Aliment Pharmacol Ther 2002; 16: 197±205.
Antioxidant effects of herbal therapies used by patients
with in¯ammatory bowel disease: an in vitro study
L. LANGMEAD, C. DAW SON, C. HAWK INS, N . BANNA, S. LOO & D. S. RAMPTON
Academic Department of Adult and Paediatric Gastroenterology, Barts and the London, Queen Mary's, School of Medicine and
Dentistry, London, UK
Accepted for publication 21 August 2001
SUMMARY
Background: Herbal remedies used by patients for treat-
ment of in¯ammatory bowel disease include slippery
elm, fenugreek, devil's claw, Mexican yam, tormentil
and wei tong ning, a traditional Chinese medicine.
Reactive oxygen metabolites produced by in¯amed
colonic mucosa may be pathogenic. Aminosalicylates
(5-ASA) are antioxidant and other such agents could be
therapeutic.
Aims: To assess the antioxidant effects of herbal remed-
ies in cell-free oxidant-generating systems and in¯amed
human colorectal biopsies.
Methods: Luminol-enhanced chemiluminescence in a
xanthine/xanthine oxidase cell-free system was used to
detect superoxide scavenging by herbs and 5-ASA, and
INTRODUCTION
Conventional therapies used in in¯ammatory bowel
disease are not totally successful in achieving remission
or preventing relapse, and may cause serious side-
effects, therefore many patients seek alternative options.
Indeed, surveys of patients with in¯ammatory bowel
disease indicate that up to half have tried or currently
employ some form of complementary therapy, and that
this ®gure is steadily rising.1±3 The most commonly
used alternative remedies are herbal, the main reasons
cited by patients for their purchase being the chronicity
Correspondence to: Dr D. S. Rampton, Endoscopy Unit, Royal London
Hospital, London E1 1BB, UK.
E-mail: D.Rampton@mds.qmw.ac.uk
Ó 2002 Blackwell Science Ltd
¯uorimetry to de®ne peroxyl radical scavenging using a
phycoerythrin degradation assay. Chemiluminescence
was used to detect herbal effects on generation of
oxygen radicals by mucosal biopsies from patients with
active ulcerative colitis.
Results: Like 5-ASA, all herbs, except fenugreek, scav-
enged superoxide dose-dependently. All materials tested
scavenged peroxyl dose-dependently. Oxygen radical
release from biopsies was reduced after incubation in all
herbs except Mexican yam, and by 5-ASA.
Conclusions: All six herbal remedies have antioxidant
effects. Fenugreek is not a superoxide scavenger, while
Mexican yam did not inhibit radical generation by
in¯amed biopsies. Slippery elm, fenugreek, devil's claw,
tormentil and wei tong ning merit formal evaluation as
novel therapies in in¯ammatory bowel disease.
and poor quality of life associated with in¯ammatory
bowel disease.4±6
Reactive oxygen metabolites (ROM) are present in
excess in in¯amed colonic mucosa and are likely to play
a pathogenic role in in¯ammatory bowel disease.7±10
The antioxidant actions of aminosalicylates (5-ASA)
may constitute at least part of their therapeutic
effect.11±13 Novel drugs or therapies for in¯ammatory
bowel disease that have antioxidant activity may be
useful clinically. For example, allopurinol, a xanthine
oxidase inhibitor, appears to be effective in pouchitis,14
and in prolonging the time to relapse in ulcerative
colitis.15 Superoxide dismutase showed encouraging
preliminary results in refractory Crohn's disease.16
Although many herbal concoctions are said to be
effective in chronic in¯ammatory conditions, there is
197
198 L. LANGMEAD et al.
little clinical and pharmacological data to support
these claims. The aim of this work was to investigate
the antioxidant effects in vitro of six herbal remedies
claimed to have bene®t in in¯ammatory bowel disease
and diseases such as rheumatoid arthritis: slippery
elm, derived from the bark of the slippery elm tree;
fenugreek, an Ayurvedic therapy; Mexican yam, a
tropical staple; Devils claw, the root of an African
¯ower; tormentil, a European ¯ower; and Wei tong
ning, a traditional Chinese herbal therapy. The ®rst
®ve remedies are available for sale in health food
stores in the West, although their exact composition
may vary according to proprietary brand. Wei tong
ning is used in Chinese traditional medicine for
treatment of ulcerative colitis. All the preparations
contain entire plant extracts rather than puri®ed
chemical entities.
We have tested the antioxidant effects of each of these
herbal preparations in two cell-free radical generating
systems, xanthine/xanthine oxidase for superoxide and
the phycoerythrin degradation assay for the peroxyl
radical, and in in¯amed colorectal mucosal biopsies,
using 5-ASA as a positive control11 and orange juice as
a nontherapeutic phytic control.
METHODS
Materials
All chemicals were provided by Sigma Chemical Co. UK
except for 2,2¢-azobis (2-amidinopropane) dihydrochlo-
ride (AAPH), which was obtained from Polysciences
Inc., Warrington, PA, USA.
The sources of herbal remedies were as follows:
slippery elmÐPotter's (Herbal Supplies) Ltd, Wigan,
UK; fenugreekÐGood `N Natural Manufacturing Corp.
(USA) for Holland and Barrett, Nuneaton, UK; devils
clawÐBio-Health Ltd, Rochester, Kent, UK; Mexican
yamÐHigher Nature, Burwash Common, East Sussex,
UK; tormentilÐBioforce (UK) Ltd, Irvine, Scotland,
UK. Wei tong ning was obtained from the China
Academy of Traditional Chinese Medicine, Beijing,
China.
Reagents
Luminol was made up as a stock solution, 50 mg/mL, in
dimethyl sulfoxide (DMSO) and stored at 4 °C for up to
1 month. On the day of experiments it was diluted to
300 lM in Dulbecco's phosphate buffered saline solu-
tion (PBS) with glucose 5 mmol. A stock solution of
xanthine, 5 mmol, was made up by dissolving xanthine
in 1 M sodium hydroxide and dilution in PBS and stored
for up to 1 month; it was added to luminol on the day of
experiments to a ®nal concentration of 50 lM. Xanthine
oxidase was stored at 4 °C and diluted to a concentra-
tion of 0.017 U/mL in the ®nal experimental mixture.
0.34 lM aliquots of 1 mL phycoerythrin were prepared
by dissolving the protein precipitate in PBS, centrifuging
for 4 min and re-dissolving the pellet in PBS before
removing (NH4)2SO4 on a NAP-5 column (a prepacked
disposable chromatography column containing Sepha-
dex G-25 Medium of DNA grade) with Chelex-treated
PBS; aliquots were stored at 4 °C wrapped in foil. AAPH
was prepared on the day of the experiment in Chelex-
treated PBS to give a ®nal cuvette concentration of
10 nM.
To prepare mixtures of herbs for use in the assay
systems, a single dose of the crude preparation (tablet,
capsule contents, liquid or pellet) was crushed and
powdered if necessary and dissolved in the smallest
possible volume of water (slippery elm, fenugreek,
Mexican yam, Devil's claw) or DMSO (Wei tong ning)
according to solubility. Any excess particulate matter
was removed by centrifugation. For each herb, the
supernatant resulting from this procedure was kept as
stock for up to 4 weeks at 4 °C and used in all
subsequent dilutions. Orange juice was prepared fresh
on the day of experiment. The juice from an orange was
homogenized and diluted in PBS or Tyrodes medium.
Dilutions are expressed as parts original stock per
volume of diluent (ppv). 5-ASA was dissolved in 1 M
Na OH on the day of experiments and made up with PBS
to a 10-mmol stock solution used in all subsequent
dilutions for cell-free experiments. For biopsy incuba-
tions a ®nal concentration of 20 mmol 5-ASA was
prepared using Tyrodes medium.
Patients
Sigmoid or rectal mucosal biopsies were obtained from
patients with active ulcerative colitis undergoing rout-
ine colonoscopy for assessment or surveillance. Diag-
nosis had been con®rmed previously by conventional
clinical, colonoscopic and histological criteria. Disease
activity was assessed using clinical17 and sigmoidoscopic
scores.18 Details of age, sex, disease extent and current
therapy are shown in Table 1. Written informed
Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205
 ANTIOXIDANT EFFECTS OF HERBAL THERAPIES IN IBD
Table 1. Characteristics of patients with ulcerative colitis used for
in vitro studies of colorectal mucosal biopsies incubated with herbs
and 5 ASA
n ˆ 23
Age (years, median[IQR]) 45 [34±51]
Sex, male:female 14:9
Disease extent
Total 11
Sub-total 4 540 nm and emission at 575 nm, thermostated to
Left-sided 8 37 °C and ®tted with an electronic cuvette stirrer.
Disease activity scores
Powell-Tuck (median [IQR]) 5 [2±8]
Baron sigmoidoscopy score 2 [1±3]
Oral treatment
5-ASA 16
Prednisolone 5 against time. The coef®cient of variation of assays was
Azathioprine 4 10.1% (n ˆ 21).20 (Orange juice was not tested using
consent was obtained prior to colonoscopy at which
8±12 biopsies from macroscopically in¯amed mucosa
were taken. The study was approved by the East London
and City Health Authority Ethics Committee.
Xanthine/xanthine oxidase cell-free system
The xanthine-xanthine oxidase cell-free system produ-
ces superoxide subsequently detected by luminol-
dependent chemiluminescence.11, 19 One millilitre of
preoxygenated luminol + xanthine solution, with
added herbs or orange juice to give ®nal dilutions of
1 in 50, 1 in 500, 1 in 5000, 1 in 50 000 and 1 in
500 000 ppv were added to vials, adjusted to pH 7.4
and precounted. 5-ASA (10±4±10±7 M) was used as a
positive control. Xanthine oxidase was added to start
the reaction and vials were immediately recounted
over 15 min in repeated sequence for 0.5 s per vial in
an automated LB 953 luminometer (Bertholdt, Bad
Wildbad, Germany). The ®nal result for each control or
herbal dilution was recorded as the mean value of the
integral count in counts/s for duplicate vials. The
median coef®cient of variation for assays was 16%
(n ˆ 21).20
Phycoerythrin degradation cell-free assay
Free radicals were generated using AAPH, a chemical
that decomposes at a temperature-controlled rate to
form carbon-centred radicals that react with oxygen to
Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205
199
yield peroxyl radicals.21 Phycoerythrin is a ¯uorescent
pigment protein derived from algae; its exposure to
oxygen radicals causes zero-order decay in ¯uores-
cence. Phycoerythrin was desalted as stated earlier.
Fluorescence of 1 mL 20 nM phycoerythrin with or
without addition of a known dilution of herbal sample
or 5-ASA was measured in a spectro¯uorimeter
(Kontron Instruments, Milan, Italy) with excitation at
Baseline ¯uorescence in the presence of sample was
measured for 2±3 min before addition of 10 nM AAPH.
Fluorescence was then monitored for 25 min. The rate
of decay of phycoerythrin ¯uorescence was calculated
from the slope of the curve of ¯uorescence plotted
this assay.)
Mucosal biopsy incubations
Four to six paired biopsy samples from adjacent areas of
similar macroscopic in¯ammation18 from each individ-
ual patient were placed initially in preoxygenated
incubation medium (Tyrodes medium) before transfer
to a tube containing 1 mL of either Tyrodes medium as
a control, or Tyrodes with added herbal mixtures or
orange juice to ®nal dilutions of 1 in 100 ppv (1/1000
ppv for tormentil), or 20 mmol 5-ASA. Herbal dilutions
were chosen to approximate those likely to be present in
the colonic lumen. This was estimated assuming
minimal absorption or digestion of phytic constituents
from the small intestine, and an ileal ef¯uent of about
1 L per day. Tormentil was tested at 1 in 1000 because
of its greater potency in the cell-free systems (see below).
The 5-ASA concentration chosen was used because it
resembles colonic luminal concentration in vivo22 and
has previously been demonstrated to inhibit release of
ROM from colonic biopsies.11 Biopsies were incubated at
37 °C with a continuous ¯ow of O2 95%/CO2 5%
bubbling through the medium. At 20 min, fresh
medium was exchanged for old. After 40 min, biopsies
were washed to remove residual test herb or 5-ASA, and
transferred to precounted scintillation vials containing
1 mL 300 lM luminol with 5 mmol glucose, pH 7.4.
Vials were immediately counted for 4 mins each in the
LB953 luminometer. After assay, biopsies were blot-
dried and weighed. Chemiluminescence was recorded as
the mean for paired biopsies expressed as counts/min/mg
200 L. LANGMEAD et al.

Figure 1. Dose±response curves for inhibitory effects of herbs, orange juice and 5-aminosalicylic acid (5-ASA) on chemiluminescence
produced by oxygen radicals released by the xanthine/xanthine oxidase system. Dilution of herbs and orange juice is expressed as parts
per volume (ppv) and 5-ASA concentration in mols (n ³ 3 for each agent).

Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205

 ANTIOXIDANT EFFECTS OF HERBAL THERAPIES IN IBD 201

tissue weight, after subtraction of background. The

coef®cient of variation of chemiluminescence counts for
paired biopsies was 47% (n ˆ 30).20
Calculations and statistics
For all assays, the effect of each herb or 5-ASA was
expressed as a percentage of the mean result of the
control pair for that experiment. Results were com-
pared using the Wilcoxon signed ranks test for paired
data using 2-tailed P-values and taking < 0.05 as
statistically signi®cant. All comparisons were planned
prospectively, therefore no correction was made for
multiple comparisons.23 Dose±response curves were
drawn using linear logistic regression curves for the
mean values for three experiments for each herb at
each dilution. The IC50 was obtained from the line of
best ®t. Coef®cients of variation were calculated using
the method of Bland.20
RESULTS
Xanthine/xanthine oxidase cell-free system
Slippery elm, Mexican yam, Devil's claw, tormentil,
Wei tong ning and 5-ASA all had dose-dependent
antioxidant effects (Figure 1). Fenugreek and orange
juice had no effects except undiluted (Figure 1). The
potency for each herb, as assessed by the dilution that
produced 50% inhibition of ROM detection (IC50),
ranged between 10±4 for Mexican yam and 10±6 for
tormentil (Table 2).
Phycoerythrin degradation assay
All the herbs and 5-ASA had dose-dependent peroxyl-
radical scavenging effects (Figure 2). The IC50 for this
effect varied between 10±2 for Mexican yam and 10±5
for tormentil (Table 2).
In¯amed colorectal biopsies
In comparison with controls, chemiluminescence was
signi®cantly reduced in biopsies incubated with 1 in
100 dilutions of slippery elm (P < 0.02), fenugreek
(P < 0.05), Devil's claw (P < 0.05), and Wei tong
ning (P < 0.02), and 1 in 1000 dilution of tormentil
(P < 0.05). There was no signi®cant antioxidant effect
with Mexican yam or orange juice (Figure 3).
DISCUSSION
In previous studies, we have used cell-free and mucosal
biopsy chemiluminescence-based methods alone for
assessing the antioxidant effects of conventional and
potentially novel drugs for in¯ammatory bowel dis-
ease.11, 24 In the present work we have also employed
a ¯uorescence-based cell-free assay. With the two cell-
free systems, scavenging of both superoxide and
peroxyl can be assessed. While the cell-free techniques
have the advantages of being technically straightfor-
ward, readily available, highly reproducible, speci®c in
relation to the radical generated, and allowing dose±
response curves to be generated, the use of a tissue-
based method is likely to provide a closer model of
what may occur in vivo. The mucosal luminol-based
chemiluminescence assay sensitively but nonspeci®cal-
ly detects production of multiple different ROMs, of
which the predominant species in active ulcerative
colitis are probably hypochlorite, hydrogen peroxide
and hydroxyl ion as well as superoxide.24 The results
obtained with each assay method therefore comple-
ment each other: as indicated below, however they do
not precisely mimic each other.
In vivo, antioxidants may exert their activity at
several different points in the pathways of oxidant
metabolism.25, 26 Prevention of radical generation may
be achieved at a tissue level as a result of down-
regulation of recruitment and/or activation of neu-

Table 2. IC50 for herbs and 5 ASA in

IC50 SE FG DC MY T WTN 5 ASA
xanthine/xanthine oxidase and
phycoerythrin degradation assays X/XO 3 ´ 10)4 N/A 3 ´ 10)4 3 ´ 10)4 1 ´ 10±6 5 ´ 10)4 10)5 M
PE 1 ´ 10)3 1 ´ 10)3 7 ´ 10)3 8 ´ 10)2 1 ´ 10)5 1 ´ 10)3 10)5 M

The concentration (expressed as dilution in parts per volume (ppv) for herbs and mols for 5
ASA) to produce 50% inhibition of ROM detection (IC50) is shown for the xanthine/xanthine
oxidase system (X/XO) (Figure 1) and the phycoerythrin degradation assay (PE) (Figure 2)
(n ³ 3 for each agent). SE ˆ slippery elm, FG ˆ fenugreek, DC ˆ devil's claw, MY ˆ Mexican
yam, T ˆ tormentil, WTN ˆ Wei tong ning, 5 ASA ˆ 5-aminosalicylic acid. Results for
orange juice are not shown because no inhibition was seen.

Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205

202 L. LANGMEAD et al.

Figure 2. Dose±response curves for inhibitory effects of herbs and

5-aminosalicylic acid (5-ASA) on phycoerythrin degradation
produced by peroxyl radicals released by the thermal decomposi-
tion of AAPH. Herb dilution is expressed as parts per volume (ppv)
and 5-ASA concentration in mols (n ³ 3 for each agent).

Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205

 ANTIOXIDANT EFFECTS OF HERBAL THERAPIES IN IBD
Figure 3. Effects of herbs and orange juice
at 1 in 100 dilution (except tormentil ˆ 1
in 1000) and 5-aminosalicylic acid
20 mmol (5-ASA) on release of reactive
oxygen metabolites from in¯amed
colorectal mucosal biopsies measured by
luminol-ampli®ed chemiluminescence after
40 min incubation. Results are expressed
as percentage of chemiluminescence
produced by control biopsies incubated in
inert vehicle. Median, IQR and range are
shown.*P < 0.02, **P < 0.05 compared
with controls, n ³ 5 for each.
trophils or monocytes. At a cellular level activation of
in¯ammatory cells may be reduced by blockade of
membrane receptors or inhibition of intracellular
metabolic pathways. Prevention of the activity of
radicals after their generation and release can occur
as a result of scavenging by antioxidants. Protection of
target tissues from radical attack from the lumen of the
intestine could also be a result of enhancement of
physico±chemical barriers, for example by increased
mucus production. The cell-free methods used in the
present studies provide information about superoxide
and peroxyl scavenging, while the mucosal biopsy
results give an indication of the net effects of the herbs
tested on tissue oxidant activity without elucidating
their mechanism.
We have employed these three assays to investigate
the antioxidant effects of six widely used herbal
preparations.
Slippery elm bark from the slippery elm, or red elm tree
native to North America, is claimed to have ``soothing''
properties in in¯ammation of the gastrointestinal
tract. It is popular among in¯ammatory bowel disease
patients in the UK.6
Fenugreek is an aromatic herb that produces Methi
seeds, which are used in Indian cookery for their
distinctive pungent aroma. In traditional Ayurvedic
medicine, fenugreek, like slippery elm, is said to have a
Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205
203
bene®cial effect on in¯amed intestines. Fenugreek
contains steroidal saponins, which form the basis for
physiological steroid production and might be able to
in¯uence the local in¯ammatory response.27±29
Mexican yam is a tropical perennial whose starch-rich
tuberized root is a food staple in the areas in which it
grows. Like fenugreek, it contains saponins in the form
of dioscorea, a steroid extract used in commercial
steroid synthesis. Mexican yam is used by herbalists
for the treatment of menstrual irregularities as well as
joint and gut in¯ammation.
Devil's claw (Harpagophytum procumbens) is a ¯ower-
ing plant native to southern Africa. The tuberized
secondary roots are used in parts of Africa for liver and
stomach ailments and for joint pains; it is said to have
analgesic and anti-in¯ammatory properties. Uncon-
trolled studies in rheumatic and degenerative joint
disease have shown an improvement in mobility and
reduction of pain.30 Ex-vivo and in vitro studies failed
to show any effect on eicosanoid production in whole
blood,31, 32 suggesting that the anti-in¯ammatory
effects (if any) of devil's claw are mediated through
other mechanisms. Extracts of devil's claw contain
¯avonoids, which are proven free radical scaven-
gers,33, 34 and plant phenols, which are able to donate
hydrogen to oxygen radicals to form stable inert
compounds.35, 36
204 L. LANGMEAD et al.
Tormentil (Potentilla tormentilla) is a member of the
rose family that grows wild all over Europe. The root is
thick and red on the inside giving rise to the herb's
alternative name, bloodroot. It is said to be effective in
the treatment of diarrhoea and bowel in¯ammation.
Tormentil contains a high concentration of tannins
known to have a potent superoxide-scavenging
effect.37, 38
Traditional Chinese medicine uses both physical (e.g.
acupuncture) and herbal treatments to redress imbal-
ances in energy meridians said to be associated with
disease. The remedy we have tested, wei tong ning, is
prescribed by Chinese herbalists for treatment of
patients with peptic ulceration and other intestinal
in¯ammatory disorders.
All the herbs tested are likely to contain numerous
biologically active and, speci®cally, antioxidant com-
pounds. These include vitamins C and E, in varying
concentrations, both naturally and in some prepara-
tions as preservative additives. To address the fact that
many vegetable and other food products also contain
potentially biologically active components and yet are
not claimed to have any therapeutic bene®ts, we have
used orange juice as a phytic control.
Our results have shown that, like 5-ASA, slippery elm,
devil's claw, tormentil and wei tong ning are antioxid-
ant in all three assay systems used. Fenugreek failed to
scavenge superoxide except at high concentrations, but
did have peroxyl-radical scavenging effects and reduced
release of ROMs from in¯amed mucosa. It seems likely
that different components of fenugreek are responsible
for its con¯icting effects. The observation that Mexican
yam is a scavenger in cell-free systems but has no effect
in biopsies could be explained by failure of its antioxid-
ant component(s) to penetrate mucosal tissue during
the short incubation period used. Alternatively, it is
possible that Mexican yam contains substances that
have pro-in¯ammatory effects in tissue that are not
detectable in cell-free assays.
That orange juice fails to have an antioxidant effect
except at nearly neat concentrations is perhaps surpri-
sing, given its vitamin C content. However, it suggests
that the potency shown by the herbal remedies tested
may be therapeutically signi®cant and adds weight to
the claim that particular herbs may have antioxidant
constituents and actions over and above those seen in
many common foods. This conclusion is supported by
the observation that they behave similarly to the
puri®ed pharmacological agent, 5-ASA.
Further exploration of the antioxidant chemistry of
herbal extracts would require separation and puri®ca-
tion of the multiple individual potentially bioactive
components of each. However, in the absence of
controlled trials indicating their ef®cacy in in¯amma-
tory bowel disease or other in¯ammatory diseases, this
time-consuming and expensive exercise may be prema-
ture and contradicts the concept of use of whole plant
extracts by herbalists.
The results reported here indicate that a range of
widely used herbal preparations have antioxidant effects
both in cell-free and mucosal biopsy assay systems.
Consideration should be given to formal evaluation
in vivo of the therapeutic potential of slippery elm,
fenugreek, devil's claw, tormentil and wei tong ning in
patients with in¯ammatory bowel disease and other
chronic in¯ammatory conditions.
REFERENCES
1 Eisenberg DM, Davis RB, Ettner SL, et al. Trends in alternative
medicine use in the United States, 1990±97: results of a
follow-up national survey [see comments]. JAMA 1998;
280(18): 1569±75.
2 Hilsden RJ, Scott CM, Verhoef MJ. Complementary medicine
use by patients with in¯ammatory bowel disease [see com-
ments]. Am J Gastroenterol 1998; 93(5): 697±701.
3 Moody GA, Eaden JA, Bhakta P, et al. The role of complementary
medicine in European and Asian patients with in¯ammatory
bowel disease. Public Health 1998; 112(4): 269±71.
4 Moser G, Tillinger W, Sachs G, et al. Relationship between the
use of unconventional therapies and disease-related concerns:
a study of patients with in¯ammatory bowel disease.
J Psychosom Res 1996; 40(5): 503±9.
5 Hilsden RJ, Meddings JB, Verhoef MJ. Complementary and
alternative medicine use by patients with in¯ammatory bowel
disease: An Internet survey. Can J Gastroenterol 1999; 13(4):
327±32.
6 Langmead L, Chitnis M, Rampton D. Complementary
therapies in GI patients: who uses them and why? Gut 2000;
46(Suppl. II): A22(Abstract).
7 Simmonds NJ, Rampton DS. In¯ammatory bowel diseaseÐa
radical view. Gut 1993; 34(7): 865±8.
8 Grisham MB. Oxidants and free radicals in in¯ammatory
bowel disease. Lancet 1994; 344(8926): 859±61.
9 Harris ML, Schiller HJ, Reilly PM, et al. Free radicals and other
reactive oxygen metabolites in in¯ammatory bowel disease:
cause, consequence or epiphenomenon? Pharmacol Ther
1992; 53(3): 375±408.
10 Middleton SJ, Shorthouse M, Hunter JO. Increased nitric oxide
synthesis in ulcerative colitis [see comments]. Lancet 1993;
341(8843): 465±6.
11 Simmonds NJ, Millar AD, Blake DR, et al. Antioxidant effects
of aminosalicylates and potential new drugs for in¯ammatory
Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205
 ANTIOXIDANT EFFECTS OF HERBAL THERAPIES IN IBD
bowel disease: assessment in cell-free systems and in¯amed
human colorectal biopsies. Aliment Pharmacol Ther 1999;
13(3): 363±72.
12 Yamada T, Grisham MB. Role of neutrophil-derived oxidants in
the pathogenesis of intestinal in¯ammation. Klin Wochenschr
1991; 69(21±23): 988±94.
13 Keshavarzian A, Fields JZ. A radical perspective on ulcerative
colitis [editorial; comment]. J Gastroenterol Hepatol 1995;
10(2): 208±9.
14 Levin KE, Pemberton JH, Phillips SF, et al. Role of oxygen free
radicals in the etiology of pouchitis. Dis Colon Rectum 1992;
35(5): 452±6.
15 Jarnerot G, Strom M, Danielsson A, et al. Allopurinol in
addition to 5-aminosalicylic acid based drugs for the main-
tenance treatment of ulcerative colitis. Aliment Pharmacol
Ther 2000; 14(9): 1159±62.
16 Emerit J, Pelletier S, Tosoni-Verlignue D, et al. Phase II trial of
copper zinc superoxide dismutase (CuZnSOD) in treatment of
Crohn's disease. Free Radic Biol Med 1989; 7(2): 145±9.
17 Powell-Tuck J, Day DW, Buckell NA, et al. Correlations
between de®ned sigmoidoscopic appearances and other
measures of disease activity in ulcerative colitis. Dig Dis Sci
1982; 27(6): 533±7.
18 Baron J, Connell A, Lennard-Jones A. Variation between
observers in describing mucosal appearances in proctocolitis.
BMJ 1964; 1: 189±92.
19 Millar AD, Rampton DS, Chander CL, et al. Evaluating the
antioxidant potential of new treatments for in¯ammatory bowel
disease using a rat model of colitis. Gut 1996; 39(3): 407±15.
20 Bland J. Clinical measurementÐrepeatability and precision
measurement. An Introduction to Medical Statistics, 3rd edn.
Oxford: Oxford University Press, 1991: 267±8.
21 Glazer AN. Phycoerythrin ¯uorescence-based assay for
reactive oxygen species. Meth Enzymol 1990; 186: 161±8.
22 Staerk Laursen L, Stokholm M, Bukhave K, et al. Disposition of
5-aminosalicylic acid by olsalazine and three mesalazine
preparations in patients with ulcerative colitis: comparison of
intraluminal colonic concentrations, serum values, and
urinary excretion. Gut 1990; 31(11): 1271±6.
23 Perneger TV. What's wrong with Bonferroni adjustments [see
comments]. BMJ 1998; 316(7139): 1236±8.
24 Simmonds NJ, Allen RE, Stevens TR, et al. Chemilumines-
cence assay of mucosal reactive oxygen metabolites in
in¯ammatory bowel disease. Gastroenterology 1992; 103(1):
186±96.
25 Hensley K, Robinson KA, Gabbita SP, et al. Reactive oxygen
species, cell signaling, and cell injury. Free Radic Biol Med
2000; 28(10): 1456±62.
Ó 2002 Blackwell Science Ltd, Aliment Pharmacol Ther 16, 197±205
205
26 Bauer V, Bauer F. Reactive oxygen species as mediators of
tissue protection and injury. Gen Physiol Biophys 1999; 18:
7±14.
27 Sauvaire Y, Baissac Y, Leconte O, et al. Steroid saponins from
fenugreek and some of their biological properties. Adv Exp
Med Biol 1996; 405: 37±46.
28 Petit PR, Sauvaire YD, Hillaire-Buys DM, et al. Steroid sapo-
nins from fenugreek seeds: extraction, puri®cation, and
pharmacological investigation on feeding behavior and
plasma cholesterol. Steroids 1995; 60(10): 674±80.
29 Dawidar AM, Saleh AA, Elmotei SL. Steroid sapogenin
constituents of fenugreek seeds. Planta Med 1973; 24(4):
367±70.
30 Wegener T. [Therapy of degenerative diseases of the muscu-
loskeletal system with South African devil's claw (Harpago-
phytum procumbens DC)]. Wien Med Wochenschr 1999;
149(8±10): 254±7.
31 Whitehouse LW, Znamirowska M, Paul CJ. Devil's Claw
(Harpagophytum procumbens): no evidence for anti-in¯amma-
tory activity in the treatment of arthritic disease. Can Med
Assoc J 1983; 129(3): 249±51.
32 Moussard C, Alber D, Toubin MM, et al. A drug used in tra-
ditional medicine, Harpagophytum procumbens: no evidence for
NSAID-like effect on whole blood eicosanoid production in
human. Prostaglandins Leukot Essent Fatty Acids 1992;
46(4): 283±6.
33 Choi HY, Jhun EJ, Lim BO, et al. Application of ¯ow injec-
tionÐchemiluminescence to the study of radical scavenging
activity in plants. Phytother Res 2000; 14(4): 250±3.
34 Dugas AJ Jr, Castaneda-Acosta J, Bonin GC, et al. Evaluation
of the total peroxyl radical-scavenging capacity of ¯avonoids:
structure±activity relationships. J Nat Prod 2000; 63(3):
327±31.
35 Liu Z, Ma LP, Zhou B, et al. Antioxidative effects of green tea
polyphenols on free radical initiated and photosensitized per-
oxidation of human low density lipoprotein. Chem Phys Lipids
2000; 106(1): 53±63.
36 Sawa T, Nakao M, Akaike T, et al. Alkylperoxyl radical-scav-
enging activity of various ¯avonoids and other phenolic
compounds: implications for the anti-tumor-promoter effect of
vegetables. J Agric Food Chem 1999; 47(2): 397±402.
37 Bos MA, Vennat B, Meunier MT, et al. Procyanidins from tor-
mentil: antioxidant properties towards lipoperoxidation and
anti-elastase activity. Biol Pharm Bull 1996; 19(1): 146±8.
38 Vennat B, Bos MA, Pourrat A, et al. Procyanidins from
tormentil: fractionation and study of the anti-radical activity
towards superoxide anion. Biol Pharm Bull 1994; 17(12):
1613±5.