Omega-3 Fatty Acids—Science, Efficacy, and Clinical Use in Cardiology

Register or Login to View PDF Permissions
Permissions× For commercial reprint enquiries please contact Springer Healthcare:

For permissions and non-commercial reprint enquiries, please visit to start a request.

For author reprints, please email
Average (ratings)
No ratings
Your rating
Copyright Statement:

The copyright in this work belongs to Radcliffe Medical Media. Only articles clearly marked with the CC BY-NC logo are published with the Creative Commons by Attribution Licence. The CC BY-NC option was not available for Radcliffe journals before 1 January 2019. Articles marked ‘Open Access’ but not marked ‘CC BY-NC’ are made freely accessible at the time of publication but are subject to standard copyright law regarding reproduction and distribution. Permission is required for reuse of this content.

Studies have shown that omega-3 fatty acids (FAs) can promote cardiovascular health by both primary and secondary prevention of cardiac disease, and most notably by reducing fatal cardiovascular events including sudden cardiac death (SCD). For example, recent evidence from secondary prevention trials has suggested that intake of 850mg/day of omega-3 FAs can reduce the risk of coronary heart disease (CHD) events by 25% and SCD by about 45%.1 Furthermore, of the three known omega-3 FAs there is evidence that eicosapentanoic acid (EPA) is particularly biologically important. Given the strength of the evidence, blood levels of omega-3 FAs may have value as biomarkers to evaluate risk for cardiovascular events, leading to opportunities for earlier intervention in high-risk cases.

What Are Omega-3 Fatty Acids?

Omega-3 FAs, like omega-6 FAs, are a type of polyunsaturated FA (PUFA) that the body is unable to synthesize and that must, therefore, be obtained through the diet. The omega-3 FAs are distinguished from other classes of FAs by the structure of their carbon chains. FAs are chains of carbon atoms that are labeled with the ‘alpha’ carbon proximal to the carboxyl group and end with the distal ‘omega’ carbon. The omega-3 FAs contain a double bond between the third and fourth carbons from the omega carbon. All of the omega-6 FAs contain their first double bond between the sixth and seventh positions from the omega carbon. Humans lack the enzymes required to place a cis double bond at the omega-3 or omega-6 positions. Thus, omega-6 and omega-3 FA are essential FAs and must be supplied in our diet. Omega-6 FAs are present in grains and nuts. Omega-3 alpha-linolenic acid (ALA) is an 18-carbon FA that is found in plants but cannot be easily elongated to the more important 20-carbon form by humans. The omega-3 FAs EPA and docosahexaenoic acid (DHA) are synthesized by phytoplankton and then consumed and concentrated by fish. They may be acquired in the diet by fish consumption, by foods fortified in omega-3, or by further concentration and specification in high-concentrate nutritional omega-3 supplements.

The 1982 Nobel Prize2 was awarded for the discovery of omega-6 FA in the form of prostaglandins and related biologically active substances that are metabolites of arachidonic acid (AA), the 20-chain omega-6 FA that is a foundation of the inflammatory response. These omega-6 AA metabolites include prostaglandins, such as the thromboxanes and leukotrienes. Subsequent to that award, it has become appreciated that the pro-inflammatory 20-chain omega-6 arachidonic pathway is counterbalanced by an anti-inflammatory pathway based on the 20-chain omega-3 EPA. The scientific literature is growing regarding the tremendous health benefits of omega-3 supplementation both for overall health and disease prevention and for the treatment of many inflammatory conditions.

Grains are rich in omega-3 FA. The agricultural revolution greatly increased the human food supply by permitting the large-scale production of grains. The caloric plethora afforded by the agricultural revolution, however, has resulted in disproportionate ratios of omega-6 to omega-3 FA in the modern diet: while the optimal ratio in the body of omega-6 to omega-3 is approximately 1:1, the current ratio in Westernized societies is approximately 16:1,3 with a resultant increase in inflammatory disorders.

This increase in the pro-inflammatory omega-6 from grains may also bring an increase in diverse pro-inflammatory conditions, including heart disease, asthma, depression, arthritis, bowel, and even cancer. Accordingly, there appear to be multidimensional health benefits from dietary omega-3 supplementation beyond the cardiovascular realm. For example, in addition to the American Heart Association (AHA) guidelines on omega-3 FA intake,4 the American Psychiatric Association (APA) has issued a consensus statement recommending 1g EPA+DHA per day for patients with mood, impulse-control, or psychotic disorders.5 Moreover, the American Diabetes Association (AHA) recommends that diabetics increase their omega-3 FA consumption.6 The explosion of scientific evidence surrounding the importance of omega-3 FAs in health grows daily. The current PubMed database contains 11,081 references for omega-3.

Dietary Sources of Omega-3 Fatty Acids

Omega-3 FAs are present as ALA in a few vegetable oils, e.g. flaxseed, canola, and soybeans; green leafy vegetables, e.g. purslane; and nuts, e.g. black walnuts. However, the main dietary source of EPA and DHA is from fish. The marine omega-3 FAs originate in micro-algae such as phytoplankton, which are then ingested by fish as part of their diet.

Fish particularly high in omega-3 FAs tend to be oily rather than lean white fish, and include salmon, mackerel, sardines, and herrings. Of course, the quantity of omega-3 FAs in particular breeds of wild fish can vary widely depending on season, maturity, the diet of the fish, post-catch processing, and cooking methods. However, farmed fish, e.g. salmon and rainbow trout, may contain no omega-3 if raised on cornmeal or may contain stable levels of EPA and DHA if fed fishmeal.7

A further source of omega-3 FAs is from foods such as eggs, bread, and milk that have been fortified with omega-3 FAs, derived either from oily fish or from micro-algae. These can provide an alternative source in place of fish, but are not generally recommended as a substitute for fish consumption. It is not possible at this time to obtain the high level of EPA useful in cardiology from fortified foods.

There is concern over the potentially harmful effects of mercury, dioxins, and polychlorinated biphenyls present in some fish species, especially larger fish. Those containing particularly high concentrations of mercury include shark, swordfish, king mackerel, and tilefish. In a 2006 meta-analysis, Mozaffarian and Rimm concluded that the benefits of fish intake far exceed any potential risks;8 however, those species particularly high in contaminants should be avoided by pregnant women and nursing mothers. Omega-3 supplements provide a safe way to obtain omega-3, as the mercury remains bound to the protein of the fish and is not concentrated into the oil. Fish oil can be distilled to high levels of purity and concentration of particular omega-3s (EPA or DHA).

Omega-3 supplements can also be taken to boost intake levels, and in 1997 The US Food and Drug Administration (FDA) granted Generally Recognized As Safe (GRAS) status to refined menhaden fish oil, and ruled that consumption of up to 3g/day of EPA plus DHA from all sources would be safe for American adults.9 Also, omega-3 supplements provide a good indication of the volumes of EPA and DHA being consumed, since a Consumer’s Report analysis of 16 brands found label claims to be generally accurate. The selection of a high-EPA omega-3 nutritional supplement from a company that tests each individual batch of omega-3 for purity, potency, and label claim is essential. Furthermore, the selection of a supplement that is manufactured under nitrogen or xenon to prevent oxidation provides an omega-3 free of aftertaste—an important factor for patient compliance.10

Theoretical Cardioprotective Mechanisms

The cardioprotective mechanisms of omega-3 FAs appear to be based on the incorporation of EPA and DHA into cell membranes, altering the membrane’s physical characteristics and the activity of membrane-bound proteins.11 EPA undergoes conversion into a wide variety of bioactive prostaglandins, thromboxanes, leukotrienes, and eicosanoids resolvins. It also acts as a ligand for several nuclear transcription factors via peroxisome proliferator-activated receptors and their role in lipid metabolism, inflammation, and atherosclerosis via modulation in cell-cholesterol trafficking and inflammatory activity,12 consequently altering gene expression.13 Also, a variety of intracellular messages are triggered.

Recent studies point to the importance of EPA not only in terms of decreasing the formation of atherosclerotic plaque and decreasing embolic stroke, but also in terms of their ability to improve endothelial function, reverse atherosclerotic plaques over time, and increase arterial elasticity.14 While the benefit is clearly multifactorial, the relative importance of each of these actions, the manner in which they co-ordinate, and their ability to explain clinical results are being explored. The overall health benefits of omega-3 allow the clinician to provide the patient with a cardiac treatment that has positive health benefits without adverse effects. The following observations have been made regarding the effects and use of omega-3 FAs on cardiovascular health, risk factors, and disease.

Antiarrhythmic and Antifibrillatory Effects

A principal mechanism by which omega-3 FAs reduce adverse cardiac events is by reducing myocardial susceptibility to lethal arrhythmias. Omega-3 FAs have been shown to decrease the resting heart rate and improve heart rate recovery and variability parameters associated with risk for SCD.15 Much of the evidence for the antiarrhythmic effect comes from animal and cell culture studies. For example, an infusion of an omega-3 emulsion prevented ischemia-induced fibrillation in dogs, and when cultured neonatal rat cardiomyocytes were incubated with omega-3 FAs they were protected from various cardiotoxins. There are supportive data from human trials too, such as the reduction in sudden death in post-myocardial infarction patients.15 This benefit may accrue from the relatively small doses of omega-3 FAs (250–500mg/day of EPA+DHA) that may be ingested via dietary sources.

Effect on Hyperlipidemia

Although omega-3 FAs do not have a significant effect on blood cholesterol, they do lower blood triglycerides—an important blood lipid that is produced by the liver—when EPA and DHA are given at doses of about 3–4g/day. These higher doses require the use of purified supplements high in EPA and DHA. In these higher doses, they inhibit the liver from synthesizing triglycerides and appear to accelerate the removal of triglycerides from the blood. All in all, they typically lower triglyceride levels by between 20 and 50%. They are effective even when the triglyceridemia is severe, and their benefit appears to be additive to statins and other antihyperlipidemics. These observations prompt the recommendation of a combination of omega-3 and statin therapy for serious hyperlipidemia.16

Effect on Blood Pressure

A meta-analysis of several studies revealed changes in blood pressure with the use of fish oil.17,18 The benefits tended to occur in older populations (>45 years) and generally required higher doses. Although the blood pressure changes were minor, they may nonetheless be clinically significant since even small changes in blood pressure can influence mortality rates from CHD and stroke. It is likely that the positive effects of fish oil on blood pressure are the result of an improvement in endothelial function.

Effect on Thrombosis, Atherosclerosis, and Inflammation

Numerous effects of omega-3 FAs on thrombosis, atherosclerosis, and inflammation have been reported.19 These benefits require the higher doses of EPA afforded by highly concentrated high-EPA omega-3 supplements.

Effect on Sudden Cardiac Death

Strong evidence for the cardioprotective mechanism of omega-3 FAs at the lower doses achievable by fish ingestion relates to effects on fatal cardiac events rather than a general reduction in all cardiac events. For example, in the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto (GISSI) Miocardico Prevenzione trial there was no decrease in cardiac events overall, but those given EPA+DHA were at lower risk for fatal myocardial infarctions. Therefore, a reduction in SCD is most likely to be effected by an increased intake of omega-3 FAs at lower doses.14

In animal models, all three omega-3s—ALA, EPA, and DHA—have been shown to be equally capable of inhibiting the accumulation of high concentrations of cytosolic-free calcium in cardiac myoctes and stabilizing the electrical activity of the cell, potentially being the means of reduction of sudden death seen in the GISSI trial.20

Effect on Cardiac Events and Cardiac Death

The cardioprotection and decrease in adverse cardiac events from high-dose EPA omega-3, even in patients on a fish-rich diet and a statin, was demonstrated in the Japan EPA Lipid Intervention Study (JELIS), where 18,645 patients were randomized to receive 1.8g/day EPA or act as a control.21 Doses of 1.8g/day EPA yield significant improvement when the end-point is any major coronary event, fatal or otherwise.22


The link between omega-3 FAs and improved heart health was first observed by Bang et al. in an epidemiological study carried out on Greenland Eskimos in the 1970s. Comparing myocardial infarction rates among Eskimos with those among Danes they observed that, in spite of a diet high in fat from seal, caribou, and fish, the Eskimos suffered significantly fewer cardiac events than the Danes. The unusually high blood levels of omega-3 FAs in the Eskimos were proposed as the link to improved heart health.21

In spite of a substantial amount of evidence suggesting the positive effects of omega-3 FAs on cardiovascular events, a British Medical Journal (BMJ) meta-analysis, based on a survey of randomized controlled trial results, reported that the evidence was ‘not strong’ for an effect of omega-3 FAs on overall mortality rates or on risk for cancer or cardiovascular disease.23 The authors noted that the inclusion of one large study by Burr et al., published in the European Journal of Clinical Nutrition, skewed the results to the null.24 That study, often referred to as the Diet and Reinfarction Trial (DART) II (because the same researchers used a design similar to that in the original DART study25), has been criticized as having serious limitations in design and study implementation, partially due to an interruption in study funding. Therefore, it is questionable whether valid conclusions can be drawn from its finding; this is also true for the BMJ meta-analysis.14

A more recent meta-analysis conducted by Mozzaffarian and Rimm has several advantages over the BMJ study.8 It analyzed the effects of fish and fish oil as opposed to only omega-3, it covered many years of intake rather than just a couple of years, and it included a large number of population studies estimating dietary intake levels. Looking at the benefits versus the risks of intake of fish and fish oil, it concluded that the benefits of fish oil vastly outweigh the potential risks.

GISSI was a large and well-controlled trial assessing the combined effects of EPA and DHA on death, non-fatal myocardial infarction, and stroke. Using a pharmaceutical omega-3 preparation, the trial tested the hypothesis that relatively small intakes of omega-3 FAs (<1g) could reduce the risk for death from CHD in high-risk patients. Over 11,000 patients who had experienced a heart attack less than three months previously and who were receiving modern cardiac pharmacotherapy were randomized to 850mg/day of EPA plus DHA, 300mg/day of vitamin E, both, or no treatment. After 3.5 years of follow-up, the group given just the EPA plus DHA (n=2,836) experienced a reduction in death from any cause of 20%, a decrease in cardiac death of 35%, and a reduction in sudden death of 45% (all p<0.01 compared with the untreated control group, n=2,828).15 The reduction in risk for sudden death was regarded as particularly relevant and became statistically significant after four months. After six to eight months, a similarly significant pattern was observed for cardiovascular, cardiac, and coronary deaths.

Recently, the JELIS demonstrated the benefit of added high-concentration EPA over and above the other omega-3 FAs in reducing negative cardiac events.22 The study showed that the addition of 1.8g of EPA could reduce the risk for cardiac events even in patients already consuming high levels (800–1,000mg/day) of dietary EPA+DHA and who were also taking statins. There was no difference in SCD between the control and the EPA groups; however, the high background intake of omega-3 FAs in the Japanese population most likely virtually eliminated this disease from the study cohort to begin with. However, the improved outcomes of the experimental group over the broad range of cardiac end-points imply a benefit of higher EPA doses for both fatal and non-fatal cardiac events.

The Omega-3 Index as a Measurement of Risk

In light of this evidence, it has been suggested that omega-3 FA blood levels may serve as a risk factor for cardiovascular disease, and particularly for SCD. Red blood cell (RBC) membrane EPA+DHA levels reflect dietary omega-3 intake and are a valid surrogate of human cardiac membrane omega-3 FA content, and high levels were found to be associated with a 90% reduction in odds for SCD. This biomarker—the Omega-3 Index—is proposed as a new, physiologically relevant, modifiable, independent, and graded risk factor for death from CHD.25

An Omega-3 Index of ≥8% was proposed as a healthy target value since in a survey of published studies this level was associated with the greatest cardioprotection, whereas an index of ≤4% was associated with the least.23 Consistent with the data quoted above, moderate elevations in the Omega- 3 Index may predict reductions in fatal cardiac events, and higher elevations may predict reductions in all cardiac events, both fatal and non-fatal.

Although the method is still in development, the Omega-3 Index is potentially easy to measure, given the right equipment and methodology. RBC membrane composition is analyzed by direct methylation of packed RBCs using boron triflouride in methanol. This generates FA methyl esters, which are analyzed by capillary gas chromatography.13

The Omega-3 Index is most useful as a predictor for SCD, for which risk factors are currently limited. Compared with other classic CHD risk factors, the Omega-3 Index performed more favorably in predicting SCD than any other, including blood lipids, C-reactive protein, and homocysteine. The Omega-3 Index is a consistent predictor in epidemiological studies, has a plausible mechanism of action, is a reproducible assay, and is predicting-risk independent.26 In addition, the Omega-3 Index is easily, safely, and cheaply modifiable and, most importantly, raising the Omega-3 Index has been demonstrated to reduce the risk for fatal cardiac events.23

Future studies with major epidemiological studies such as the Framingham study are planned. These investigations will provide important evidence regarding the ability of omega-3 levels to add significantly to current risk prediction algorithms that already include known risk factors such as age, gender, cholesterol, blood pressure, and smoking. It is important to determine whether the Omega-3 Index can improve on these factors to provide a more accurate risk prediction.

Use of Omega-3 Nutritional Supplements in Clinical Practice

Omega-3 FA supplements are food products under the regulatory guidance of the FDA. High-quality high-EPA omega-3 nutritional supplements are now available for the cardiologist to use. The FDA allows a qualified health claim for foods and supplements containing EPA and DHA omega-3 FAs that the product may reduce the risk of heart disease.7 The JELIS study further demonstrates the value of high EPA in an omega-3 supplement, providing benefit when combined with statin for preventing both fatal and non-fatal cardiac events.22

The following criteria may be helpful in selecting an omega-3 supplement: high concentration of 90% omega-3 or greater to allow for smaller dosing and more compliance; high EPA concentration of 70% or higher for greater benefit; third-party testing of each individual batch for purity and label claim; no aftertaste provided by careful manufacturing without exposure to oxygen; small capsule size for compliance—500mg is easy, whereas 1,000mg can be difficult for many patients; and manufacture at FDA-approved facilities with full Good Manufacturing Practices (GMP) compliance.

The recent approval of one omega-3 product to be called a drug for use in reduction of triglycerides has brought confusion in two areas. First, what does it mean for a nutritional supplement to be FDA-approved as a drug? Is it superior to an omega-3 that is not FDA-approved as a drug when the molecular structure is the same? FDA approval is for proven safety and efficacy for a disease indication. Given that omega-3s have a qualified FDA health claim already and are GRAS, FDA approval is of less importance than selecting the supplement with the desired ratio and concentration of omega-3. FDA approval allows the manufacturer to make the label claim to treat hyperlipidemia, which all EPA and DAH omega-3s have been shown to do. The manufacturer is allowed to make the disease claim in advertising. Given the FDA-qualified claim for omega-3 DHA and EPA for reducing the risk of heart disease and the GRAS status, for use beyond reduction of triglycerides the cardiologist may choose to select a higher EPA omega-3 for greater benefit.

Second, why has only one omega-3 been FDA approved despite other products having higher concentration, higher EPA, and equal or higher purity? The answer is that the FDA grants a five-year exclusivity period for the first nutritional supplement granted FDA approval for treatment of a disease. This means that no other omega-3 companies can even apply for FDA approval for five years from the date at which the FDA approval was granted for the first omega-3.

It may be helpful for cardiologists to understand that there is a period of exclusivity that prevents other omega-3s from applying for FDA approval despite identical or even higher concentration and purity. FDA approval is not an indication of purity, concentration of the omega-3, or a superior manufacturing process. Given the FDA-qualified claim for EPA and DHA omega-3s, the FDA GRAS status, and the availability of omega-3 supplements with the same level or higher concentration omega-3, higher ratio of EPA, and proven purity and label claim by third-party testing on every batch, the cardiologist can choose from a range of omega-3 nutritional supplements to meet his or her clinical needs; the cardiologist can therefore choose the higher EPA supplement over the lowerconcentration FDA-approved omega-3 supplement.

The JELIS study indicated the importance of high EPA for both preventing and treating cardiac disease beyond simply lowering of triglycerides; therefore, cardiologists want a high EPA supplement of proven purity and quality for their patients. OmegaBrite provides such an option, having a higher concentration of EPA and a lower cost to the patient than the drug omega-3 FDA-approved as an adjuvant therapy for hyperlipidemia. The hyperlipidemia treatment Lovasa, previously Omacor, contains only 84% omega-3 EPA and DHA and provides only 465mg of EPA at a retail cost of $193.49 for 120 capsules (1,000mg size). In contrast, the nutritional supplement OmegaBrite provides 90% omega-3 and contains 700mg of EPA per gram in a smaller capsule, and is less than half the price. At $21.99 for 60 capsules, the cost is approximately $54.95 a month to achieve a daily dose of 1.8g of EPA, as in the JELLIS study. To achieve the same 1.8g of EPA using the lower EPA concentration Lovaza, the cost is $195.00 per month. This represents a significant difference for the patient; in addition, the smaller, more highly concentrated OmegaBrite capsule is easier to swallow, leading to greater compliance.

OmegaBrite was recently awarded the Product Merit award by the Nutrition Business Journal for creating the category of high-concentrate omega-3 supplements, creating the first high-concentrate EPA omega-3 supplement, and elevating the standards of the industry in terms of purity, concentration, and scientific evidence. OmegaBrite is third-party tested in every batch by the manufacturer, Omega Natural Science, and has also won the Consumer Labs Seal of Approval by exceeding its highest standards of purity, potency, and label claim. With this proven level of purity and quality and a higher EPA and concentration, OmegaBrite allows the cardiologist to offer the patient a superior product at a far more affordable price.


There is substantial evidence to suggest that the long-chain omega-3 FAs, particularly EPA, may be the most potent cardioprotective factors in the human diet. Achieving optimal omega-3 FA intakes — 1,000mg/day for reduction of fatal events and >1.8g/day of EPA for reduction of all cardiac events — will have a profound effect in safely reducing the risk of cardiac disease, promoting cardiac health, and providing the cardiologist with a safe and effective treatment option. Furthermore, the JELIS study demonstrated the advantage of high EPA supplementation in reducing cardiac events of all types, even when on a high-fish diet and on a statin. With high-EPA omega- 3 supplements of proven high purity such as OmegaBrite, the cardiologist is now armed with a profoundly effective, safe, and affordable means of promoting cardiac health and preventing and treating cardiac disease.


  1. Marchioli R, et al., Circulation, 2002;105:1897–1903.
    Crossref | PubMed
  3. Franz MJ, et al., Diabetes Care, 2003;26(Suppl. 1):S51–61.
    Crossref | PubMed
  4. Kris-Etherton PM, et al., Arterioscler Thromb Vasc Biol, 2003;23: 151–2.
    Crossref | PubMed
  5. Freeman MP, et al., J Clin Psychiatry, 2006;67:1954–67.
    Crossref | PubMed
  6. USDA Nutrient Data Laboratory. Available at: (accessed October 3, 2002).
  8. Mozaffarian D, Rimm EB, JAMA, 2006;296(15):1885–99.
    Crossref | PubMed
  9. Consum Rep, 2003;30–32.
  10. Harris WS, Pharmacol Res, 2007;55(3):217–23.
    Crossref | PubMed
  11. Swan JS, et al., Cardiovasc Res, 2003;1;60(2):337–46.
    Crossref | PubMed
  12. Food and Drug Administration, Fed Regist, 1997;30751–7.
  13. Hjerkinn EM, et al., Eur J Cardiovasc Prev Rehabil, 2006;13(3): 325–33.
    Crossref | PubMed
  14. von Schacky C, Harris WS, J Cardiovasc Med (Hagerstown), 2007;8(Suppl. 1):S46–9.
    Crossref | PubMed
  15. GISSI-Prevenzione Investigators, Lancet, 1999;354:447–55.
    Crossref | PubMed
  16. Nambi V, Ballantyne CM, Am J Cardiol, 2006;98(Suppl.): 34i–38i.
    Crossref | PubMed
  17. Appel LJ, et al., Arch Intern Med, 1993;153(12):1429–38.
    Crossref | PubMed
  18. Geleijnse JM, et al., J Hypertension, 2002;20(8):1493–9.
    Crossref | PubMed
  19. Lopez-Garcia E, et al., J Nutr, 2004;134(7):1806–11.
  20. Kang JX, et al., Proc Natl Acad Sci U S A, 1995;92:3997–4001.
    Crossref | PubMed
  21. Bang HO, et al., Acta Med Scand, 1976;200:69–73.
    Crossref | PubMed
  22. Yokoyama M, Origasa H, JELIS Investigators, Am Heart J, 2003;146(4):613–20.
    Crossref | PubMed
  23. Hooper L, et al., Br Med J, 2006;332:752–60.
    Crossref | PubMed
  24. Burr ML, et al., Lancet, 1989;2:757–61.
    Crossref | PubMed
  25. Harris WS, von Schacky C, Prev Med, 2004;39(1):212–20.
    Crossref | PubMed
  26. Biomed Pharmacother, 2006;60(9):502–7. Epub 2006 Aug 28.
    Crossref | PubMed