How Much Farm-Raised Salmon Can You Safely Eat? Quantifying the Risk 
January 13, 2003

    I have been trying for several days to address the issues raised by the farm-raised salmon safety study showcased in today's Science News. It's proved quite difficult. A key question concerns the EPA's safety guideline that PCB's at a level of 50 parts per billion corresponds to a cancer risk of 1 in 100,000. That's meaningless. If I ate a one-milligram crumb of salmon containing 50 parts per billion of PCB's (viz., Scottish salmon), I would get 50 picograms of PCB's. If I ate a one-kilogram serving of that salmon, I would take on 50 micrograms of PCB's. Risk factors have to be attached to absolute amounts (nanograms, micrograms, milligrams) of toxins ingested rather than to concentrations.
    I have contacted someone with the EPA, but so far, I haven't been able to resolve this question.
    A short answer to the dilemma posed by this and other articles is as follows.
    First, the serving size specified in the study described in these articles is 8 ounces. That's a generous portion. If we're willing to back off to a four-ounce serving, that would halve the risk.
    Second, most of the farm-raised salmon in the United States comes from Chile or from the State of Washington. This has half the average level of PCB's (and presumably, of other contaminants such as dioxin, dieldrin, and toxaphene). So this would reduce your level of exposure to one-fourth that cited in the study.
    Third, most of the pollutants are concentrated in the skin, and in the thin layer of subcutaneous fat located just beneath the skin. If you remove the skin and its subcutaneous prior to cooking, this should reduce the amount of toxins in the cooked fish by another factor of several (particularly if you grill the salmon and let the fat that's released run off).
    The study didn't remove the skin or subcutaneous fat layer in arriving at their numbers.
    The report recommends wild salmon, including canned salmon, finding an average level of PCB's in wild Alaska salmon of 4.75 parts per billion.
    By observing the above precautions, you ought to be able to reduce the concentrations of toxins in farmed salmon to, or close to the levels of wild salmon.
    The major problem with farm-raised salmon, and the reason Tommie Jean and I quit buying it, is the fact that some of it is fattened on soybean oil and animal parts that give it no omega-3 fatty acids. The contaminants primarily show up when farmed salmon are fed "junk fish" and fish meal, but so do the omega-3 fatty acids. Canola oil can be used because it contains omega-3 oils, but you never know what's
in the salmon filets on display at the supermarket.
    Of course, you could also apply these precautions to wild salmon, further reducing their risks. One of the problems with wild salmon is that it costs about $15.00 a pound compared with $5.00 a pound for farm-raised salmon. Fresh wild salmon is only available for four months out of the year, although it can probably be purchased in frozen form.
    Canned salmon includes skin and subcutaneous fat, so the 4.75 parts per billion figure probably applies to it, also.
    Tommie and I are eating the canned, wild Alaska sockeye salmon recommended by Nicholas Perricone (M. D.) in his book, "The Perricone Prescription", pgs. 59-60. Is it safe? I hope so, but how do you tell? Our primary source of the omega-3 oils are fish oil capsules. 
    I've just ordered two bottles of fish oil capsules from the Life Extension Foundation. I'm hoping they'll be a reliable source.
    An alternative source might be Pure Encapsulations. Pure Encapsulations sells only to health care professionals, but some third-party vendors sell their products. Pure Encapsulations was allegedly founded by M. D.'s who wanted a trustworthy source of nutritional supplements. Pure Encapsulations tests its products. The best price I've found for Pure Encapsulations omega-3 oils is Kaysville Clinic. ($25.12 for 180 capsules, each containing 300 mg. of EPA and 200 mg. of DHA.). (However, I didn't search very thoroughly, and you may be able to find a better price.)
    One concern I have with these claims stems from my recent experience with resveratrol. I bought a bottle of resveratrol capsules from Pure Encapsulations through a pharmacist friend. Longevinex has given me the impression that it may be the only company in the world that sells biologically active resveratrol. David Sinclair is a consultant to the company, and he and his colleagues used Longevinex' resveratrol for their experiments with yeast.
    So where does that leave Pure Encapsulations, which says it tests its products? Longevinex' resveratrol is "encapsulated in a patented (by Pfizer) air tight capsule with internal nitrogen bubble. (Exclusive license LicapsŪ Capsugel.) Pure Encapsulations' resveratrol is a brown powder in an ordinary sleeve-type gelatin capsule. I haven't broached this subject yet with Pure Encapsulations. It will be interesting to see what they say. In the meantime, I've been reminded again of how unregulated this field of nutritional supplements is.
    Most of the major nutritional supplements mongers will sell you resveratrol.

    The least-contaminated form of wild salmon is Kodiak chum.
    One way to look at this is to ask how much salmon we would have to eat to meet the experts' recommendation of 3 grams a day of EPA + DHA, or the American Heart Association's recommendation of 2 to 4 grams of EPA + DHA a day for someone who has had coronary artery problem.
    Four ounces of salmon would contain 1.5 to 2 grams of EPA + DHA. Thus, we'd need an 8-ounce serving a day to get 3 to 4 grams of EPA + DHA. At 5 parts of PCB's per billion, that would imply a cancer risk of 1 in 1,000,000 per day, leading to a ten-year integrated cancer risk of 3,650 millionths, or 0.365%. However, one of the key questions that isn't mentioned in the articles is whether or not you'd have to sum up all the cancer risks introduced by all the toxins (PCB's, dieldrin, dioxins, etc.) in fish to arrive at a total cancer risk for all of them.


Future Index