Effects of Moderate-Dose Omega-3 Fish Oil on Cardiovascular Risk Factors and Mood After Ischemic Stroke
A Randomized, Controlled Trial
Background and Purpose— Fish-derived omega-3 fatty acids have long been associated with cardiovascular protection. In this trial, we assessed whether treatment with a guideline-recommended moderate-dose fish oil supplement could improve cardiovascular biomarkers, mood- and health-related quality of life in patients with ischemic stroke.
Methods— Patients with CT-confirmed stroke were randomized to 3 g/day encapsulated fish oil containing approximately 1.2 g total omega-3 (0.7 g docosahexaenoic acid; 0.3 g eicosapentaenoic acid) or placebo oil (combination palm and soy) taken daily over 12 weeks. Serum triglycerides, total cholesterol and associated lipoproteins, selected inflammatory and hemostatic markers, mood, and health-related quality of life were assessed at baseline and follow-up. The primary outcome was change in triglycerides. Compliance was assessed by capsule count and serum phospholipid omega-3 levels (Australian Clinical Trials Registration: ACTRN12605000207617).
Results— One hundred two patients were randomized to fish oil or placebo. Intention-to-treat and per-protocol (>85% compliance) analyses showed no significant effect of fish oil treatment on any lipid, inflammatory, hemostatic, or composite mood parameters measured. Adherence to treatment based on pill count was good (89%) reflected by increased serum docosahexanoic acid (P<0.001) and eicosapentaenoic acid (P=0.0006) in the fish oil group. Analysis of oil composition, however, showed some degradation and potentially adverse oxidation products at the end of the study.
Conclusions— There was no effect of 12 weeks of treatment with moderate-dose fish oil supplements on cardiovascular biomarkers or mood in patients with ischemic stroke. It is possible that insufficient dose, short duration of treatment, and/or oxidation of the fish oils may have influenced these outcomes.
Some controversy surrounds cardiovascular protection and encapsulated omega-3 polyunsaturated fatty acids (PUFA) derived from fish.1,2⇓ Although long chain PUFAs docosahexaenoic acid (DHA, C22:6n−3) and eicosapentaenoic acid (EPA, C20:5n−3) have long been associated with decreased cardiovascular disease,3,4⇓ not all trials have been supportive of clinical improvement.1 Trials such as Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico (GISSI)- Prevenzione,4 the largest prospective omega-3 trial to date, and the first Diet and Reinfarction Trial (DART)5 showed benefits after myocardial infarction as did the recent GISSI-HF.6 However, the second DART trial of >3000 men with stable angina showed an unexpected increase in cardiac deaths after advice to supplement with encapsulated oils, albeit lessened with oily fish.7
Most studies of fish oils have involved patients with coronary artery disease. Less is known of effects in patients with stroke in which epidemiology shows modest beneficial effects of fish/fish oil capsules, more commonly for prevention of ischemic rather than hemorrhagic stroke,8–10⇓⇓ but in which the evidence is inconclusive.11,12⇓ In ischemic stroke atherosclerosis is a common underlying etiology, and adverse lipid profile and arterial inflammation are important risk factors for recurrent events. Total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) are important determinants of atherosclerotic plaque formation, and even moderately raised triglycerides may increase cardiovascular risk. Low-density lipoprotein particle size (LDLps) is also a key factor in the genesis of atheroma13 with small, dense LDL particles having a greater rate of flux into the arterial wall and greater susceptibility to oxidation.14 Triglycerides play a central role in regulating LDL subfractions and can lead to catabolism of protective high-density lipoprotein cholesterol (HDL-C) and generation of small, dense, highly atherogenic LDL particles. Fish oils may decrease triglycerides and redistribute LDLps toward larger, less dense particles characteristic of the low cardiovascular risk “pattern A” phenotype,13,15⇓ although they may also adversely increase concentrations of LDL-C.16 Other putative mechanisms for fish oils include antiarrhythmic, anticoagulant, and anti-inflammatory effects17,18⇓; antiatherosclerotic effects on triglycerides and small dense LDL-C13,19–21⇓⇓⇓; and blood pressure-lowering.22 Hence, fish oil supplementation is recommended in high-risk patients and is popularly used across the general community for broader health benefits.
There have been no large intervention trials investigating effects of fish oils in patients with stroke. An early pilot study showed no effect of fish oils on cholesterol fractions or platelet function.23 The aim of the Fish Oils in Stroke (FOILS) study was to evaluate cardiovascular effects of moderate-dose (1 g DHA+EPA) fish oil supplementation as recommended by American Heart Association guidelines for patients with established coronary artery disease.3 Indicators of mood and quality of life were also assessed given that stroke-associated depression is common and fish oils have been associated with improvement in mood.24
The FOILS study was a randomized, double-blind trial to investigate the effect of moderate-dose (3 g/day) encapsulated fish oils on cardiovascular risk and mood in patients with ischemic stroke. The trial was undertaken at a single hospital site in Auckland, New Zealand, from July 2004 to March 2006 with the first and last patients randomized on July 29, 2004, and December 23, 2005, respectively. Treatment was a 12-week intervention with encapsulated fish oils (approximately 1.2 g/day omega-3 PUFA) or matching placebo (palm and soy oils25–27⇓⇓). The randomization was a blocked (varying block sizes), unstratified design sequence. It was implemented using treatment packs blinded for treatment with sequential allocation of packs to patients after confirmation of inclusion criteria and was generated using an automated Internet-based system held at the Clinical Trials Research Unit, University of Auckland. Participants were dispensed capsules with instructions to take 3×1-g capsules each morning with food, to abstain from other fish oil supplements, and maintain their habitual diet. Outcomes at baseline and 12-week follow-up included anthropometry, hematology, biochemistry, blood pressure, questionnaires to assess health-related quality of life and mood, compliance, and adverse events (AEs). Blood samples were collected fasted. Compliance was assessed by capsule count and through analysis of serum phospholipid fatty acid methyl esters (FAMEs). Telephone contact was made at 6 and 11 weeks. The trial was approved by the Auckland Human Ethics Committee X, Auckland, New Zealand, and all participants provided written informed consent.
Participants were aged >45 years, clinically stable, and with a history of CT-confirmed first-ever or recurrent ischemic stroke of probable noncardioembolic etiology >3 months before registration. Exclusion criteria included intolerance/hypersensitivity to fish/fish oils, current use of fish oil supplements, malabsorptive bowel diseases, or participation in a concurrent clinical trial. Stroke was defined using World Health Organization criteria of “rapidly developing clinical signs of focal (at times global) disturbance of cerebral function lasting more than 24 hours with no apparent cause other than of vascular origin.”28 Recruitment was predominantly by hospital admissions with one patient recruited from newspaper advertisements.
Fish Oil and Placebo Supplements
The 3-g/day fish oil supplement contained approximately 1.2 g/day total omega-3 PUFA. Table 1 outlines the fatty acid composition of the fish oil and placebo. The fish oil was made from processed Hoki liver oil, a southern ocean fish found in waters surrounding Australia and New Zealand. Random samples of fish oil and placebo capsules were stored at ambient temperature and later analyzed for fatty acid profile to confirm randomization.
The primary outcome was change in serum triglycerides between baseline and 12 weeks. Secondary outcomes included parameters of lipid profile TC, LDL-C, LDLps, HDL-C; inflammatory markers, including high-sensitivity C-reactive protein, liver function tests, white cell count, erythrocyte sedimentation rate, and ferritin; fibrinogen; and blood pressure. Mood was assessed using a 28-item General Health Questionnaire,29 and mental and physical health-related quality of life were assessed using the 36-item Short Form questionnaire (SF-36).30
The fatty acid composition of the fish oil was verified using gas chromatography methodology after generation of fatty acid esters (MIRINZ Centre Analytic Laboratory, Hamilton, New Zealand). Blood lipids, inflammatory, and hemostatic markers were analyzed by a diagnostic laboratory (Lab Plus, Auckland, New Zealand). LDLps was measured by Skylight Biotech Inc (Akita, Japan) using a high-sensitivity duel-detection lipoprotein-profiling high-performance liquid chromatography separation system.31 High-performance liquid chromatography with gel permeation columns was used as an alternative method to ultracentrifugation for classification and quantification of lipoproteins on the basis of particle size. The method quantified 20 lipoprotein subclasses from both cholesterol and triglyceride components within the major classes of chylomicrons (>80 nm particle size), very-low-density lipoprotein (VLDL) (30 to 80 nm), LDL (18 to 30 nm), and HDL (8 to 18 nm) fractions. Subclasses of relevance were large LDL, 28.6 nm; medium LDL, 25.5 nm; small LDL, 23.0 nm; very small LDL1, 20.7 nm; very small LDL2, 18.6 nm; and very small LDL3, 16.7 nm.
FAME analyses of serum phospholipids provided a qualitative assessment of treatment compliance and were extracted and analyzed using standard methods.32
Initial sample size calculations indicated that 220 participants would be required for 90% power to detect a 15% difference in change in serum triglycerides between treatment and placebo groups. However, on the basis of a blinded interim analysis during data safety review of AEs of the first 50 randomized participants, lower than predicted variance in the primary outcome serum triglycerides led to 102 participants randomized into the trial. Data were analyzed using intention-to-treat in which no dropouts were replaced and all randomized participants were included in analyses. Missing at follow-up data were imputed using last value carried forward. Analysis of covariance (ANCOVA) was used to deal with potentially unbalanced baseline data. A blinded analysis prereview was performed to determine required adjustments for potential confounders and to establish rules for the per-protocol analysis. Probability values as presented in data tables were calculated from ANCOVA for the difference between fish oil and placebo treatment over time adjusted for 3 factors: baseline measurement, gender, and age. The blinded prereview included examination of adjustments for factors such as concurrent medications and years since ischemic stroke performed as exploratory or sensitivity analyses to ensure the robustness of estimates in the primary analysis and for which all were shown to have no effect on the trial outcomes. A per-protocol analysis was also performed based on >85% compliance. Distribution of continuous end points was assessed for normality; however, there was no requirement for transformation. No adjustments for multiplicity were carried out for secondary end points, AEs, or other end points. P<0.05 level of statistical significance was used. Analyses were carried out using SAS Version 9.1 (SAS Institute Inc, Cary, NC) and SPLUS (Insightful Corp, Seattle, Wash). Data are mean±SD.
One hundred two participants were randomized into the trial with 51 participants allocated to each treatment group (Figure). Seven participants withdrew consent or were lost to follow-up at 12 weeks and hence 95 participants completed the intervention. There were 42 AEs (fish oil: n=22; placebo: n=20) of which 6 were serious AEs (fish oil: n=4; placebo: n=2), including one participant randomized to placebo who died as a result of an acute myocardial infarction. Serious AEs in the fish oil group were chest pain, palpitations, congestive heart failure, and elective hip replacement, none of which were assessed as related to treatment by the principle investigator.
Participants were predominantly overweight, older men (n=72) and women (n=30) with central adiposity with an ischemic stroke approximately 1 year previously. Across treatment groups, participants had similar serum lipids, including triglycerides of 1.6 mmol/L, and other cardiovascular biomarkers (Table 2). In both groups, approximately 80% (n=40, n=38) of participants were on therapies for serum lipid-lowering, including statins, 60% to 70% (n=30, n=37) on blood pressure-lowering therapies, and 8% to 16% (n=8, n=4) on oral hypoglycemic agents and/or insulin for diabetes mellitus. Body weight was stable over 12 weeks.
Compliance to treatment as assessed by return capsule count at 12 weeks was approximately 90%. This was reflected in a significant increase in circulating omega-3 PUFA over 12 weeks with higher DHA (P<0.001) and EPA (P=0.0006) in the fish oil group compared with placebo (Table 3).
Serum Markers of Cardiovascular Disease
There were no significant effects of treatment on serum lipids, including the primary outcome variable serum triglyceride, or on inflammatory or hemostatic markers (Table 4) or indicators of mood (Table 5). It is notable that there was no decrease in serum triglycerides in this trial; rather, there was an unexpected increase on fish oil treatment, albeit nonsignificant, and a decrease on placebo of +7% and −3%, respectively. The absence of triglyceride-lowering was confirmed on per-protocol (>85% compliance to treatment) analysis and with the exclusion of one participant with an outlying extreme triglyceride value (9.0 mmol/L) in the fish oil group at 12 weeks, which normalized after the end of the trial. Also notable alongside the lack of triglyceride-lowering was an absence of AEs on LDL-C. Table 4 also shows mean LDLps to be unchanged from baseline as were circulating concentrations of large, medium, small, or very small LDL-C. Blood pressure was also unaffected by fish oil treatment.
The fish oil treatment had no effect on health-related quality of life. The 28-item General Health Questionnaire total score was significantly different between treatments, a consequence of a decrease on placebo (fish oil: −0.8 units, placebo: −2.3 units, P=0.04; Table 5). Similarly, of the 28-item General Health Questionnaire components, only social dysfunction differed between treatments (fish oil: −0.8 units, placebo: −2.0 units; P=0.027) with no effect on somatic symptoms, anxiety, insomnia, or depression. It is unlikely that these small changes in mood scores are of clinical significance.
In this trial of patients with ischemic stroke, there was no evidence that treatment with moderate-dose, encapsulated fish oils over 12 weeks improved cardiovascular markers, including triglycerides and LDLps, inflammatory and hemostatic parameters, blood pressure, health-related quality of life, or mood. The absence of triglyceride-lowering was particularly unexpected because this cardiovascular biomarker has the strongest evidence base as summarized in a recent systematic review, which concluded that most studies assessing triglycerides have shown a decrease of at least 15% with fish oils.19 It is notable, however, that Balk and colleagues19 concluded that supplementation with fish oils is associated with lower rates of cardiovascular mortality and sudden death but not stroke. Similarly, GISSI-HF has recently shown lower rates of cardiovascular death, sudden cardiac death, myocardial infarction, and hospital admission but not stroke in n-3 PUFA-treated patients.6 Although evidence of cardiovascular protective effects of fish and/or fish oil capsules in patients at risk of ischemic stroke has been shown in some epidemiological studies,8–10⇓⇓ not all are confirmatory.11,12⇓ Skerrett and Hennekens33 showed consistency of evidence between consumption of fish/fish oils across ecological, cross-sectional, and case control studies but inconclusive in prospective cohort studies. No consistent relationship between fish consumption and stroke was again shown in a recent European cohort study.12 Although some inconsistencies may be due to differences in study design, it has been suggested that different types of fish consumed and/or methods of preparation (eg, baking, frying) may play a role12 as may consumption of encapsulated oils rather than dietary fish. A randomized, placebo-controlled intervention is the most reliable way to assess effects of fish oils.
There are several possible explanations for lack of efficacy in our study. First, 1 g/day DHA+EPA may have been too low a dose, although similar doses were used in GISSI-Prevenzione4 and GISSI-HF6 in which low–moderate-dose fish oil was cardioprotective and led to guidelines recommending approximately 1 g/day DHA+EPA for patients with documented cardiovascular disease.3 Certainly trials have shown triglyceride-lowering with modest dose fish oil,27,34⇓ although higher doses may result in greater effects.19 The proportion of DHA:EPA may also be important, but although they cannot be regarded as equivalent, there is no consensus over the optimal ratio.21,35–37⇓⇓⇓ We cannot attribute lack of effect to poor adherence, which was good when estimated from capsule counts and serum phospholipids. Other potential explanations include the relatively short 12-week intervention period and the moderately elevated baseline triglycerides. Higher baseline levels may be associated with greater triglyceride-lowering.19 Approximately 80% of participants in FOILS were prescribed statins, which lower TC by inhibiting HMG-CoA reductase, stimulating LDL receptors, and increasing clearance of LDL-C from circulation. Statins, however, have less effect than fibrates or niacin in decreasing triglycerides, and it has been shown that 3-month treatment with DHA-rich fish oil can improve lipids in patients taking statins.38 Another factor is sample size, which may have been too small to detect more modest beneficial effects of fish oils, which, when applied widely across whole populations, may have important public health implications but have arguably questionable benefits in the management of individual patients.
It is possible that lack of efficacy is a true effect. There are clear differences between increasing fish consumption in which fish replaces other animal proteins, and supplementation with fish oil capsules in which omega-3 PUFAs are added without the parallel benefit of decreased saturated fats. Much of the epidemiology showing protective benefits in patients with stroke has identified relationships between fish and morbidity/mortality, only sometimes accounting also for encapsulated oils. Another factor with encapsulated oils is the propensity for oxidation and rancidity over time.39 Fish oils are easily oxidized during processing and after encapsulation. High numbers of double bonds render DHA and EPA more susceptible to lipid oxidation than other fatty acids with rapid formation of hydroperoxides (HPODE), the initial degradation product in free radical-catalyzed lipid oxidation. Whether oxidized fish oils are less efficacious for cardioprotection is not known, but there is evidence of adverse effects on lipid and chylomicron metabolism from circulating oxidized products such as conjugated dienes (a marker of lipid hydroperoxides) and adverse changes in hemostatic clotting and endothelium-dependent vascular dilatation. It is possible, therefore, that oxidation of fish oil during the study may have influenced the results.
This study is one of few clinical trials to investigate fish oil supplementation after ischemic stroke. Contrary to other studies in different cardiovascular risk populations, there was no evidence that a moderate-dose encapsulated fish oil improved cardiovascular biomarkers in this high-risk group. Insufficient dose and oxidation of encapsulated oils are potential explanations for lack of cardiovascular improvement. Until results of large-scale studies of fish oil supplementation in poststroke populations such as the ongoing SU.FOL.OM3 (SUpplementation with FOlate, vitamin B6 and B12 and/or OMega-3 fatty acids) trial40 are available, these factors may need to be considered when formulating guidelines and public health policy regarding health benefits of fish oil supplementation in patients with ischemic stroke.
Appendix of Investigators
Sally D. Poppitt, PhD, fundraiser, trial steering committee, trial design and interpretation, senior author; Colin A. Howe, PhD, trial steering committee member, trial manager, trial design and interpretation, coauthor; Fiona E. Lithander, PhD, trial steering committee member, trial design and interpretation, coauthor; Karen M. Silvers, PhD, trial steering committee member, trial design and interpretation, coauthor; Ruey-Bin Lin, MSc, biostatistician; John Croft, trial steering committee member, provision of treatment (n-3 PUFA, placebo) oils; Yogini Ratnasabapathy, MBBS, FRACP, patient recruitment, clinical oversight. Robert A. Gibson, PhD, head, analytic laboratory; Craig S. Anderson, MBBS, FRACP, PhD, fundraiser, trial steering committee chair, trial design and interpretation, coauthor.
We thank Dr Ajay Kumar, Mrs Helen Walter, and Ms Francie Birch for the clinical aspects of the study; Roxanne Portolesi for the FAME analyses; and Michele Barlow, Simon Pink, and John Fa’atui for database management. We also thank the participants of this intervention trial.
Sources of Funding
The Health Research Council of New Zealand provided funding for this intervention trial. The Maurice and Phyllis Paykel Trust, New Zealand, funded the LDL particle size and FAME analyses. Sea Dragon, New Zealand, provided the fish oil and Nutrition Laboratories, New Zealand, provided the placebo treatment and encapsulated the oils.
J.C. has a consultancy relationship with Nutrition Labs New Zealand.
- Received April 8, 2009.
- Revision received June 11, 2009.
- Accepted July 7, 2009.
- ↵Hooper L, Thompson RL, Harrison RA, Summerbell CD, Ness AR, Moore HJ, Worthington HV, Durrington PN, Higgins JP, Capps NE, Riermersma RA, Ebrahim SB, Smith GD. Risks and benefits of omega 3 fats for mortality, cardiovascular disease and cancer: a systematic review. BMJ. 2006; 332: 752–760.
- ↵Brunner E. Oily fish and omega 3 fat supplements [Editorial]. BMJ. 2006; 332: 739–740.
- ↵Kris-Etherton PM, Harris WS, Appel LJ. Fish consumption, fish oil, omega-3 fatty acids, and cardiovascular disease. Circulation. 2002; 106: 2747–2757.
- ↵GISSI-Prevenzione Investigators. Early protection against sudden death by n-3 polyunsaturated fatty acids after myocardial infarction: time-course analysis of the results of the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico (GISSI)-Prevenzione. Circulation. 2002; 105: 1897–1903.
- ↵Orencia AJ, Daviglus ML, Dyer AR, Shekelle RB, Stamler J. Fish consumption and stroke in men. 30-year findings of the Chicago Western Electric Study. Stroke. 1996; 27: 204–209.
- ↵Layne KS, Goh YK, Jumpsen JA, Ryan EA, Chow P, Clandinin MT. Normal subjects consuming physiological levels of 18: 3(n-3) and 20:5(n-3) from flaxseed or fish oils have characteristic differences in plasma lipid and lipoprotein fatty acid levels. J Nutr. 1996; 126: 2130–2140.
- ↵Leaf A, Albert CM, Josephson M, Steinhaus D, Kluger J, Kang JX, Cox B, Zhang H, Schoenfeld D; Fatty Acid Antiarrhythmia Trial Investigators. Prevention of fatal arrhythmias in high-risk subjects by fish oil n-3 fatty acid intake. Circulation. 2005; 112: 2762–2768.
- ↵Green D, Barreres L, Borensztajn J, Kaplan P, Reddy MN, Rovner R, Simon H. A double-blind, placebo-controlled trial of fish oil concentrate (maxEPA) in stroke patients. Stroke. 1985; 16: 706–709.
- ↵Caslake MJ, Miles EA, Kofler BM, Lietz G, Curtis P, Armah CK, Kimber AC, Grew JP, Farrell L, Stannard J, Napper FL, Sala-Vila A, West AL, Mathers JC, Packard C, Williams CM, Calder PC, Minihane AM. Effect of sex and genotype on cardiovascular biomarker response to fish oils: the FINGEN study. Am J Clin Nutr. 2008; 88: 618–629.
- ↵World Health Organization. International Classification of Functioning, Disability and Health. Geneva; World Health Organization; 2001.
- ↵Goldberg DP, Williams P. A User’s Guide to the General Health Questionnaire. Basingstoke: National Foundation for Educational Research (NFER)-Nelson, UK. 1988.
- ↵Ware JE, Kosinski M, Keller SD. SF-36 Physical and Mental Health Summary Scales: A User’s Manual. Boston; MA: The Health Institute, New England Medical Center; 1994.
- ↵Usui S, Hara Y, Hosaki S, Okazaki M. A new on-line dual enzymatic method for simultaneous quantification of cholesterol and triglycerides in lipoproteins by HPLC. J Lipid Res. 2002; 43: 805–814.
- ↵Schwellenbach LJ, Olson KL, McConnell KJ, Stolcpart RS, Nash JD, Merenich JA. The triglyceride-lowering effects of a modest dose of docosahexaenoic acid alone versus in combination with low dose eicosapentaenoic acid in patients with coronary artery disease and elevated triglycerides. J Am Coll Nutr. 2006; 25: 480–485.
- ↵Mori TA, Bao D, Burke V. Purified EPA and DHA have differential effects on serum lipids and lipoproteins, LDL particle size, glucose and insulin, in mildly hyperlipidemic men. Am J Clin Nutr. 2000; 71: 1085–1094.
- ↵Woodman RJ, Mori TA, Burke V, Puddey IB, Watts GF, Best JD, Beilin LJ. Docosahexanoic acid but not eicosapentaenoic acid increases LDL particle size in treated hypertensive type 2 diabetic patients. Diabetes Care. 2003; 26: 253.
- ↵Galan P, de Bree A, Mennen L, Potier de Courcy G, Preziozi P, Bertrais S, Castetbon K, Hercberg S. Background and rationale of the SU.FOL.OM3 study: Double-blind randomized placebo-controlled secondary prevention trial to test the impact of supplementation with folate, vitamin B6 and B12 and/or omega-3 fatty acids on the prevention of recurrent ischemic events in subjects with atherosclerosis in the coronary or cerebral arteries. Nutr Health Aging. 2003; 7: 428–435.