The Role of Fish Oils in the Treatment of Rheumatoid Arthritis

Abstract

Fish oils are a rich source of omega-3 long chain polyunsaturated fatty acids (n-3 LC PUFA). The specific fatty acids, eicosapentaenoic acid and docosahexaenoic acid, are homologues of the n-6 fatty acid, arachidonic acid (AA). This chemistry provides for antagonism by n-3 LC PUFA of AA metabolism to pro-inflammatory and pro-thrombotic n-6 eicosanoids, as well as production of less active n-3 eicosanoids. In addition, n-3 LC PUFA can suppress production of pro-inflammatory cytokines and cartilage degradative enzymes.
In accordance with the biochemical effects, beneficial anti-inflammatory effects of dietary fish oils have been demonstrated in randomised, double-blind, placebo-controlled trials in rheumatoid arthritis (RA). Also, fish oils have protective clinical effects in occlusive cardiovascular disease, for which patients with RA are at increased risk.
Implementation of the clinical use of anti-inflammatory fish oil doses has been poor. Since fish oils do not provide industry with the opportunities for substantial profit associated with patented prescription items, they have not received the marketing inputs that underpin the adoption of usual pharmacotherapies. Accordingly, many prescribers remain ignorant of their biochemistry, therapeutic effects, formulations, principles of application and complementary dietary modifications. Evidence is presented that increased uptake of this approach can be achieved using bulk fish oils. This approach has been used with good compliance in RA patients. In addition, an index of n-3 nutrition can be used to provide helpful feedback messages to patients and to monitor the attainment of target levels.
Collectively, these issues highlight the challenges in advancing the use of fish oil amid the complexities of modern management of RA, with its emphasis on combination chemotherapy applied early.

Full text

Drugs 2003; 63 (9): 845-853
L
EADING
A
RTICLE
0012-6667/03/0009-0845/$33.00/0
 Adis Data Information BV 2003. All rights reserved.
The Role of Fish Oils in the
Treatment of Rheumatoid Arthritis
Leslie G. Cleland, Michael J. James and Susanna M. Proudman
Rheumatology Unit, Royal Adelaide Hospital, Adelaide, South Australia, Australia
Fish oils are a rich source of omega-3 long chain polyunsaturated fatty acids
Abstract
(n-3 LC PUFA). The specific fatty acids, eicosapentaenoic acid and docosahexae-
noic acid, are homologues of the n-6 fatty acid, arachidonic acid (AA). This
chemistry provides for antagonism by n-3 LC PUFA of AA metabolism to
pro-inflammatory and pro-thrombotic n-6 eicosanoids, as well as production of
less active n-3 eicosanoids. In addition, n-3 LC PUFA can suppress production of
pro-inflammatory cytokines and cartilage degradative enzymes.
In accordance with the biochemical effects, beneficial anti-inflammatory
effects of dietary fish oils have been demonstrated in randomised, double-blind,
placebo-controlled trials in rheumatoid arthritis (RA). Also, fish oils have protec-
tive clinical effects in occlusive cardiovascular disease, for which patients with
RA are at increased risk.
Implementation of the clinical use of anti-inflammatory fish oil doses has been
poor. Since fish oils do not provide industry with the opportunities for substantial
profit associated with patented prescription items, they have not received the
marketing inputs that underpin the adoption of usual pharmacotherapies. Accord-
ingly, many prescribers remain ignorant of their biochemistry, therapeutic effects,
formulations, principles of application and complementary dietary modifications.
Evidence is presented that increased uptake of this approach can be achieved
using bulk fish oils. This approach has been used with good compliance in RA
patients. In addition, an index of n-3 nutrition can be used to provide helpful
feedback messages to patients and to monitor the attainment of target levels.
Collectively, these issues highlight the challenges in advancing the use of fish
oil amid the complexities of modern management of RA, with its emphasis on
combination chemotherapy applied early.
Dietary fish oil supplementation has been shown 1. Mechanisms of Action
to reduce symptoms in chronic rheumatoid arthritis
Fish oil contains two biologically important ome-
(RA). In spite of substantial data from clinical trials,
ga-3 long chain polyunsaturated fatty acids (n-3 LC
a compelling mechanistic rationale and absence of
PUFA), eicosapentaenoic acid (20:5 n-3; EPA) and
serious unwanted effects, the application of fish oil
docosahexaenoic acid (22:6 n-3; DHA). Both fatty
treatment in RA is not applied universally and igno-
acids, and EPA in particular, have close homology
rance prevails regarding important aspects of mode
with arachidonic acid (20:4 n-6; AA), with EPA and
of action and practical implementation. AA differing only in the presence or absence of the

846 Cleland et al.
CH
3
COOH
Arachidonic acid
(AA; 20:4 n-6)
COOH
CH
3
Eicosapentaenoic acid
(EPA; 20:5 n-3)
CH
3
COOH
Docosahexaenoic acid
(DHA; 22:5 n-3)
Fi
g
. 1. Structures of lon
g
-chain ome
g
a-6 and ome
g
a-3 fatt
y
acids.
omega-3 (n-3) double bond respectively (figure 1). EPA is converted to an array of products that
differ from their n-6 (AA-derived) counterparts in
Both EPA and DHA are competitor substrates that
the presence of the n-3 double bond. Since the
inhibit oxidation of AA by the cyclooxygenase
subscript designation of eicosanoids, e.g. AA de-
(COX) and lipoxygenase enzymes that are pivotal in
rived prostaglandin (PG)-E
2
, refers to the number of
the production of the C20 oxylipids, known as
double bonds present, the EPA derived eicosanoids
eicosanoids (eicosa means twenty in Greek) [figure
have a subscript designation that is one integer
2]. These mediators are important in regulating vari-
higher (e.g. PGE
3
) than those derived from AA. The
ous homeostatic functions, including the gastric mu-
presence of the additional n-3 double bond can have
cosal integrity, vascular patency, haemostasis and
an important effect on biological activity, as is the
inflammation. case with EPA-derived leukotriene B
5
(LTB
5
),
5LO 5LO
COX COX
Monocyte
TNFα, IL-1β
Production of
MMPs
PGE
2
pain, swelling
PGE
3
pain swelling
TXA
2
TXA
3
Joint
destruction
+
AA
LTB
4
Potent neutrophil
chemotaxin and activator
LTB
5
EPA
LTB
5
Weak neutrophil
activity
Inhibition
PGH
2
PGH
3
Metabolism
Poor metabolism
Fig. 2. Metabolism of omega-6 and omega-3 20-carbon fatty acids. 5LO = 5-lipoxygenase; AA = arachidonic acid; COX = cyclooxygenase-1
or -2; EPA = eicosapentaenoic acid; IL = interleukin; LTB = leukotriene B; MMP = matrix metalloproteinase; PGE = prostaglandin E; PGH =
p
rosta
g
landin H; TNF = tumour necrosis factor; TXA = thromboxane.
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

Fish Oils and Rheumatoid Arthritis 847
which has little of the chemotactic or stimulatory land
[25]
). The benefits of fish oil in RA have been
effect on leucocytes of LTB
4
.
[1]
Similarly, thrombo- endorsed by meta-analysis and mega-analysis.
[26]
xane (TX)-A
3
has little of the activity of TXA
2
as a Importantly, the effects are generally delayed from
vasoconstrictor and stimulus for platelet aggrega- 2–3 months,
[25,26]
as is the case with the disease
tion.
[2]
The relative lack of agonist effect of TXA
3
modifying anti-inflammatory drugs (DMARDs) and
seen in the vascular system seems likely to extend to in contrast to the immediacy of the analgesic action
the up-regulatory effect of TXA on tumour necrosis of NSAIDs. Anti-inflammatory effects have been
factor (TNF)-α and interleukin (IL)-1β synthesis by shown with doses between 2.6 g/day and 7.1 g/
monocytes, since this effect is also TX receptor day,
[25]
but were not seen with a dose of 1.0 g/day.
[23]
mediated.
[3]
The observed inhibitory effect of fish Kremer and co-workers showed that dose of 90 mg/
oil supplementation on TXA
2
synthesis may thus kg/day EPA/DHA (3 : 2 ratio) was associated with a
explain its inhibitory effect on TNFα and IL-1β shorter period to response than 45 mg/kg/day with-
synthesis.
[4]
This effect of fish oil supplements is out little additional benefit otherwise.
[14]
Thus,
especially important, since TNFα and IL-1β up- higher doses might be considered for loading, with
regulated release of collagenase, stromelysin and the lesser doses within the anti-inflammatory range
other enzymes that have been implicated in the for maintenance.
irreversible joint damage that is the hallmark of
RA.
[5]
In contrast, PGE
2
and PGE
3
have similar
3. Strategies for Implementation in
potency in inducing oedema
[6]
and down-regulating
the Clinic
TNFα and IL-1β synthesis by monocytes,
[7]
but
One barrier to the broader use of fish oil in RA is
little, if any, PGE
3
is formed.
[8,9]
the cost of anti-inflammatory doses when purchased
Further studies are required to further under-
at typical prices for a standard gelatin encapsulated
standing of the ways in which n-3 LC PUFA alter
preparation. Pricing for both pharmaceutical and
intracellular signalling and mediator release in mo-
healthcare products appears generally to be pitched
nocytes and other cells resident within or infiltrating
toward market tolerance for usual doses. In the case
synovium and other sites of inflammation. Recent
of fish oil supplements, one to two 1g capsules daily
studies showing the production of novel anti-inflam-
is the usual dose for self-medication. As the most
matory lipids from EPA via COX in the presence of
commonly used preparations contain 300mg n-3 LC
aspirin and nonsteroidal anti-inflammatory drugs
PUFA per gram, this dose falls well short of the
(NSAIDs)
[10]
and from DHA in the presence of
doses of more than 2.5g n-3 LC PUFA that have
aspirin
[11]
provide additional candidate mechanisms
been used in RA studies, which require more than
for the beneficial effects of fish oil in the manage-
eight capsules daily. For higher doses of n-3 LC
ment of anti-inflammatory diseases, such as RA.
PUFA, the cost of fish oil treatment can exceed
substantially that of many other medications, and in
2. Clinical Studies of Fish Oil in
some jurisdictions may experience a further disad-
Rheumatoid Arthritis (RA)
vantage through subsidies for drugs that do not
include fish oil.
To date, there have been at least 13 randomised
controlled trials of fish oil in RA (table I).
[12-24]
All This extra cost can be largely avoided by using
have been undertaken in patients with late disease unencapsulated fish oil taken on juice using the ‘two
with mean disease duration of more than 10 years. glass’ technique. This involves floating 10–15mL
The most consistent benefits have been reduced fish oil on about 30mL of fruit or vegetable juice in a
morning stiffness and decreased tender joint count. small (e.g. sherry or liqueur) glass. This is placed in
In each of three studies, in which NSAIDs have been the mouth to allow the contents to be swallowed
given on an as required basis, use of these agents without contacting the lips, thereby avoiding the fish
was spared by fish oil (reviewed in James and Cle- oil taste. Immediately after swallowing, a further
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

848 Cleland et al.
40mL or so of juice is taken from a separate glass to contents into the duodenum is another strategy for
avoiding ‘repeating’ fish oil taste.
rinse the mouth. If taken before a solid meal without
additional fluid, a repeating taste can also be avoid-
However, there is a need to bring more fish body
ed. Lying on the left side for 10 minutes immed-
oils to market in unencapsulated liquid form, since
iately after ingestion in order to drain the stomach cod liver oils, which traditionally have been taken
Table I. Randomised, controlled trials of fish oil in patients with rheumatoid arthritis
Study n Treatment Medication Omega-3 fat supplement Outcome measures which improved
periods (wks) (g/day) significantly in the fish oil group
Kremer et al.
[12]
38 12 Continued 1.8g EPA, 1.2g DHA No. of tender joints, duration of morning
stiffness
Kremer et al.
[13]
33 14 Continued 2.7g EPA, 1.8g DHA ARA class, physician’s global
assessment, no. of tender joints, no. of
swollen joints, time to fatigue
Cleland et al.
[16]
46 12 Continued 3.2g EPA, 2.0g DHA No. of tender joints, grip strength
Kremer et al.
[14]
49 24 Continued (change was Low dose: 1.7g EPA, No. of swollen joints, no. of tender
a withdrawal criterion) 1.2g DHA or joints, grip strength, physician’s global
high dose: 3.5g EPA, assessment, duration of morning
2.4g DHA
a
stiffness (high dose only)
Tulleken 27 12 Continued 2.0g EPA, 1.3g DHA No. of swollen joints, joint pain index
et al.
[17]
van der Tempel 14 12 Continued 2.0g EPA, 1.3g DHA No. of swollen joints, duration of
et al.
[18]
morning stiffness
Skoldstam 43 24 Continued SAARD. 1.8g EPA, 1.2g DHA Physician global assessment, no. and
et al.
[19]
Free to change NSAID severity of tender joints (Ritchie Index),
decreased NSAID use
Kjeldsen-Kragh 67 16 Continued SAARD. 3.8g EPA, 2.0g DHA No. and severity of tender joints (Ritchie
et al.
[20]
NSAID continued in Index), duration of morning stiffness in
group A and stopped at both groups
10 wks in group B (both
groups received n-3
PUFA)
Nielsen et al.
[21]
51 12 Continued (change was 2.0g EPA, 1.2g DHA Duration of morning stiffness, no. of
a withdrawal criterion) tender joints, C-reactive protein levels
Lau et al.
[22]
64 52 None on SAARD. 1.7g EPA, 1.1g DHA Reduced NSAID use
at entry
b
Change in NSAIDs was
the end-point
Geusens 60 52 Varied as required Low dose: 0.86g Physician pain assessment, patient
et al.
[23]
during study EPA, 0.18g DHA global assessment, and decreased
or high dose: 1.7g NSAID and /or SAARD use (high dose
EPA, 0.36g DHA only)
Kremer et al.
[15]
49 26 or 30 Continued SAARD & 4.6g EPA, 2.5g DHA
a
No. of tender joints, duration of morning
NSAID, but NSAID was stiffness, physician pain assessment,
stopped at 18 or 22 wks physician and patient global assessment
(all at 18 or 22 weeks)
Volker et al.
[24]
26 15 Continued (change was 1.3g EPA, 1.0g DHA
a
Within group: swollen joint count,
a withdrawal criterion). morning stiffness, pain score, physician
Linoleic acid intake was and patient global assessment, HAQ.
<10 g/day Between groups: morning stiffness,
HAQ
a Based on 65kg individuals.
b Variable; numbers had dropped at each 3-monthly assessment over a 15-month period.
ARA = American Rheumatoid Association; DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid; HAQ = health assessment
questionnaire; n = No. of participants for analysis; PUFA = polyunsaturated fatty acids; SAARD = slow acting anti-rheumatic drug.
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

Fish Oils and Rheumatoid Arthritis 849
Omega-6 fats
Linoleic acid
(LA; 18:2)
Arachidonic acid
(AA; 20:4)
Omega-3 fats
α-linolenic acid
(ALA; 18:3)
Eicosapentaenoic
acid (EPA; 20:5)
Sunflower, corn,
soybean,
cottonseed oil
Flaxseed, canola,
soybean oil
Fish, fish oil
Docosahexaenoic
acid (DHA; 22:6)
Fi
g
. 3. Metabolism of 18-carbon fatt
y
acids to lon
g
er chain fatt
y
acids.
unencapsulated, are often the only fluid preparations ble reduction in the pro-inflammatory cytokines
available. While fish liver oils are a good source of IL-1 and TNFα produced by peripheral blood
n-3 LC PUFA (typically 20% w/w), they are also mononuclear cells ex vivo.
[4]
Eating fatty fish and
rich in fat-soluble vitamins and some preparations manufactured foods, which are enriched in fish oil,
can deliver undesirably high levels of vitamin A at also increases tissue n-3 LC PUFA levels and can
doses that deliver anti-inflammatory quantities of reduce mononuclear cell IL-1 synthesis.
[28]
n-3 LC PUFA.
Thus, an optimum dietary enrichment strategy
will involve the diet per se, as well as a fish oil
4. Changes in Diet as a Complement to
supplement. However, variations in application of
Fish Oil Supplements
this multifactorial dietary approach, coupled with
inter-individual putative variability in remodelling
While administration of fish oil supplements pro-
dietary C18 n-3 PUFA to n-3 LC PUFA and/or
vides an effective way to increase tissue levels of
variability in tissue incorporation of dietary n-3 LC
n-3 LC PUFA, selection of foods rich in n-3 PUFA
PUFA, achieves a range of biochemical out-
and avoidance of unnecessary amounts of n-6 PUFA
comes.
[27-29]
For this reason, an index of n-3 nutri-
can achieve or facilitate attainment of this end (fig-
tion is needed for use in the clinic as a guide to the
ure 3).
effectiveness and adherence to advice to increase
In the study of Volker et al., patients with RA
dietary n-3 PUFA intake.
were selected for an intake of n-6 fats in the back-
ground diet of <10 g/day and were given a dose of
5. Plasma Phospholipid EPA as an Index
n-3 LC PUFA of 40 mg/kg/day.
[24]
Trials of fish oil
of Omega-3 Nutrition
in patients with RA have not otherwise addressed
the issue of n-6 fat in the background diet. Cleland, With this discussion in section 4 in mind, we
James and co-workers in a series of investigations in have examined plasma phospholipid EPA as a po-
healthy volunteers, have established that choices of tential index of n-3 nutrition. Our choice of this
visible fats (spreads, cooking oils, dressing and marker was influenced by our observations that
mayonnaise oils), which are rich in n-3 and monoun- plasma phospholipid EPA level has a very close
saturated fats, with avoidance of products that are linear relationship with peripheral blood mononu-
rich in n-6 PUFA, can increase tissue EPA levels clear cell (PBMC) phospholipid EPA (r = 0.97)
[30]
and/or enhance incorporation of n-3 LC PUFA from and is easier to measure. PBMC EPA has a signifi-
fish oil supplements.
[27-29]
Use of visible fats rich in cant inverse curvilinear relationship with TNFα and
n-3 α-linolenic acid (a potential vegetable n-3 pre- IL-1β production by these cells.
[4]
As discussed in
cursor of EPA found in flaxseed and some other section 1, these cytokines have been implicated as
seed oils) causes a modest increase in EPA relative up-regulators of the enzymatic and free-radical me-
to that seen with fish oil, but can achieve a measura- diated destructive events that lead to irreversible
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

850 Cleland et al.
tissue damage in RA and other inflammatory disor- animal studies (see Kang and Leaf
[33]
for review)
ders.
[5]
We observed that, at PBMC EPA levels of and in preliminary human cardiac electrophysiologi-
1.5% total fatty acids or greater, production of these cal studies.
[34]
Other observed cardiovascular bene-
inflammatory cytokines is largely suppressed.
[4]
fits include reduction of raised plasma triglycerides,
This equates to a plasma phospholipid EPA level of a modest reduction in blood pressure, and an amelio-
3.2%, which we have therefore chosen as a notional ration of the effects of atherogenic diets in experi-
indicator of effective dietary n-3 PUFA fortification. mental animals (see O’Keefe and Harris
[35]
for re-
view). These considerations are especially pertinent
We have applied this index in our Early Arthritis
in RA, in which the incidence of cardiovascular
Clinic, which has been established to enable early
events is increased to an extent that reducing cardio-
diagnosis and treatment of RA with combination
vascular events should be seen as component of the
therapy. The treatment regimen involves a combina-
management of RA.
[36]
tion of drugs (methotrexate, sulfasalazine and
hydroxychloroquine in the first instance) coupled
7. Fish Oil and Diet Drug Interactions
with advice to increase n-3 and reduce n-6 PUFA in
the diet and to take a fish oil supplement. At first
There are a number of potentially favourable
assessment few, if any, patients have plasma phos-
interactions between fish oil treatment and chemo-
pholipid EPA levels above the target. At 6 and 12
therapy in RA. These include reduced NSAID use
months approximately 60% (n = 24) had levels
(discussed in section 2), and amelioration of the
above target. Some patients had substantially higher
dose limiting nephrotoxic and hypertensive effects
levels with a third overall exceeding 5% EPA as a
of cyclosporin A.
[37]
There are several anti-rheumat-
proportion of total plasma phospholipid fatty acids.
ic agents with the potential to increase cardiovascu-
At 12 months, biochemically validated continuation
lar risk. These include the highly selective COX-2
rate for dietary n-3 PUFA enrichment was similar to
inhibitors, which can dysregulate TXA
2
/PGI
2
bal-
admitted continuation rates for sulfasalazine and
ance in favour of prothrombotic TXA
2
,
[38]
an effect
somewhat less than that for hydroxychloroquine
likely to be reduced by dietary fortification with n-3
(80%) and methotrexate (96%). Observations have
LC PUFA. The combination of methotrexate and
been taken over a period of up to 3 years and raised
sulfasalazine elevates plasma homocysteine le-
plasma phospholipid EPA levels have been ob-
vels,
[39]
a risk factor for cardiovascular disease. Cor-
served consistently throughout this period in some
ticosteroids increase vascular risk and reduce bone
patients.
[30]
Correlations between levels achieved
density.
[40]
Preliminary data suggest dietary n-3 LC
and long-term outcomes in terms of need for escala-
PUFA may have a favourable effect on bone densi-
tion of chemotherapy and bone and joint structure
ty.
[41]
Finally, supplementation with dietary n-3 LC
and function are the subject of ongoing investiga-
PUFA, by virtue of its ability to inhibit TNFα and
tion.
IL-1β synthesis, is a logical, inexpensive companion
intervention for highly expensive biological agent
6. Fish Oil Supplements and
therapies that have been designed to block these
Cardiovascular Risk in RA
cytokines.
There is now substantial evidence for protective
8. Recent Onset RA
effects of n-3 LC PUFA on the cardiovascular sys-
tem, with the strongest case being for protection Studies of fish oil in RA to date have dealt
against sudden cardiac death after myocardial in- principally with late disease. The use of fish oil
farction
[31]
or as a primary event.
[32]
This effect is remains to be evaluated within the context of best
attributable to a membrane stabilising action of n-3 practice combination treatment of recent onset
LC PUFA on cardiac myocytes that has been shown RA.
[42]
Ideally treatment should be introduced with-
in isolated cardiomyocytes and corroborated in in 3–4 months of onset. The complexity of early
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

Fish Oils and Rheumatoid Arthritis 851
management of RA and assessment of long-term chondrocytes of enzymes implicated in the patho-
outcomes gives impetus to implementation within genesis of OA
[47]
provides ample rationale for a trial.
the trial setting of an index of n-3 nutrition as To date, there are no in vivo data from animal or
discussed in section 5. End points for correlative human studies that address this question. Consider-
analyses should include the need for escalation of ing the emerging basis for broad recommendations
treatment according to predetermined criteria, un- to increase dietary n-3 PUFA intakes for cardiovas-
wanted events, peri-articular bony erosions and peri- cular benefit, the possible positive or negative ef-
articular and more remote bone density, functional fects of fish oil/n-3 LC PUFA on OA need to be
outcomes, including activities of daily living and defined. However, the nature of these assessments is
participation in paid employment. These studies will such that studies need to be designed carefully and
need to be well designed and multicentred. Studies prospectively. The prevalence of OA is sufficient to
that evaluate rules for applying combinations of make an analysis of effects on OA potentially feasi-
agents of known efficacy in series or in parallel are ble within the context of a large long-term interven-
difficult intrinsically, but are crucial to the definition tion study into the effects of fish oil on cardiovascu-
of orderly treatment that achieves best outcomes. In lar outcomes (e.g. primary sudden cardiac death) in
the case of dietary n-3 fats, biochemical effects are elderly individuals.
likely to be more readily established than correla-
tions with clinical effects, although logically one
10. Barriers to Implementation
would expect these to follow.
In seeking to promote the use fish oil supple-
9. Other Applications
ments in the clinical situation where benefit has
been shown, one needs to consider a number of
barriers to implementation. Not least is the perva-
9.1 Inflammatory Diseases
sive influence that drug company marketing has on
Lessons learned in the practice of introducing
the continuing education of physicians. Messages
fish oil in RA should be extended to management
about the health benefits of n-3 LC PUFA must
and trials of other inflammatory diseases. The re-
compete at a disadvantage with the sophisticated
ported favourable influence of fish oil on relapse
and lavishly resourced promotions for patented pre-
rates in Crohn’s disease
[43]
and progression rates in
scription medications. A related problem is the lack
IgA nephropathy
[44]
warrants further attention. The
of instructional skills among physicians with regard
known cardiovascular benefits of n-3 LC PUFA and
to the theoretical and practical aspects of dietary n-3
beneficial effects of fish oil diets on disease mani-
PUFA enrichment. There is also the issue of the
festations and survival in lupus mice
[45]
suggest
opportunity costs for doctors who commit to de-
studies into the influence of fish oil supplements on
tailed explanations about diet. Patients can be
long-term renal outcomes and cardiovascular events
daunted by substantial procurement costs for fish oil
in systemic lupus could be rewarding. The positive
capsules, which may not be covered by pharmaceu-
results of infusions with n-3 LC PUFA in patients
tical subsidy arrangements. Suitable unencapsulated
with psoriasis
[46]
suggest a novel approach to treat-
fish body oil preparations and broader education in
ing psoriatic arthritis, which can be resistant to usual
the ‘two glass’ technique would help resolve this
anti-rheumatic treatments.
problem. Finally, there is the lack of short-term
therapeutic targets for a treatment with delayed ef-
9.2 Osteoarthritis
fects. It is in relation to the latter that the proposed
The possible influence of dietary n-3 LC PUFA index of n-3 nutrition may be especially useful. If
on the occurrence and progression of osteoarthritis validated, a target EPA level, or a related index,
(OA) is an obvious and neglected question. The could prove as useful as other indices used to reduce
inhibitory effects of n-3 LC PUFA on release by long-term risks for unwanted events associated with
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

852 Cleland et al.
high pressure liquid chromatography with fluorometric detec-
chronic disordered homeostasis. In this regard there
tion. Prostaglandins 1991; 42: 355-68
are parallels with now standard medical measures
9. Knapp HR, FitzGerald GA. The antihypertensive effects of fish
oil: a controlled study of polyunsaturated fatty acid supple-
such as blood pressure, blood lipids (cholesterol and
ments in essential hypertension. N Engl J Med 1989; 320:
triglycerides) and glycosylated haemoglobin.
1037-43
10. Serhan CN, Clish CB, Brannon J, et al. Novel functional sets of
11. Conclusion
lipid-derived mediators with antiinflammatory actions gener-
ated from omega-3 fatty acids via cyclooxygenase 2-nonster-
oidal antiinflammatory drugs and transcellular processing. J
There exist high level standards of evidence for
Exp Med 2000; 192: 1197-204
clinically beneficial effects of dietary fish oil in
11. Serhan CN, Hong S, Gronert K, et al. Resolvins: a family of
inflammatory disease, particularly RA. The barriers
bioactive products of omega-3 fatty acid transformation cir-
cuits initiated by aspirin treatment that counter proinflamma-
to uptake of this approach in the standard treatment
tion signals. J Exp Med 2002; 196: 1025-37
of RA are not difficult to discern and they involve a
12. Kremer JM, Bigauoette J, Michalek AV, et al. Effects of
mix of issues ranging from a marketing vacuum due
manipulation of dietary fatty acids on clinical manifestations
of rheumatoid arthritis. Lancet 1985; I: 184-7
to a lack of patentability of fish oil (and fish) to
13. Kremer JM, Jubiz W, Michalek A, et al. Fish-oil fatty acid
physical and financial barriers in ingesting daily
supplementation in active rheumatoid arthritis. Ann Intern
anti-inflammatory doses of fish oil. However, use of
Med 1987; 106: 497-503
14. Kremer JM, Lawrence DA, Jubiz W, et al. Dietary fish oil and
bulk fish oils can overcome the latter two issues. In
olive oil supplementation in patients with rheumatoid arthritis.
addition, use of an index of n-3 nutrition can provide
Arthritis Rheum 1990; 33: 810-20
helpful feedback to both patients and clinicians on
15. Kremer JM, Lawrence DA, Petrillo GF, et al. Effects of high-
dose fish oil on rheumatoid arthritis after stopping nonsteroidal
the success of the intervention in attaining target
antiinflammatory drugs. Arthritis Rheum 1995; 38: 1107-14
levels.
16. Cleland LG, French JK, Betts WH, et al. Clinical and biochemi-
cal effects of dietary fish oil supplements in rheumatoid arthri-
Acknowledgements
tis. J Rheumatol 1988; 15: 1471-5
17. Tulleken JE, Limburg PC, Muskiet FAJ, et al. Vitamin E status
Funding from NHMRC is acknowledged. The authors
during dietary fish oil supplementation in rheumatoid arthritis.
have no conflicts of interest that are directly relevant to the
Arthritis Rheum 1990; 33: 1416-9
content of this manuscript.
18. van der Tempel H, Tulleken JE, Limburg PC, et al. Effects of
fish oil supplementation in rheumatoid arthritis. Ann Rheum
Dis 1990; 49: 76-80
References
19. Skoldstam L, Borjesson O, Kjallman A, et al. Effect of six
1. Goldman DW, Pickett WC, Goetzl EJ. Human neutrophil chem-
months of fish oil supplementation in stable rheumatoid arthri-
otactic and degranulating activities of leukotriene B
5
(LTB
5
)
tis: a double blind, controlled study. Scand J Rheumatol 1992;
derived from eicosapentaenoic acid. Biochem Biophys Res
21: 178-85
Commun 1983; 117: 282-8
20. Kjeldsen-Kragh J, Lund JA, Riise T, et al. Dietary omega-3
2. Moncada S, Vane JR. The role of prostacyclin in vascular tissue.
fatty acid supplementation and naproxen treatment in patients
Fed Proc 1979; 38: 66-71
with rheumatoid arthritis. J Rheumatol 1992; 19: 1531-6
3. Caughey GE, Pouliot M, Cleland LG, et al. Regulation of tumor
21. Nielsen GL, Faarvang KL, Thomsen BS, et al. The effects of
necrosis factor alpha and interleukin-1 beta synthesis by
dietary supplementation with n-3 polyunsaturated fatty acids
thromboxane A
2
in non-adherent human monocytes. J
in patients with rheumatoid arthritis: a randomized double
Immunol 1997; 158: 351-8
blind trial. Eur J Clin Invest 1992; 22: 687-91
4. Caughey GE, Mantzioris E, Gibson RA, et al. The effect on
22. Lau CS, Morley KD, Belch JJF. Effects of fish oil supplementa-
human tumor necrosis factor α and interleukin-1β production
tion on non-steroidal anti-inflammatory drug requirement in
of diets enriched in n-3 fatty acids from vegetable oil or fish
patients with mild rheumatoid arthritis: a double blind placebo
oil. Am J Clin Nutr 1996; 63: 116-22
controlled study. Br J Rheumatol 1993; 32: 982-9
5. Westacott CI, Sharif M. Cytokines in osteoarthritis: mediators
23. Geusens P, Wouters C, Nijs J, et al. Long-term effect of
or markers of joint destruction? Semin Arthritis Rheum 1996;
omega-3 fatty acid supplementation in active rheumatoid ar-
25: 254-72
thritis. Arthritis Rheum 1994; 37: 824-9
6. Hawkes JS, James MJ, Cleland LG. Biological activity of pros-
24. Volker D, Fitzgerald P, Major G, et al. Efficacy of fish oil
taglandin E
3
with regard to oedema formation in mice. Agents
concentrate in the treatment of rheumatoid arthritis. J Rheu-
Actions 1992; 35: 85-7
matol 2000; 27: 2343-6
7. Dooper MM, Wassink L, M’Rabet L, et al. The modulatory
25. James MJ, Cleland LG. Dietary n-3 fatty acids and therapy for
effects of prostaglandin-E on cytokine production by human
rheumatoid arthritis. Semin Arthritis Rheum 1997; 27: 85-97
peripheral blood mononuclear cells are independent of the
prostaglandin subtype. Immunology 2002; 107: 152-9
26. Fortin PR, Lew RA, Liang MH, et al. Validation of a meta-
8. Hawkes JS, James MJ, Cleland LG. Separation and quantifica- analysis: the effects of fish oil in rheumatoid arthritis. J Clin
tion of PGE
3
following derivatization with panacyl bromide by Epidemiol 1995; 48: 1379-90
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)

Fish Oils and Rheumatoid Arthritis 853
27. Cleland LG, James MJ, Neumann MA, et al. Linoleate inhibits 39. Haagsma CJ, Blom HJ, van Riel PL, et al. Influence of
EPA incorporation from dietary fish oil supplements in human
sulphasalazine, methotrexate, and the combination of both on
subjects. Am J Clin Nutr 1992; 55: 395-9
plasma homocysteine concentrations in patients with rheuma-
toid arthritis. Ann Rheum Dis 1999; 58: 79-84
28. Mantzioris E, Cleland LG, Gibson RA, et al. Biochemical
effects of a diet containing foods enriched with n-3 fatty acids.
40. Frauman AG. An overview of the adverse reactions to adrenal
Am J Clin Nutr 2000; 72: 42-8
corticosteroids. Adverse Drug React Toxicol Rev 1996; 15:
29. Mantzioris E, James MJ, Gibson RA, et al. Dietary substitution
203-6
with an α-linolenic acid-rich vegetable oil increases
41. Kruger MC, Coetzer H, de Winter R, et al. Calcium, gamma-
eicosapentaenoic acid concentrations in tissues. Am J Clin
linolenic acid and eicosapentaenoic acid supplementation in
Nutr 1994; 59: 1304-9
senile osteoporosis. Aging (Milano) 1998; 10: 385-94
30. Cleland LG, Proudman SM, Hall C, et al. A biomarker of n-3
status in patients taking fish oil for inflammatory diseases.
42. Emery P, Breedveld FC, Dougados M, et al. Early referral
Lipids. In press
recommendation for newly diagnosed rheumatoid arthritis:
evidence based development of a clinical guide. Ann Rheum
31. Marchioli R, Barzi F, Bomba E, et al. Early protection against
Dis 2002; 61: 290-7
sudden death by n-3 polyunsaturated fatty acids after myocar-
dial infarction: time-course analysis of the results of the Grup-
43. Belluzzi A, Brignola C, Campieri M, et al. Effect of an enteric-
po Italiano per lo Studio della Sopravvivenza nell’Infarto
coated fish-oil preparation on relapses in Crohn’s disease. N
Miocardico (GISSI)-Prevenzione. Circulation 2002; 105:
Engl J Med 1996; 334: 1557-60
1897-903
44. Donadio JV. The emerging role of omega-3 polyunsaturated
32. Albert CM, Campos H, Stampfer MJ, et al. Blood levels of
fatty acids in the management of patients with IgA nephropa-
long-chain n-3 fatty acids and the risk of sudden death. N Engl
thy. J Ren Nutr 2001; 11: 122-8
J Med 2002; 346: 1113-8
33. Kang JX, Leaf A. Antiarrhythmic effects of polyunsaturated
45. Robinson DR, Prickett JD, Makoul GT, et al. Dietary fish oil
fatty acids: recent studies. Circulation 1996; 94: 1774-80
reduces progression of established renal disease in (NZB X
NZW)F1 mice and delays renal disease in BXSB and MRL/1
34. Metcalf RG, James MJ, Cowie R, et al. Effect of fish oil on
strains. Arthritis Rheum 1986; 29: 539-46
electrically induced ventricular tachycardia in high risk cardiac
patients [abstract]. Australian Society for Medical Research
46. Mayser P, Mrowietz U, Arenberger P, et al. Omega-3 fatty acid-
(SA). Annual Scientific Meeting; 2002 May 30; Adelaide
based lipid infusion in patients with chronic plaque psoriasis:
35. O’Keefe Jr JH, Harris WS. From Inuit to implementation:
results of a double-blind, randomized, placebo-controlled,
omega-3 fatty acids come of age. Mayo Clin Proc 2000; 75:
multicenter trial. J Am Acad Dermatol 1998; 38: 539-47
607-14
47. Curtis CL, Rees SG, Little CB, et al. Pathologic indicators of
36. Bacon PA, Townend JN. Nail in the coffin: increasing evidence
degradation and inflammation in human osteoarthritic carti-
for the role of rheumatic disease in the cardiovascular mor-
lage are abrogated by exposure to n-3 fatty acids. Arthritis
tality of rheumatoid arthritis. Arthritis Rheum 2001; 44:
Rheum 2002; 46: 1544-53
2707-10
37. Darlametsos IE, Varonos DD. Role of prostanoids and endothe-
lins in the prevention of cyclosporine-induced nephrotoxicity.
Correspondence and offprints: Dr Leslie G. Cleland, Rheu-
Prostaglandins Leukot Essent Fatty Acids 2001; 64: 231-9
matology Unit, Royal Adelaide Hospital, North Terrace,
38. Cleland LG, James MJ, Stamp LK, et al. COX-2 inhibition and
Adelaide, SA 5000, Australia.
thrombotic tendency: a need for surveillance. Med J Aust
2001; 175: 214-7
E-mail: Lcleland@mail.rah.sa.gov.au
 Adis Data Information BV 2003. All rights reserved. Drugs 2003; 63 (9)