Risk Assessment and Evaluation Approach Applied in Study of Contaminants in Farmed Salmon

1. Risk Assessment and Evaluation Approach Applied in Study of Contaminants in Farmed Salmon David O. Carpenter, MD Institute for Health and the Environment University at Albany Rensselaer, NY 12144 European Food Safety Colloquium June 28, 2004 Supported by the Environmental Division of the Pew Charitable Trust

2. SALMON A very popular fish with high levels of omega-3 fatty acids, known to be beneficial in preventing sudden cardiac death. Farming of salmon has grown very rapidly, now at levels of over 1 million tons per year. Farmed salmon are relatively cheap and are available throughout the year.

3. Purpose of Our Study To analyze Atlantic salmon from farms in eight salmon-farming regions, Atlantic salmon fillets from supermarkets in 16 cities and wild Pacific salmon of five species for the presence of environmental organic and metal contaminants.

4. We purchased 459 whole farmed salmon from 51 farms in eight farming regions in six countries (Scotland, Norway, Faroe Islands, Eastern Canada, Maine, Western Canada, Washington State, Chile). • We purchased 135 wild Alaskan salmon, including chum, coho, chinook, pink and sockeye, from suppliers in Alaska and Western Canada. • We purchased salmon fillets in supermarkets in 16 North American and European cities (Vancouver, Seattle, Los Angeles, San Francisco, Denver, Chicago, Toronto, New Orleans, Washington, D.C., New York, Boston, London, Edinburgh, Paris, Frankfurt, Oslo). • Composites of three fish or three fillets (for a total of 246 samples) were analyzed for 14 organic contaminants, nine metals and lipids. • We purchased 13 samples of farmed salmon feed from different parts of the world for analysis.

8. EPA Fish Consumption Advisories For Single Contaminants: Eq. 1 CRlim =ARL ● BW CSF ● Cm For Maximum Allowable Meals per Month: Eq. 3 CRmm = CRlim ● Tap MS For Multiple Contaminants: Eq. 2 CRlim = ARL ● BW x  Cm ● CSF m=1 Crlim = maximum allowable fish consumption rate (kg/d) ARL = maximum acceptable individual lifetime risk level (unitless) (1:100,000) BW = body weight CSF = cancer slope factor (upper 95% CI) Cm = concentration of contaminant m (mg/kg) CRmm = maximum allow fish consumption (meals/mo) Tap = Time energy period (30.44 d/mo) MS = meal size (0.227 kg fish/meal)

11. The WHO uses an exposure-based approach for dioxins by issuing thresholds for contaminant intake rather than fish consumption advice (WHO, 1998).  This assessment stressed that the upper range of the TDI (4 pg TEQ/kg b.w.) should be considered a maximal tolerable intake on a provisional basis and that the ultimate goal is to reduce human intake levels below 1 pg TEQ/kg body weight/day.  Consumption advice for dioxins was generated using an intake level of 2 pg/kg/day, which is consistent with tolerable intakes established by the WHO (1998).  For consistency, we converted WHO intake levels for dioxins to fish consumption rates.  For all risk estimates and consumption calculations, it was assumed that an average meal size was 227 g (0.5 pounds) and the body weight of an average individual was 70 kg.

12. Patterns in consumption advice generated by the WHO Tolerable Daily Intake for dioxins are consistent with patterns in consumption advice for the other contaminants.  Generally, consumption of farmed Atlantic salmon at a rate of 1-2 meals/week (the American Heart Association’s recommended dietary guideline for fish consumption) would result in dioxin intake levels that reach or exceed the WHO limit of total human intake from all sources (WHO 1998), while consumption of wild salmon at rates as high as 14 meals/wk would be required to achieve the same dioxin intake levels.

13. B1 = Sufficient C = Possible B2 = Probable based on animals studies D = Not classifiable

14. The presence of contaminants may counteract the beneficial effects of omega-3 fatty acids: • Omega-3 fatty acids are well documented to prevent cardiac arrhythmias, especially in persons who have had one heart attack. They do not protect against cancer. However, PCBs and dioxins cause an elevation of serum lipids, which is the greatest risk factor for a heart attack. • Although less well established, there is some evidence that intake of omega-3 fatty acids during gestation increase cognitive function in children. But PCBs and dioxins are very well documented to cause deficits in IQ. • Therefore, while eating uncontaminated salmon is healthy, eating contaminated salmon or any other fish is going to counteract the beneficial effects.

15. CONCLUSIONS Farmed salmon have significantly greater levels of organochlorine compounds than do wild salmon, and the source appears to be the fish food. Farmed salmon from Northern Europe have significantly higher levels than those from North America which, in turn, are higher than those from South America. Salmon have significant amounts of omega-3 fatty acids, but the beneficial effects of omega-3 fatty acids on sudden cardiac death must be balanced against the increased risk of cancer from the contaminants. Using EPA cancer risk assessment methods for PCBs, dieldrin and toxaphene and WHO dioxin TEQ methods for cancer risk assessment, farmed salmon from all regions studied elicited highly restrictive fish consumption advisories, while those for wild salmon are much less stringent.

16. However, even this advisory does not consider other carcinogenic substances found to be present in farmed salmon, nor does it consider non-cancer endpoints. Some of the non-cancer endpointssuch as neurobehavioral decrements, endocrine disruption and immune system suppression, may be of greater importance than cancer from a public health perspective.

17. When considering food safety it is inappropriate to consider only one class of fat soluble contaminants, since there are a number of different chemicals that coexist in fish and animal fats, many of which are carcinogenic and also have adverse non-cancer actions.