Eating oily fish every week may lower a person's risk of suffering a stroke, however taking fish oil supplements does not appear to have the same effect, a new study suggests.

The consumption of fish has long been linked to a reduction in heart disease, particularly oily fish, which contain omega-3 fatty acids. Oily fish includes salmon, mackerel and herring. However, until now, a link between fish consumption and a lower risk of stroke has been less clear.

European scientists carried out a detailed analysis of 38 studies involving almost 800,000 people in 15 countries. The studies included people with heart disease and without.

The scientists looked specifically at the effect of fish consumption on the risk of stroke and mini-stroke, known as TIA (transcient ischaemic attack).

Almost 35,000 stroke or mini-stroke events were recorded in the studies.

The scientists found that people who ate between two and four servings of oily fish per week had a ‘moderate but significant' 6% reduced risk of suffering a stroke or TIA.

Those who ate five or more servings every week had a 12% reduced risk.

Even just two servings per week was linked to a 4% reduced risk.

However, fish oil supplements did not have the same effect and did not appear to reduce the risk of stroke.

The scientists said there could be a number of reasons for this. Eating higher amounts of fish may mean that people are eating less of other foods that are not considered as good for heart health, such as red meat. Alternatively, eating a lot of fish may simply indicate an overall healthier diet or higher socioeconomic status, both of which are linked to better heart health.

They also suggested that the differences noted between oily fish and white fish, such as cod and haddock, may be down to how these foods are cooked. White fish is often battered and/or deep fried.

The team said that the findings support current guidelines in relation to the consumption of fish and added that future nutritional guidelines should be mainly ‘food based'.

Details of these findings are published in the British Medical Journal.