Good fish/bad fish: a composite benefit-risk by dose curve.

Balancing risks and benefits of fish consumption is now a high visibility public health topic. Many studies identify health benefits of eating fish, both for prenatal development and adult cardiovascular conditions, partly attributed to omega-3 (n-3) polyunsaturated fatty acids or PUFAs). Many reports raise concerns about methylmercury and polychlorinated biphenyl effects on the developing fetal brain (although adults, too, can manifest methylmercury effects). Most reports and advisories focus on recreational or subsistence fish, but the vast majority of people obtain most or all of their fish from commercial sources. Our analysis of the nine most common fish in New Jersey markets, yielded a weighted average methylmercury concentration of 0.23 ug/g (ppm wet weight). There are great disparities in the amount and distribution of both PUFAs and contaminants) in different fish species. Recognizing that both benefits and harm must be related to dose, we propose a compound dose-response curve, currently based on limited data, to identify a zone of benefit, above the benefit threshold and below the harm threshold. The duration of pregnancy and birth weight improve at a benefit threshold of about 8-15 g/day maternal fish intake. Meta-analyses reveal adult cardiovascular benefits around 7.5-22.5 g/day bracket (assuming an 8 ounce/227 g typical meal), yielding a midpoint also at 15 g/day, but this is an artifact of the intake stratification. Benefit asymptotes are harder to extract, but are above 45 g/day, and in some studies exceed 100g/day. Using the EPA Reference Dose of 0.1 ug/kg day as a methylmercury threshold, The fish intake threshold for harm converts to 27 g/day (for a selection of common commercial fish averaging 0.23 ppm MeHg) to 65 g/day for someone choosing fish low in MeHg (0.1 ppm). However, these are worst case thresholds since the RfD includes uncertainty factors. Some people eat much more than 65 g/day. The shape of the dose-benefit and dose-harm curves require better data for estimating thresholds and asymptotes, which will impact the composite curve. We propose this approach clarifies the kinds of data needed to improve risk communication on "what should I eat". Benefits from fish consumption are confounded by socioeconomic class and/or by the avoidance of more harmful foods that fish replaces, which may be as important a benefit mechanism as the PUFA content. Additional studies with better dose-reconstruction are needed and large scale intervention studies are desirable.