GASTROENTEROLOGY 73:715-718, 1977 Vol. 73, No.

4, Part 1
Copyright © 1977 by the American Gastroenterological Association Printed in U.S A.

SYMPTOMATIC TREATMENT OF DIARRHEA WITH BISMUTH

SUBSALICYLATE AMONG STUDENTS ATTENDING A
MEXICAN UNIVERSITY
HERBERT L. DuPoNT, M.D., PEGGY SuLLIVAN, LARRY K . PICKERING, M.D.,
GERRY HAYNES, R.N., M.P.H., AND P. B. AcKERMAN, M.S.
The University of Texas Medical School at Houston and The Norwich Pharmacal Company , Division of
Morton-Norwich Products, Inc ., Houston, Texas and Norwich , New York

Students attending a Mexican university who developed diarrhea were randomly

treated with bismuth subsalicylate or a placebo. One hundred and eleven were given 30
ml each 112 hr until eight doses (total dose of active drug 4.2 g) were given and 58
students received twice this dose (8.2 g of active drug) over the 3112-hr treatment period.
The number of unformed stools was significantly decreased in both bismuth subsalicy-
late treatment groups compared to the placebo controls for the period 4 to 24 hr after
therapy. A reduction in diarrhea was additionally noted for the duration of the 48-hr
surveillance period for the students receiving the higher dose of active drug. Subjective
relief within 24 hr of therapy of the symptoms of diarrhea, nausea, and abdominal pain
or cramps was reported with a significantly increased frequency in the bismuth subsali-
cylate group. The most pronounced effect of the treatment occurred in the United States
students with diarrhea who had recently arrived in Mexico. This appeared to be related
to the favorable effect of bismuth subsalicylate on the course of toxigenic Escherichia
coli infection. Students with shigellosis did not experience a prolonged illness in either
treatment group.

Persons from the United States often develop diar-
rhea when traveling to Latin America. Recently, toxi-
genic Escherichia coli and shigella strains have been
shown to be the most important causative agents of
diarrhea of United States travelers to Mexico. 1-a How-
ever, in a percentage of travelers, Giardia lamblia,
Entamoeba histolytica, salmonella strains, and other
agents can be identified as etiological agents.
Despite the fact that there are few published studies
which offer convincing evidence that commonly used
antidiarrheal compounds are helpful in the sympto-
matic treatment of infectious diarrhea, the sales in the
United States of both prescription and over-the-counter
products used in the treatment of diarrhea exceeds 50
million dollars per year. 4 One of the most widely used
over-the-counter compounds is Pepto-Bismol. The active
Received January 31, 1977. Accepted March 28, 1977.
Address requests for reprints to: Dr. H . L. DuPont, 6400 West
Cullen Street, Houston, Texas 77030.
This study was supported by a grant from the Norwich Pharmacal
Company, Division of Morton-Norwich Products, Inc., and Grant AI
12699-01 from the National Institutes of Health.
The authors acknowledge the enthusiasm and support of the
faculty, staff and students at Universidad de las Americas. They
thank Drs. M. Rendon, H. Kirkpatrick, E. Simmen, E. Rice-Wray,
P. Carrillo, and C. Moctezuma. Members of the field team included:
J. Wittenberg, W. Edwards, P. Hamilton, C. Breece, M. Ceresko, B.
Skeel, C. Ricart , C. Cooper, K. Valentine. Drs. T. Magenis, E.
Buchman and R. Keenan helped to design the study, Dr. B. Hsi
offered statistical help, and K. Boyle helped in the preparation of the
manuscript.
715
ingredient of this preparation is bismuth subsalicylate
(bs). Whereas bismuth-containing drugs have been em-
ployed as antidiarrheal agents for many years, recent
attention has been given salicylates, which may de-
crease diarrhea through inhibition of intestinal secre-
tion. 5 The present study was designed to determine the
value of bs in the symptomatic treatment of students
who acquire diarrhea while attending school in Mexico.
Materials and Methods
A clinic for the study of diarrhea was established at U niver-
sidad de las Americas, Cholula, Puebla, Mexico, from June
1975 to April 1976. Students were enrolled in larger studies
looking at enteric pathogens and stools were collected in many
cases before illness developed . Students with three or more
unformed stools in the preceding 24 hr associated with one or
more symptoms of enteric infection (fever, nausea, vomiting,
or abdominal cramps) were interviewed, a consent form was
signed, and they were enrolled in the study. Those taking
other antidiarrheal compoundR or antimicrobials were ex-
cluded from the study.
Students were treated with either a suspension of bs (17.5
mg per ml) or a placebo formulation . The suspension was a
blind modification of the commercial product Pepto-Bismol
containing bs, flavoring, and coloring in a suspending vehicle
of magnesium aluminum silicate and methylcellulose. The
two suspensions were prepared to be as identical as possible in
color, flavor, and viscosity . Packaged in amber glass with
identical labeling, the 8-ounce bottles were ordered and num-
bered in accordance with a blind randomized code generated
by computer. As students were entered in the study the bottles
were dispensed sequentially, providing a double blind assign-
716 DUPONT ET AL. Vol. 73, No.4, Part 1

ment as neither the investigator nor the student knew the Results
identity of the formulation .
Students were studied in one of the two following schedules: Tables 1 and 2 show the number of unformed stools
(1) 111 students with three or more unformed stools in the passed in the two treatment groups during the 48-hr
preceding 24 hr were given 30 ml every 112 hr (525 mg ofbs per surveillance period. One hundred and eleven students
30 ml) until eight doses or one bottle (total dose 4.2 g of bs) with three or more unformed stools in the preceding 24
were taken, and (2) 58 students with five or more unformed hr were given 240 ml of bs (containing 4.2 g) or placebo
stools in the preceding 24 hr received 60 ml (1 g of bs) for eight
doses or two bottles (total dose 8.4 g of bs). In both groups
during the 3112 hr treatment period. They were divided
treatment was continued for 3112 hr. Students were given wide by their prior geographic location and according to drug
mouth Styrofoam cartons in which to collect all stools passed taken. There were no differences in the number of stools
during the subsequent 48 hr. They were seen at specified passed before entry into the study between the various
intervals-4, 8, 12, 24, and 48 hr-to assure optimal stool student groups separated by geographical background
collection and to determine subjective relief of diarrhea, nau- receiving bs or placebo, or between the total groups
sea, vomiting, or abdominal cramps, through verbal question- receiving bs or placebo, thus making the paired treat-
ing. ment groups comparable. There were fewer diarrheal
Stools were processed for enteric bacterial or protozoal path- stools passed among each student group receiving bs
ogens, and serum antibody rises to the heat-labile enterotoxin compared to placebo controls from 4 to 24 hr after initia-
of E. coli were sought as described recently." To be considered
an etiological agent, a pathogen had to be present and detected tion of therapy. The differences in numbers of stools
within 24 hr of disease onset. Pathogens were considered passed after therapy were significant for United States
unimportant if found on a pre-illness base-line stool. Stools summer students from 4 to 48 hr (P < 0.025), for Latin
received in the laboratory were examined for form and consist- American students in all time intervals (P < 0.025 toP
ency, and the entire volume was weighed. Stool water content < 0.05), and for all students combined for the period of
was measured in each sample by placing approximately 2 to 5 time 4 to 24 hr after therapy (P < 0.05).
g of stool in a centrifuge tube, determining the wet weight, Table 2 shows similar data for a second group of 58
and then drying for 1 hr at ll0°C in a Lab-line Multi-temp heat students with five or more unformed stools in the pre-
block (Lab-Line Plaza, Melrose Park, Ill.) followed by weigh- ceding 24 hr given twice the dose of bs over a similar
ing the dried stool. Percentage of stool water was determined 3112-hr treatment period. A statistically significant de-
by dividing the wet weight minus dry weight by the wet
weight and multiplying by 100. Stool water content of a nor- crease occurred in the number of diarrheal stools in
mal stool collected before onset of diarrhea or after recovery United States summer students receiving the active
was determined for a baseline in as many students as possible. drug for the following time intervals post-therapy: 0 to
Statistical evaluation of data obtained was performed by the 48 hr (P < 0.01), 4 to 24 hr (P < 0.025), and 4 to 48 hr (P
Student's t-test (tables 1 to 3), and x" analysis (fig. 1) . < 0.01). The number of unformed stools in the 4- to 24-
TABLE 1. Bismuth subsalicylate and placebo therapy in diarrhea in students from differing geographical locations"
No. of unformed stools•
Student status No. studied Preparation (hr post-treatment)
0-4 4-24 24-48
United States
Summer 27 Bismuth subsalicylate 47 (1.7) 62 (2.3) 56 (2.1)
16 Placebo 25 (1.6) 70 (4.4) 77 (4.8)
Full-time 23 Bismuth subsalicylate 29 (1.3) 46 (2.0) 59 (2.6)
28 Placebo 27 (1.0) 64 (2.3) 57 (2.0)
Latin American 6 Bismuth subsalicylate 0 1 (0.2) 0
11 Placebo 20 (1.8) 31 (2.8) 13 (1.2)
Total 56 Bismuth subsalicylate 76 (1.4) 109 (1.9) 115 (2.1)
55 Placebo 72 (1.3) 165 (3 .0) 147 (2. 7)
" Dose, 240 ml (8 ounces) over 3112 hr.
b Total number of stools (mean per student).
TABLE 2. Bismuth subsalicylate and placebo therapy of diarrhea in students from differing geographical locations"
No. of unformed stools•
Student status No. studied Preparation (hr post-treatment)
0-4 4-24 24-48
United States
Summer 19 Bismuth subsalicylate 25 (1.3) 25 (1.3) 8 (0.4)
17 Placebo 22 (1.3) 44 (2.6) 29 (1. 7)
Full-time 6 Bismuth subsalicylate 4 (0.7) 10 (1. 7) 1 (0.2)
11 Placebo 13 (1.2) 29 (2 .6) 12 (1.1)
Total 27'' Bismuth subsalicylate 32 (1.2) 39 (1.4) 10 (0.4)
31" Placebo 40 (1.3) 77 (2.5) 44 (1.4)
" Dose, 480 ml (16 ounces) over 3 1/ 2 hr.
" Total number of stools (mean per student).
,. Five Latin American students were included in the tota l analysis.
October 1977 DIARRHEA OF TRAVELERS TO MEXICO
and 4- to 48-hr periods after therapy was significantly
decreased for the total students receiving two bottles of
bs (P < 0.05 andP < 0.01, respectively for the two time
periods).
Table 3 shows the intensity of diarrhea among treated
students according to the probable etiology of their diar-
rhea. In this analysis only students having stools proc-
essed within 24 hr of disease onset were included. With
the exception of 8 students with shigellosis, all students
in this analysis were given the lower dose of medication
717
Figure 1 shows the subjective relief of symptoms
within 24 hr after beginning the medication, as verbally
reported by the students when questioned. The 32 stu-
dents with shigellosis were not considered in this analy-
sis owing to failure of subjective response to therapy
among them. Although there was a slightly increased
frequency of relief of symptoms among the students
receiving two bottles ofbs compared to 1 (4.2 versus 8.4
a Bismuth Subsalicylate

(8 ounces). Twelve students in this portion of the study

100 29/
0 Placebo
had mild illness (one to two unformed stools) 24 hr ~1
90- 45/
before initiation of therapy, whereas the remaining 85 !J2
students had three or more stools during this time pe- 80-
riod. There was an insufficient number of students hav- 70-
ing a protozoal pathogen to analyze as a separate group. 60-
Twenty-eight students had toxigenic E. coli isolated 50-
from stool or had a 4-fold rise in serum antibodies toE. c i
coli heat-labile enterotoxin. These students with toxi- 40j
"'c.;
u
30~
genic E. coli infection showed a favorable response to 8 a.
ounces of bs (4.2 g) therapy given for 3112 hr. A signifi- 20-
cantly (P < 0.025) fewer number of unformed stools 10-
were passed between 4 and 24 hr after initiation of
therapy and between 4 and 48 hr, by students who Diarrhea Nausea Vomiting Abdominal
received the active drug when compared to those who Cramps
were given placebo. FIG. 1. Subjective relief of symptoms within 24 hr of initiation of
Thirty-two students with diarrhea had a shigella therapy, from diarrhea, nausea, and abdominal cramps occurred
strain isolated from stool. Twenty-four students with with a significantly increased frequency in bismuth subsalicylate-
shigellosis received 240 ml of medication and 8 students treated students (shigella cases are excluded) .
were given 480 ml. No statistically significant reduction
in the occurrence of diarrhea between 4 and 48 hr was 8-
seen in either group, although the students treated with 7 - -
a Bismuth Subsalicylate

bs averaged fewer stools numerically. None of the bs- 0 Placebo

treated students with shigellosis had a prolonged ill- c"' 6

ness. "'
"0
::l
5 f- r-
In 37 students with diarrhea where etiological agents u; 4
:;;
were sought, an enteric pathogen was not isolated from .0
E 3 f- r- r-
stool and a serum antibody titer rise toE. coli entero- ::l

toxin was not detected. These students are shown as

z 2!- - n
having "nonspecific" diarrhea in table 3. Altough there
appeared to be a reduction in number of diarrheal stools
between 4 and 24 hr (mean 1.2 stools per student versus
1-

0 2 3 4
t-1
LJ
5
n
6 7
n
8 9 ~10

2.5 stools per student) the differences were not signifi-
cant. Students were not included if their stools were not
studied within 24 hr of disease onset because we were
unable to state with confidence the probable etiology of
their diarrhea.
Number Unformed Stools
FIG. 2. Severity of diarrhea as determined by number of unformed
stools after therapy differed in the two treatment groups. The bis-
muth subsalicylate-treated students tended to experience milder
illness (they received 16 ounces over 3 1/ 2 hr) .

TABLE 3. Bismuth subsalicylate and placebo therapy of diarrhea in students according to etiology"
No. of unformed stools'
Etiology No. studied Preparation (hr post-treatment)
0-4 4-24 24-48

Toxigenic Escherichia coli 16 Bismuth subsalicylate 17 (1.1) 19 (1.2) 35 (2.2)

12 Placebo 19 (1.6) 42 (3 .5) 52 (4.3)
Shigella 15 Bismuth subsalicylate 30 (2.0) 43 (2.9) 41 (2.7)
17 Placebo 26 (1.5) 62 (3.6) 64 (3 .8)
"Non specific" (no pathogens 16 Bismuth subsalicylate 14 (0.9) 19 (1.2) 23 (1.4)
isolated)
21 Placebo 22 (1.0) 52 (2.5) 22 (1.0)
- - - - - - - - - - - -·- ------·-- - - - -- -- - - - - - - - -- - -- - - ·
" All but 8 students with shigellosis received 240 mi.
b Total number of stools (mean per student).
I
718 DUPONT ET AL.
g), they did not statistically differ and so data were
combined. Subjective relief of diarrhea, nausea, vomit-
ing, and abdominal pain or cramping within 24 hr of
therapy initiation were more commonly reported by
drug-treated when compared to placebo-treated stu-
dents. Relief of three of the symptoms was significant
among the bs-treated students: diarrhea (P < 0.025),
nausea (P < 0.05), and abdominal cramps (P < 0.025).
The differences in relief of vomiting were not statisti-
cally different in the two treatment groups, which ap-
peared to relate to the small number of students with
that symptom. There was a general acceptability of both
preparations by all students, even those with vomiting.
No student failed to take the prescribed therapy because
of objection to it.
Figure 2 graphically illustrates the beneficial effect of
480 ml of drug in students with severe diarrhea (five or
more stools per 24 hr); bs-treated students tended to
have milder illness as measured by number of unformed
stools and the placebo group experienced more intense
diarrhea.
Before the onset of diarrhea the mean stool water was
69% for the bs group and 68% for the placebo group.
When diarrhea occurred, the concentration of stool wa-
ter rose to 85 and 87%, respectively, for the two groups,
which were significantly (P < 0.05) greater than the
base-line value. During the 48-hr period, there were no
differences in stool water content or daily stool weight
in the two treatment groups.
Discussion
This study indicates that a 3112-hr course of therapy
with either a 4.2 g (240 ml) or 8.2 g (480 ml) dosage
regimen of bs can effectively decrease the number of
unformed stools passed as well as the subjective com-
plaints of diarrhea, nausea, and abdominal cramping
associated with enteric infection acquired in Mexico.
The favorable effect was delayed in onset and appeared
to occur generally after the 31/2-hr treatment period.
The number of unformed stools was decreased through-
out the 48-hr surveillance period even though the drug
was administered for only 31 /2 hr. There was no differ-
ence in water content or total weight of stools passed in
the drug compared to placebo student groups, making it
difficult to postulate a mechanism for improvement in
symptoms. The present study failed to show any useful-
ness of stool-water determination in evaluating drug
efficacy. The student group showing the most favorable
response to bs was the newly arrived United States
summer students, which probably relates to a favorable
effect of the agent on toxigenic E. coli diarrhea, the
most common cause of diarrhea in this group. 1- 3 This is
further supported by our evidence that students having
diarrhea wherein a toxigenic E. coli was found in stool
or those with a rise in antibody to E. coli enterotoxin
had a statistically significant decrease in unformed
Vol. 73,No.4,Partl
stools in the 4- to 48-hr period after taking one bottle of
drug. Ericsson et al. 6 in our laboratory have shown that
bs blocks the diarrheogenic effect of E. coli and Vibrio
cholerae enterotoxin in an experimental animal model,
further supporting the idea that the compound has spe-
cific activity against enterotoxigenic bacteria or the
heat-labile toxin they produce.
Previous studies have shown that the commercial
preparation of bs, Pepto-Bismol, protected the gastric
mucosa of rats from the action of aspirin, ethyl alcohol,
or cold plus restraint stress. i They postulated that a
protective mucosal barrier was produced by the drug.
The same group of investigators were able to prevent an
enhanced charcoal-meal transport induced by castor oil
in mice and rats with bs. 8 In this animal study, fecal
output and stool frequency were also decreased. Al-
though bismuth may be the active antidiarrheal agent,
there is growing evidence that salicylates have antidi-
arrheal properties, 5 perhaps through their effect on
prostaglandin metabolism.
Regardless of etiology, bs was effective in reducing
diarrhea in our study. Of particular interest was the
fact that no enhancement of illness was found in 32
students with shigellosis when they were treated, an
occasional problem with drugs which inhibit intestinal
motility.9 This is an important point since these drugs
are so widely prescribed and a significant proportion of
travelers' diarrhea is caused by shigella strains. 2• 3 Fur-
ther studies are contemplated to determine the mecha-
nism of action of bs, and its possible usefulness as a
prophylactic agent in the prevention of travelers' diar-
rhea.
REFERENCES
1. Gorbach SL, Kean BH, Evans DG, et al: Travelers' diarrhea and
toxigenic Escherichia coli. New Engl J Med 292:933-936, 1975
2. Merson M, Morris GK, Sack DA, et al: Travelers' diarrhea in
Mexico: a prospective study of physicians and family members
attending a congress. N Engl J Med 294:1299-1305, 1976
3. DuPont HL, Olarte J, Evans DG, et al: Comparative susceptibil-
ity of Latin American and United States students to enteric
pathogens. N Engl J Med 295:1520-1521, 1976
4. US Department of Commerce Bureau of the Census Current
Industrial Report, Series MA-28G (74)-1. US Government Print-
ing Office, Washington DC, 1974
5. Mennie AT, Dalley VM, Dinneen LC, et al: Treatment of radia-
tion-induced gastrointestinal distress with acetylsalicylate.
Lancet 2:942-943, 1975
6. Ericsson CD, Evans DG, DuPont HL, et al: Inhibition of crude E.
coli and cholera toxin activity by bismuth subsalicylate. J Infect
Dis (in press), 1977
7. Goldenberg MM, Honkomp LJ, Burrous SE, et al: Protective
effect of Pepto-Bismol liquid on the gastric mucosa of rats.
Gastroenterology 69:636-640, 1975
8. Goldenberg MM, Honkomp LJ, Castellion AW: The antidi-
arrheal action of bismuth subsalicylate in the mouse and the rat.
Dig Dis 20:955-960, 1975
9. DuPont HL, Hornick RB : Adverse effect of Lomotil therapy in
shigellosis. JAMA 226:1525-1528, 1973