WHAT are fish made of? 鈥淔ive per cent protein, 95 per cent politics鈥� goes the reply of one former fisheries official. In the case of farmed salmon, it seems the ingredient list should also include dioxins, PCBs and several other organochlorine pollutants. And as the war of words raging over last week鈥檚 study confirms, these too come laced with politics (see 鈥淎 fishy tale of salmon, dioxins and food safety鈥�).
Nobody is seriously faulting the study鈥檚 sample sizes or raw data. But do the detected traces of contamination really justify warning people to eat no more than a portion a month of such heart-friendly fare 鈥� and even less if the farmed fillets hail from Scotland or the Faroe islands? The study鈥檚 authors, armed with an approach to risk assessment championed by the US government鈥檚 Environmental Protection Agency (EPA), are adamant the warning is justified. Experts at the US Food and Drug Administration and in the UK at the Food Standards Agency insist it is scaremongering.
The two camps are not even close. The EPA sets limits for PCBs and similar chemicals in fish at 24 to 48 parts per billion (the average levels in the farmed salmon were around 36 parts per billion). The other agencies set thresholds of up to 2000 parts per billion. Each side enthusiastically rubbishes the other鈥檚 methods as outdated. Neither has adequately explained the scientific assumptions 鈥� or uncertainties 鈥� underpinning their statements.
Advertisement
Both camps ought to come clean on the fact that there is no foolproof way of assessing the health risks of PCB-like chemicals. Toxicologists can give a range of largish doses to lab rats, or cells in culture, and look for cancer and other abnormalities. But they then have to extrapolate the findings to bigger-bodied humans exposed to much smaller doses. Not only is this an imprecise science, experts are bitterly divided on the fundamentals.
The EPA鈥檚 approach, favoured by greens, assumes that there is a linear relationship between the dose of a chemical and any health effects, and that the line can be extrapolated seamlessly back from the large doses used in animal tests to tiny doses of the kind seen in farmed salmon. By this 鈥渓inear model鈥�, the cancer-causing powers of a dioxin or PCB shrink as the dose falls, but never drop to zero: there is no completely safe threshold.
The competing view, backed by the WHO, holds that there is a threshold below which PCB-like chemicals lose all their cancer-causing powers, and that the linear model therefore exaggerates the dangers of the lower doses. It is a complex argument, but the gist is that PCB-like chemicals 鈥� unlike, say, ionising radiation 鈥� do not make cells cancerous directly, by damaging their core DNA. The chemicals act more subtly, stimulating receptors in cells in a variety of ways that can make it easier for any existing cancer cells to grow and form tumours. There comes a point, says the threshold theory, at which the contaminant is simply too dilute to stimulate enough receptors to influence cells.
Neither approach explains all the facts or is free of assumptions. The linear model assumes, for instance, that mammals have few or no ways to eliminate or repair cancerous cells, and that our bodies metabolise low doses of chemicals in the same way as high doses. The threshold model assumes that whenever toxicologists see a pollutant鈥檚 biological effects disappear at low doses, they are witnessing the safety threshold kicking in rather than the limits of their experiment鈥檚 sensitivity. And of course, cancer is not the only worry: evidence that PCBs and dioxins may disrupt the immune and endocrine systems goes back years.
So if the toxicology is unclear, how should we respond? We can push for regulations that ensure all farmed salmon is labelled with its origins. That would enable us to seek out fish from the 鈥渃leanest鈥� producers, while putting pressure on the industry to clean up the fish meal it uses. But be aware that shunning salmon on the basis of the EPA鈥檚 risk model means giving up not only the benefits of omega-3 fish oil, but, possibly, other contaminated foods. In the past, even butter has been found to carry comparable PCB levels.
Nor will eliminating the chemicals from feed solve the biggest threats to farming鈥檚 blue revolution. Despite the industry鈥檚 efforts to clean up, sea lice, disease, insecticides and escaping fish all continue to make salmon farming an environmental pariah. The contamination issue isn鈥檛 the most important reason to worry about farmed fish. Let鈥檚 hope it might just be the one to spur some action.