Content uploaded by Elliot M Berry
Author content
All content in this area was uploaded by Elliot M Berry on Apr 08, 2015
Content may be subject to copyright.
W heat Gra ss Juice in the Treatment of Active Distal Ulcerative Colitis
A R andomized D ouble-blind Placebo-controlled Trial
E. Ben -Arye, E. Goldin, D. Wengrower, A. Stamper, R. Kohn & E. Berry
Dept. of Family Medicine, The Bruce Rappaport Faculty of Medicine, The Technion, Israel Institute
of Technology , Haifa, Israel; Klalit Health Services, Haifa and Wester n Gallill e District;
Gastroenterolog y Unit, Hadassah Hospital. Jerusalem , Israel ; Dept. of Pharmacology , the Hebrew
University, Hadassah School of Medicine, Jerusalem , Israel; Dept. of Metabolism and Nutrition, the
Hebrew University, Hadassah Hospital, Jerusalem, Israel
Ben-Arye E, Goldin E, Wengrower D, Stamper A, Kohn R, Berry E. Wheat grass juice in the treatment of
active distal ulcerativ e colitis : a randomized double-blin d placebo-controlle d trial. Scand J Gastroentero l
2002;37:444 –449.
Background: The use of wheat grass (Triticum aestivum) juice for treatment of various gastrointestina l
and other conditions had been suggeste d by its proponent s for more than 30 years, but was never
clinically assessed in a controlle d trial. A preliminary unpublished pilot study suggeste d ef cacy of wheat
grass juice in the treatment of ulcerative colitis (UC). Methods: A randomized, double-blind , placebo-
controlled study. One gastroenterolog y unit in a tertiary hospital and three study coordinatin g centers in
three major cities in Israel. Twenty-three patients diagnosed clinically and sigmoidoscopicall y with active
distal UC were randomly allocated to receive either 100 cc of wheat grass juice, or a matching placebo,
daily for 1 month. Ef cacy of treatment was assessed by a 4-fol d disease activity index that included
rectal bleeding and number of bowel movements as determined from patient diary records, a
sigm oidoscopi c evaluation, and global assessment by a physician. Results. Twenty-one patients
com pleted the study, and full information was available on 19 of them. Treatment with wheat grass juice
was associated with signi cant reductions in the overall disease activity index (P = 0.031) and in the
severity of rectal bleedin g (P = 0.025). No serious side effects were found. Fresh extract of wheat grass
dem onstrate d a prominent tracing in cyclic voltammetry methodology, presumably correspondin g to four
groups of compounds tha t exhibit anti-oxidativ e properties . Conclusion. Wheat grass juice appeare d
effective and safe as a single or adjuvant treatment of active distal UC.
K ey words: Clinical trial; complementar y alternativ e medicine (CAM); randomized; ulcerativ e colitis;
wheat grass juice
Eran Ben-Arye, M.D., Moshav Bet-Shearim, Israel 30046 (fax.
‡
97 2 463 90127, e-mail. eranben @
netvision.net.il)
W
heat grass juice is an extrac t squeezed from mature
sprouts of wheat seeds (Triticum aes tivum). The
use of wheat grass juice for therapeutic purposes
was developed and popularized by Dr. Ann Wigmore (1).
Although proponents of wheat grass juice have re commended
it for three decades as a treatment for various a ilments,
including chroni c in ammatory conditions and malignancies,
to date no clinical trials exist. The juice is believed to possess
therapeutic qualities only when fresh and is therefore con-
sumed immediately after extraction, on an empty stomach.
The therapeutic qua lities of wheat grass juice have been
attributed to its rich nutritional content—chlorophyll, vita-
mins (A, C and E), bio avenoids, minerals (iron, calcium and
magnesium), and 17 amino acids, 8 of which are essential (2).
Several studies have a ssessed an anti-mutagenic activity of
wheat grass juice. In 1978, Lai et al. demonstrated an anti-
mutagenic activity of wheat grass extract in the Ames test
system and attributed it to chlorophyll (3). In 1992, Peryt
showed its anti-mutagenic activity agains t cyclophosphamide
and ethidium bromide, which was related to avonoids,
particularly apigenin (4).
Clinic al aspects of ulcerative colitis (UC) have been
studied thoroughly , and there have been recent advances in
drug therapy (5). Yet many patients, particularly in younger
age groups, suffer from symptoms that are dif cult to treat:
the emotional distress of feeling c ontrolled by the disease
rather than controlling it (as they constantly seek the nearest
rest-room) ; the experience of blood in the stool and constant
concern about neoplastic changes; the side effects of second-
line drugs; and the physic al and psychological distress of
multiple endoscopic procedures (6).
The use of wheat grass juice in the treatment of UC was
brought to our attention by several patients with UC who
attributed their improvement to regular use of the extract. In a
pilot study, we gave 100 cc of wheat grass juice for 14 days to
10 previously diagnosed UC patients during relapse. Eight
ORIGINAL ARTICLE
Ó 2002 Taylor & Francis
patients described clinical improvement, one had no change,
and one got worse. This study was never published. The
present randomized, double-blind, placebo-controlled trial
was designed to examine the e ffects of wheat grass juice in
patients with active distal UC.
Me thods
Patients
Patients were recruited from a variety of sources: media
publicity, an in ammatory bowel disease (IBD) support
group, and referrals from gastroenterologists and general
practitioners (Fig. 1). Of the 614 individuals who contacted
the research team, 210 presented with clinical histories
sugge stive of IBD (164 compatible with UC and 46 with
Crohn disease). Active symptoms of UC were pres ent in 124
patients. However, only 24 ful lled the clinical inclusion
criteria (Table I) as well as were willing to undergo two
sigmidoscopies, resided within reasonable travel distance,
and were able to enlist the full collaboration of their physi-
cians. These 24 patients (1 5 men and 9 women), with a mean
age of 35 and a mean duration of UC of 6.8 years, were
randomized to receive either true whe at grass juice or a
placebo of similar color and amount. There were no signi -
cant differences between the two groups in the baseline
patient characteristics, clinical and sigmoidoscopic assess-
ment and current treatment (Table II).
The protocol was approved by the Hadassah Hospital ethics
committee in Jerusalem and by the Israeli Ministry of Health
in accordance with the Declaration of Helsinki II. All patients
signed an informed consent. Patients with clinical symptoms
of UC underwent a baseline endoscopic assessment upon
entry.
Assessment of patients
Although there is a correlation between sigm oidoscopic
evaluation and clinically active UC, one cannot predict the
severity of the symptoms based on the sigmoidoscopic
examination alone. Therefore, disease activity assessments
in UC require a combination of subjective and objective
parameters. Disease activity was assessed in three ways:
symptom diary, sigmoidoscopy, and subjective improvement
scale (7).
The symptom diary consisted of eight symptoms which
patients graded daily on a 0–3 scale (no activity, mild,
moderate, or severe, respectively). The eight symptoms were
rectal bleeding, stool frequency, urgency, pain (rectal or
abdominal), distention, mucous, general well-being, and
appetite . Baseline diary information was recorded for 3–7
days prior to the beginning of treatment. Symptomatic
changes were assessed by comparing the average scores of
the rst week of diary entry to the average scores in the last
week of diary entry for each symptom.
Sigm oidoscopies were performed on each participant
within a week prior to starting treatment a nd within 3 days
following the termination of treatment. Sigmoidoscopic
results were rated on a 0–3 scale as no activity, mild,
moderate or se vere, respectively. Pretreatment and posttreat-
ment scores were compared for each participant. All
sigmoidoscopies were performed, a ssessed and rated by
gastroenterologists from the Gastroenterology Unit in
Hadassah Hospital in Jerusalem who were blinded a s to
whether the patients were treated with wheat grass or placebo.
The subjective global assessment consisted of both the
phys ician’s and participant’s estimation of change in disease
activity. It was performed upon treatment termination
separately and graded on a ¡3 to ‡3 scale.
Fig. 1. Flow chart describing progress of patients through the trial.
Table I. Inclusion criteria
Age
>
18 years
Sigmoidoscopi c nding of active Ulcerative Colitis that involve s the
left colon
Clinical activity comparable with Ulcerative Colitis
No change in drug treatment (type and dosage) in the month prior to
entry
Lack of serious systemic involvement –fever
>
38
°
C, erythem a
nodosum, arthritis
Blood hemoglobi n
>
11 g%
Negative stool culture and test for ova and parasites
Table II. Patient characteristic s
Variable Placebo Wheat grass P value
No. of patients 12 11
Age (years) 33.3
§
10.3 37.4
§
12.8 0.402
Sex (male/female) 9/3 6/5
Years of disease 6.6
§
9.4 7.0
§
9.4 0.924
Sigmoidoscopic score: (no. of patients)
Mild (score 1) 5 4
Moderate (score 2) 6 4
Severe (score 3) 1 3
Severity of disease by patient assessment
Rectal bleeding (scor e 0–3) 0.7
§
0.4 1.4
§
0.9 0.006
Bowel Movements (no.) 4.4
§
2.4 3.6
§
2.5 0.083
Hemoglobin (g/dl) 13.6
§
1.9 14.0
§
1.9 0.644
Pre-study medication : (no. of patients)
None 3 3
5-am ino-salicyli c acid oral
(enem a)
8 (4) 8 (3)
Prednisone oral (enema) 1 (1) 0
Scand J Gastroenterol 2002 (4)
Active Distal Ulcerative C olitis 445
A disease activity index (DAI) was created by combining
the four most accepted measures (according to Sutherland
(7)): stool frequency, rectal ble eding, sigmoidoscopic score
and physician’s assessment of disease activity.
Trial design and treatment
Patients were randomly allocated, through a centralized
randomization process, to receive either 100 cc of wheat grass
juice or matching placebo for 1 month. To ensur e high
qua lity, special methods of preparing and distributing the
wheat grass juice and the placebo were designed. Wheat seeds
were grown organically in indirect sunlight on a substrate
composed of soil, volcanic tuff and compost. The wheat grass
was harvested at the height of 20 cm, packaged and
distributed in a refrigerated vehicle twice weekly to the three
study coordinating centers in three main cities in Israel. Every
morning, at about 0500, workers at each study coordinating
center extracted the wheat grass juice using a manually
ope rated juice machine. Then both the true wheat grass juice
and the placebo were packaged in coded, identical, sealed,
opa que containers . A driver, blinded to the allocation scheme
and given only the address es for each package, then
distributed all the packages. All containers were delivered
within an hour of extraction to ensur e freshness. The patients
were instructed to take the juice immediately. The doses of
both the true juice and the placebo were increased gradually
from an initial dose of 20 cc on the rst day to an increase of
20 cc on each successive day. By the fth day the optimal
daily dose of 100 cc was achieved and given thereafter.
The placebo juice was manufactured from 0.18% normal
saline with a mixture of under 0.5% weight kaolin and
tragacanth, and tinted with food color (Fast Green FCF). The
placebo juice was prepared in a centra lized location and
distributed monthly to the study coordinating centers. The
placebo juice was similar to wheat grass juice in appearance,
but not in taste and smell.
Data collection and scoring
Patients’ diaries and nal subjective assessme nt of cha nge
were collected upon treatment completion. Each nal sub-
jective assessment score (after 1 month of drug or placebo)
was subtracted from the initial subjective score fo r each
patient. The values for the two groups were compared. The
proportions of persons who improved or deteriorated are
shown in Figs 2 and 3. Table III speci es the results in var ious
parameters. A change in a parameter (improvement or aggra-
vation) was de ned as larger than 0.4 in an analog scale where
¡3 designates the lowest score of aggravation, 0 no change,
and ‡3 highest score of improvement. The parameter change
for the number of bowel movements was de ned as greater
than 2 in absolute numbers.
Statis tical analysis
A between-groups analysis (tw o-tailed, Fisher exact test)
was done on the four components of the DAI: rectal bleeding,
stool frequency, sigmoidoscopic score and physician’s assess-
ment of disease activity (Table III). Analyses were con ned to
the 21 patients who completed the trial. The physician’s
assessment of disease activity was compared to the stool
frequency, rectal bleeding, sigmoidoscopic score and the
patient’s retrospective assessment score using the Spearman
rank correlation (Table IV).
R esults
Withdrawals and drop-outs
Of the 24 patients wh o were randomized, 21 completed the
double- blind study. Three withdrew e arly. One patient (from
the treatment group) withdrew within 1 da y of commencing
treatment, because she was unable to tolerate the taste of the
juice (Fig. 1). Another patient (treatment group) withdrew
after 10 days, convinced she was receiving a placebo. The
third patient (placebo group) withdrew after 14 days, due to
deterioration in her illness. Of the 21 patients who completed
the 1-month trial, 19 were evaluated with a second sigmoido-
scopy within 3 days after completion of the study.
Signi cant differences in improvements in rectal bleeding
(P = 0.025), abdominal pain (P = 0.019), the DAI score
Fig. 2. Percentage improvement in DAI (disease activity index) an d
in its 4 constituents : Sigmo (score), PGA (physician global assess-
ment), Blood (rectal bleeding-diary) , No. (daily bowel movements-
diary).
Fig. 3. Percentage deterioratio n in DAI (disease activity index) and
in its 4 constituents : Sigmo (score), PGA (physician global assess-
ment), Blood (rectal bleeding-diary) , No. (daily bowel movements-
diary).
Scand J Gastroenterol 2002 (4)
446 E. Ben-Arye et al.
(P = 0.031), the PGA (P = 0.031), and patients’ retrospective
evaluation (P = 0.0053) were demonstrated in the treatment
versus the placebo groups (Table III). No signi cant differ-
ences were found in the number of bowel movements or sense
of abdominal bloating between the two groups. A greater
proportion of participants improved on the sigmoidoscopic
evaluation in the treatment group (7 of the 9 patients (78%)),
compared to the placebo group (3 of the 10 patients (30%)).
However, this difference was not signi c ant (P = 0.13).
Credibility of blindness was ass essed by a questionnaire at
the end of the study. In the treatment group, 6 of 11 patients
believed they were getting wheat grass juice, and attributed it
to their improved well-being; three were convinced they were
getting placebo, and two did not give a de nite answer. In the
placebo group, 7 out of 12 patients did not give a de nite
answer, 2 believed they were getting wheat grass, and 3
believed they were taking placebo.
Table V summarizes both the positive and the negative
effects, which were reported by participants who received
wheat grass juice. Nausea was found to be the most adverse
effect (33% ). A prominent positive side effect was increased
vitality (41%). No side effects were reported in the placebo
group.
An attempt was made to identify likely bioactive com-
pounds tha t might explain som e of the activity of wheat grass
juice. Reactive oxygen species (ROS) have been suggested to
be involved in th e induction and prognosis of UC. Anti-
oxida nts were demonstrated to be bene cial in preventing
intestinal injury in general, and UC in particular. A signi cant
reduction in the activity of the colonic low molecular weight
anti-oxidants (LMWA) has already been reported (8); there-
fore, in this study we evaluated the overall LMWA capacity
of the wheat grass juice to see its pro le. The cyclic
volta mmetry (CV) methodology was used for this purpose,
as described in detail else where (Chevion et al.) (9). Fresh
extract of wheat grass demonstrated a prominent tracing with
at least fou r peaks (Fig. 4), presumably corresponding to four
groups of compounds that exhibit anti-oxidative properties.
These four pea ks decreased in number and amplitude over
time (Fig. 5).
D iscussion
The results of this double-blind study indicate that wheat
grass juice has a therapeutic role in treating patients with
active restricted left colon UC. We have demonstrated that
wheat grass is a generally safe substance, wit h fe w side effects
(Table V ).
Table III. Clinical state after 1 mont h of treatment
Variable Study group Improved No change Worsened P value
Disease Activity Index (DAI) Treatment (T) 10 1 0 0.0312
Placebo (P) 5 2 5
1. Rectal bleeding T 9 2 0 0.0254
P 7 0 5
2. Bowel movements (No.) T 8 1 2 0.37
P 7 0 5
3. Sigmoidoscopi c assessment * T 7 1 1 0.127
P 3 5 2
4. Physician global assessm ent (PGA) T 10 1 0 0.0312
P 5 2 5
Other indices of activity (Diary)
Mucous T 5 2 3 0.167
P 8 4 0
Pain T 8 1 0 0.0191
P 3 4 5
Abdominal bloating T 7 3 1 0.376
P 3 2 7
Patient’s retrospective evaluation T 10 0 0 0.0053
P 5 3 4
* 10 patients in placebo and 9 in treatment group.
Table IV. Correlation between physicia n global assessment and
other parameters of disease activity
Variable
No. of
subjects
Correlation
rho P value
Bowel movements 33 0.410 0.0178
Rectal bleeding 33 0.775 0.0001
Retrospective patient assessment 33 0.773 0.0001
Sigmoidoscopi c appearance 30 0.618 0.0003
Fig. 4. Cyclic voltammetry of fresh wheat grass juice.
Scand J Gastroenterol 2002 (4)
Active Distal Ulcerative C olitis 447
Both subjective and objective parameters are important in
the clinical evaluation of UC. Various disease activity assess-
ments that combine these parameters have been developed in
the past 40 years, initially by Truelove & Witt s in 1955 (10)
and later by Seo et al. (11) an d Sutherland (7). It is notable
that the correlation between sigmoidoscopic evaluation and
clinical activity in UC is well established but not absolute, as
the length of the in amed colon does not always correspond
with the severity of the clinica l appearance (12). According to
the four parameters of the DAI we used [based on Sutherland
(7)], overall clinical evalua tion of the treatment group demon-
strated a signi cant improvemen t compared to the placebo
group. Although sigmoidoscopy did not demonstrate a sig-
ni cant improvement, there was a distinct trend toward
improvement.
The present study indicates a possible clinical ef cacy of
wheat grass juice in treating UC. Several questions emerge.
First, does whe at grass possess anti-in ammatory properties?
If so, which components are the active ones? Second, does the
active substance exert a general anti-in amma tory effect via
proxim al intestinal absorption, or does it act locally at the
in amed site itself?
When wheat grass juice has been analyzed using cyclic
volta mmetry, four distinct peaks emerge. These four peaks
may indicate that there are at least four group s of active
ingre dients that possess redox potentials and anti-oxidative
prope rties. There is growing evidence that free radicals play a
role in the pathogenesis of UC (8). If this is the case, then anti-
oxida tive agents may be bene c ial therapy.
One mechanism by which wheat grass juice may improve
the symptoms of UC can involve avonoids, speci cally
apigenin, a known ingredient of wheat grass juice (4). The
well-established biological activity of avonoids includes
anti-in ammation (via the arachidonic acid pathway (13)) and
anti-oxidation (14). Apigenin, a potent bio- avonoid found in
wheat grass juice, has been shown to be an anti-oxidant in the
xantine/xantine oxidase system (15), anti-in ammatory in
inhibiting adhesion of leukocytes to endothelia (16), anti-
mutagenic in the Ames Test (17) and inhibiting towards colon
motility (18).
No answer is available at present as to the site of wheat
grass juice ac tion. This question can be addressed in future
studies by administration of wheat grass juice in enemas or
suppos itories. Many other questions are raised by ou r results.
For example, does wheat grass juice have a therapeutic
advantage over other modes of treatment for UC as a stand-
alone treatment or an adjunct to drug administration? Does it
have any role in prophylaxis or only in acute attacks? Should
it be offered to patients with extensive UC or Crohn colitis as
well? What are optimum doses and modes of administration,
and are there interactions with other drugs? The authors
believe that these questions should be tested in future studies.
Although it is noteworthy that signi cant ndings emerge
from the study with such a small sample size, ther e are some
limitations of this study because of its small sample. For
example, the sample was too small to be able to identify
characteristic of responders. Secondly, the study required
participants to undergo two sigmoidoscopies, use an uncon-
ventional treatment, an d travel. Given these requirements, the
study participants may be a non-representa tive, highly
motivate d sample of UC patients. Another trial with a larger
group of patients could address both of these shortcomings.
We can also, in retrospect, identify at least two ways that
future trials can improve on this study. First, it is recom-
mended that the sigmoidoscopy lms be read by two inde -
pendent reade rs blinded to each othe r’s assessm ent, and that
inter-rater agreements be calculated. This would highlight,
and minimize, single-reader bias. Secondly, an assessment of
whether participants believe they are receiving the placebo or
the true juice before there is any chance of a change in the
disease activity, for example at about 3 days.
We believe that wheat grass juice offers a genuine
therapeutic advantage in the disabling disease of UC. Besides
apparent ef cacy, adverse effects were relatively benign.
Notable is the potential of this regimen of empowering the
patient in becoming an active participant in the therapeutic
process. The effort of obtaining or growing wheat grass and
the daily production of fresh juice may be important in
enhancing the patients’ sense of control and alleviating some
of the patient distress de scribed earlier. In any case, use of
wheat grass juice in subsets of patients , as singl e or adjuvant
remedy in acute left-sided UC, may prove to be bene cial
therapeutically.
R eferences
1. Wigmore A. The wheatgrass book. Avery Publishing Group;
1986.
Fig. 5. Cyclic voltammetry of non-fres h wheat grass juice.
Table V. Side effects of wheat grass juice administration
No. of patients (%)
Nausea
Mild 3 (25)
Severe 1 (8.3)
Decreased morning appetite 2 (16.6)
Constipation 1 (8.3)
Increased vitality 5 (41.6)
Scand J Gastroenterol 2002 (4)
448 E. Ben-Arye et al.
2. Walters R. The alternative cancer therapy book. Avery Publish-
ing Group; 1992. p. 299–308.
3. Lai CN. Chlorophyll: The active factor in wheat sprout extract
inhibiting the metabolic activatio n of carcinogen s in vitro.
Nutrition Cancer 1979;1:19–21.
4. Peryt B, Szymczyk T, Lesca P. Mechanism of antimutagenicit y
of wheat sprout extracts. Mutation Research 1992;269:201 –
15.
5. Hanauer SB. Drug therapy: in ammatory bowel disease. N Engl
J Med 1996;334;841 –8.
6. Dudley-Brown S. Living with ulcerative colitis. Gastroenterol
Nurs 1996;19;60 –4.
7. Sutherland LR, Martin F, Greer S. 5-Aminosalicyli c acid enema
in the treatment of distal ulcerative colitis, proctosigmoiditi s and
proctitis. Gastroenterolog y 1987;92:1894 –8.
8. Buf nton GD, Doe WF. Depleted mucosal antioxidant defenses
in in ammatory bowel disease. Free R adic Biol Med 1995;19;
911– 8.
9. Chevion S, Berry EM, Kitrossky N, Kohe n R. Evaluation of
plasma low molecular weight antioxidan t capacity by cyclic
voltam metry. Free Radical Biol Med 1997;22:411 – 21.
10. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. BMJ
1955;2;1041 –8.
11. Seo M, Okada M, Yao T, Ueki M, Arima S, Okumura M. An
index of disease activity in patients with ulcerativ e colitis. Am J
Gastroentero l 1992;87;971 –6.
12. Powell-Tuck J, Day DW, Buckell NA, Wadsworth J, Lennard-
Jones JE. Correlation between de ned sigmoidoscopic appear-
ances and other measures of disease activity in ulcerative colitis.
Dig Dis Sci 1982;27;533 –7.
13. Alcara z MJ, Ferrandiz ML. Modi cation of arachidoni c metab-
olism by avonoids. J Ethnopharmaco l 1987;21;209 –29.
14. Saija A, Scalese M, Lanza M, Marzullo D, Bonina F, Castelli F.
Flavonoids as antioxidant agents: importance of their interaction
with biomembranes . Free Radical Biol Med 1995;19;481 –6.
15. Wei H, Bowen R, Cai Q, Barnes S, Wang Y. Antioxidant and
antipromotiona l effects of the soybean iso avone genistein . Proc
Soc Exp Biol Med 1995;208;124 –30.
16. Gerritsen ME, Carley WW, Ranges GE. Flavonoids inhibit
cytokine-induce d endothelial cell adhesion protein gene expres-
sion. Am J Pathol 1995;147;278 –92.
17. Peryt B, Miloszewska M, Zielenska M, Szymczyk T. Antimuta-
genic effects of several subfraction s of extract from wheat sprout
tow ards benzo(a)pyrene-induce d mutagenicity in strain TA-98 of
salmonella typhimurium . Mutat Res 1988;206;221 –5.
18. DiCarlo G, Autore G, Izzo AA. Inhibitio n of intestina l motility
and secretion by avonoids in mice and rats. J Pharm Pharmacol
1993;45;1054 –9.
Received 10 May 200 1
Accepted 20 September 2001
Scand J Gastroenterol 2002 (4)
Active Distal Ulcerative C olitis 449
