carbohydrate 34.3 g, and 250 kcal, Ensure, Ross Laboratories, Columbus, Ohio) was infused into the stomach, plus either the glucose solution, the alcohol solution, or alcoholic beverages. The gastric outlet was closed, and duodenal aspiration and reinfusion continued for 6 h. Duodenal aspirates were collected and separated in 15-min samples. Volume was measured to the closest 0.1 ml, an aliquot was saved on ice for enzyme determination, and the rest was reinfused into the intestine. Because gastric content was allowed to empty into the intestine, trypsin was chosen as the most suitable indicator of pancreatic enzyme secretion.
Assays Trypsin (10) and PEG 4000 (11) concentrations were determined in the duodenal aspirates usually within 24 h of collection. Output of trypsin in duodenal secretions was corrected by recovery of PEG 4000, which averaged 75%. Blood alcohol concentration was determined by an enzymatic method (121, and plasma trypsin-like immunoreactivity measurements were carried out by radioimmunoassay (RIA) (13) (Trypsik, CIS-kit Sorin Biomedica, Saluggia, Vercelli, Italy). The assay variability is 6%, and the sensitivity is about 1.5 ng . ml-’ (13). Cholecystokinin was measured by bioassay (14). Briefly, 2 ml of plasma samples were extracted with SEP-PAK cartridges and washed with 5 ml of methanol and 20 ml of water. The cartridges were washed again with water, and CCK was eluted in 1 ml of 80% ethanol and 0.2% trifluoroacetic acid. Extracts were incubated with 0.5 ml of a suspension of isolated dispersed rat pancreatic acini, and the amylase release in the incubation medium was determined. Dose-response curves were generated with sulphated CCK octapeptide (CCK-8). Cholecystokinin-like bioactivity, after comparison of the amylase releasing capability of the plasma samples with the dose-response curves, was expressed in pM CCK-8 equivalents. During the extraction procedure a fourfold concentration was achieved, and by this way the sensitivity of assay was as low as 0.5 pM CCK-8; intra-assay variability was 7%, and interassay variance, 12%. Radioimmunoassay of plasma gastrin was kindly performed by John H. Walsh (15). All measurements were performed in duplicate.
Study
Subjects were admitted to Center, and studies were conducted night fasting in 5 different days. On preprandial aspiration of gastric and
liquid meal plus 400 ml of a 12% glucose solution were infused into the stomach over a 20-min period. On day 2,400 ml of a 12% ethanol solution (vol/vol) was added to the meal, and on day 3,400 ml of white California wine; on day 4,850 ml of beer, and on day 5,400 ml water containing 120 ml of gin was infused with the meal. The osmolality, pH, and caloric content of each test meal are given in Table 1. Blood samples were drawn from an arm vein through an indwelling catheter at basal, 30, 60, 90, 120, 180, and 240 min following the meal. Blood samples for trypsin and hormones were collected into iced ethylenediaminetetraacetic acid tubes and immediately centrifuged to obtain plasma, which was stored at -20°C until assayed. Samples were usually assayed within 72 h of collection.
Statistical
Analysis
Mean (*SEM) trypsin outputs, plasma gastrin, CCK and trypsin levels and increments, and blood alcohol levels per each corresponding period and study days were calculated. Comparison between different periods of the same day and between similar periods on different study days were made by Student’s t-test for paired and unpaired samples, nonparametric Wilcoxon rank sum test, and analysis of variance, as appropriate. Also, the areas under the 90-min curve of plasma CCK following the meal were calculated, and comparison was carried out between study days. A p value to.05 was considered statistically significant.
Results Basal blood alcohol levels were undetectable and, as expected, increased after the administration of the alcohol solution and the different alcoholic beverages. Values were not significantly different between the study days (Figure 1).Alcohol blood levels returned to basal levels by the end of the fifth hour. Basal trypsin output in the duodenal aspirate between the different studies was similar, with a grand mean (*SEMI of 74.6 + 22.0 mg - h-‘. Trypsin output increased significantly (p < O.Ol] during the first hour after the meal plus the glucose solution to 378 f 77 mg . h-l. This ou t pu t represents approximately 80% of the enzyme secretion following submaximal doses of secretin and CCK in healthy volunteers in our Table 1. Osmolality, pH and Caloric Content of Each Test Meal
Ensure + glucose (12%) 400 ml Ensure + alcohol (12%) 400 ml
Design the Clinical Research at random after overday 1,at the end of the duodenal contents, the
Vol. 98, No. 1
Ensure + wine 400 ml Ensure + beer 850 ml Ensure + gin solution (30%) 400 ml
Caloric content
Osmolality (mosmoVkg)
pH
(kcal]
663 3708 1696 692
6.63 6.70 4.22 5.60
442 591 590 620
1809
6.32
590
ALCOHOL
January 1990
15 r
MEAL B 5 J
IO-
A--b
ETHANOL
a-0 0-o
WINE GIN
A-A
BEER
P 8 a 8
5-
id
0
30
180
120
60
240
300
MINUTES Figure 1. Mean (&EM) blood alcohol levels following the meal plus 400 ml 12% ethanol, 400 ml white wine, 650 ml beer, and 400 ml water containing 120 ml gin in 6 volunteers. In this and subsequent figures, bars represent SEM.
AND HUMAN
PANCREAS
193
laboratory [unpublished observations]. During the following hours, the trypsin activity decreased, but it remained significantly higher than basal (p < 0.02) until the fourth hour after the meal (Figure 2). When the ethanol solution was added to the test meal, there was a small and insignificant increase of the trypsin response to the meal compared with basal, and this was significantly less (p < 0.05) than that observed following the meal and glucose (Figure ZA). Similar effects were observed when beer (Figure 2BJ and gin (Figure 2C) were added to the meal. When wine was infused with the meal, there was an increase in trypsin output, but this was significantly less than the effect of glucose during the first postprandial hour (p < 0.05). However, no difference was observed during the second and third hours following the meal and wine compared with the same periods in day 1, when glucose was added to the meal. Cholecystokinin bioactivity in the plasma during fasting, 2.4 * 0.5 pM, increased after the meal and
400
d
=
p
