Fish liver consumption is not associated with increased cancer risk, say Norwegian scientists.

Fish liver is a traditional food in northern Norway and an important source of vitamin D for Norwegians during dark winter months. However, it is also one of the most contaminated food items in the country, containing high levels of persistent organic pollutants (POPs), including dioxins. POPs have been associated with negative health effects, including an increased risk of cancer. 

The Norwegian Food Control Authority (now the Norwegian Food Safety Authority) has advised that children and women of childbearing age should avoid eating fish liver. This is based on EU recommendations for tolerable weekly intakes of toxic compounds and the health-related risks associated with long-term exposure to dioxins and dioxin-like compounds.

In 2006, however, Magritt Brustad and co-workers from the University of Tromsø showed that POP levels in human blood are not significantly affected by fish liver intake.¹ Now, the team has studied whether consumption of the traditional food affects cancer rates.² From a survey of nearly 65 000 Norwegian women, they have shown that eating fish liver is not associated with increased cancer risk. In fact, they found that total cancer risk was slightly reduced for fish liver consumers.

'This is reassuring,' said Tim Key, deputy director of the Cancer Research UK Epidemiology Unit at the University of Oxford, who praised the study as 'well-designed.'

Brustad said the survey implies 'that the contribution to POPs from the traditional fish liver diet in Norway is below the limit that could give increased cancer risk.' The team suggests that the potential health gain of eating fish liver - from fat-soluble vitamins and essential fatty acids - is greater than the possible negative effects caused by contaminants.

When asked whether the Norwegian Food Safety Authority should change its recommendation regarding fish liver consumption, Brustad suggested that, as well as basing dietary advice on toxicological estimations from animal models, 'population-based research methodology should be taken into consideration.'

Freya Mearns