Data from the GISSI-Prevenzione study[Lancet 1999;354:447-55] suggested that omega-3 fatty acids may have the ability to reduce death from cardiovascular causes, mainly through a reduction in sudden cardiac death (SCD). However, subsequent studies among patients with implantable cardiac defibrillators (ICD) failed to demonstrate any reduction in device therapy with supplemental fish oils, possibly due to methodological limitations such as small sample size, poor patient adherence and high loss to follow up. Furthermore, previous systematic reviews of fish oil therapy have been inconclusive but have not included some of the most recent data, such as the JELIS trial [Lancet 2007;369:1090-8].

Leon et al performed a systematic review of 12 randomised trials of fish oil supplementation, which included a total of 32,779 patients. The primary outcome of interest was the arrhythmic end point of appropriate ICD intervention and SCD. Secondary outcomes were all cause mortality and death from cardiac causes (not specified further). Finally, subgroup analysis was performed in order to examine the effect of fish oil supplement formulation, either docosahexaenoic acid (DHA) or eicosapentaenoic acid (EPA) or both, on death from cardiac causes.

Overall, the use of fish oil supplementation was not associated with either a reduction in the risk of appropriate ICD intervention (OR 0.90, 95% CI 0.55 to 1.46) or a reduction on the incidence of SCD (OR 0.81, 0.52 to 1.25). Although all cause mortality was not affected by fish oil supplementation, a 20% reduction (OR 0.80, 0.69 to 0.92) in death from cardiac causes was, however, observed mainly due to reduced coronary events. The potential mechanism enabling fish oils to act in this way remains elusive, however it is a possible that fish oils may stabilise atherosclerotic plaque via their anti-inflammatory influence. A dose-response relationship for effects on death from cardiac causes was not seen and therefore comment cannot be made as to the most efficacious formulation of fish oils.

This review has several limitations. Firstly, 92% of the patients studied came from two large different studies; GISSI, a secondary prevention study among patients after myocardial infarction and JELIS, a study of hypercholesterolaemic subjects for both primary and secondary prevention. Secondly, significant publication bias was detected in the present review, which suggests that other neutral or negative studies may not have been published. Finally, a wide variation in dosage of fish oils was apparent among the studies included in the review (0-2000mg/day), making direct comparisons very difficult.

Nevertheless, despite its limitations, this review demonstrates that fish oil supplementation was associated with a significant reduction in deaths from cardiac causes but had no effect on arrhythmia. Further studies are needed in order to clarify the exact mechanism responsible for this reduction in cardiac death and results of the ongoing OMEGA trial [Cardiovasc Drugs Ther 2006;20:365-75] are keenly awaited.

· Leon H, Shibata M, Sivakumaran S et al. Effects of fish oils on arrhythmias and mortality: systemic review. BMJ 2009;338:a2931