European Journal of Nutrition & Food Safety,
Request from the Norwegian Food Safety Authority (NFSA)
In September 2013, the Norwegian Food Safety Authority requested VKM to update relevant parts of the benefit-risk assessment of fish in the Norwegian diet published by VKM in 2006. The background for the request was new knowledge and data on the content of some nutrients and contaminants both for wild and farmed fish since 2006. The proportion of vegetable ingredients used in farmed fish feed has in recent years increased, and new national dietary surveys for adults and children have been conducted.
The Norwegian Food Safety Authority referred to VKM’s report from 2006, which pointed out that the positive impact of fish consumption on public health was especially due to the content of polyunsaturated fatty acids and vitamin D in fish. Further, VKM concluded that the contaminants that could pose a potential risk to public health through fish consumption mainly were methylmercury, dioxins and dioxin-like PCBs (dl-PCBs).
The request included a reassessment of fish consumption in Norway with focus on specific nutrients; n-3 fatty acids (eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), docosahexaenoic acid (DHA), vitamin D, and the minerals iodine and selenium, and on specific contaminants; mercury, dioxins and dl-PCBs. VKM was asked to address the main changes in the use of raw materials in farmed fish feed and how these affect the levels of nutrients, mercury, dioxins and dl-PCBs and in fish feed. Further, VKM was asked to address to what extent levels of nutrients and contaminants in fish have changed since 2006, to describe these changes and estimate the human intake of the substances in question on the basis of recent dietary data. VKM was also requested to consider the benefits of eating fish with regard to the intake of nutrients and the risks associated with the intake of mercury, dioxins and dl-PCBs and comment on whether this change the conclusions from the report in 2006. Additionally, on the basis of updated knowledge, VKM was asked to comment whether other substances, like pesticide and residues of veterinary medicinal products, could affect the conclusions with regard to the impact on public health.
The Norwegian Food Safety Authority and the Directorate of Health will use the updated assessment as a basis for public recommendations concerning the consumption of fish and fish products.
How VKM has Addressed the Request
The VKM appointed a working group consisting of VKM members and external experts to answer the request. Several of the scientific panels of VKM reviewed the report during its preparation. The Scientific Steering Committee of VKM has given their final assessment and approval of the current report.
In the current report, VKM has mainly used data from national surveillance and monitoring programs for nutrient and contaminant concentrations in fish feed, farmed fish and wild caught fish, but occurrence data have also been derived from peer reviewed articles.
VKM has estimated fish consumption in three population groups (2-year-olds, adults and pregnant women). The estimated fish consumption was compared to national dietary guidelines.
To assess health effects of fish consumption, the current estimated fish intakes were also compared with assessments done by recognised international bodies and results from epidemiological studies addressing possible associations between fish consumption and specific health outcomes. Literature searches were done to identify relevant epidemiological studies. VKM has not systematically assessed reviews or meta-analyses nor individual studies for weight of evidence, but merely summarised the studies retrieved from the literature search. It was considered being beyond the scope of this assessment to review individual studies included in reviews or meta-analyses.
Furthermore, based on current fish consumption in the various population groups, intake of nutrients and exposure to contaminants from fish were estimated. For benefit characterisation of the specific nutrients the estimated nutrient intake was compared with national recommendations of nutrients intake and for EPA and DHA a comparison was also done with European recommendations. For risk characterization of contaminant exposure from fish, VKM used health based guidance values set by international risk assessment bodies (WHO, EFSA).
VKM noted that the request from NFSA was restricted to fish, whereas the VKM report in 2006 included both fish and other seafood.
VKM focused on specific nutrients and contaminants as requested by the NFSA. In addition, VKM also commented on other substances that could affect the risk assessment, such as residues of veterinary medicinal products including residues of antibiotics, new contaminants from fish feed like the pesticide endosulfan, polycyclic aromatic hydrocarbons (PAHs), mycotoxins, the synthetic antioxidants ethoxyquin, butylhydroxyanisole (BHA) and butylhydroxytoluene (BHT), as well as environmental contaminants like brominated flame retardants and perfluorated organic compounds.
In the Norwegian diet fish is important source of well-balanced proteins, and important nutrients such as EPA and DHA, vitamin D, iodine and selenium. On the other hand, fish is also a source of exposure to chemical contaminants like dioxins, PCBs and mercury.
Over the last 10 years there has been a great change in raw materials used in fish feeds, and in 2013 terrestrial plant proteins and vegetable oils accounted for 70% of the feed. The changes in concentrations of nutrients and contaminants in fish feed for farmed Atlantic salmon and trout are reflected in changed concentrations and compositions of the same nutrients and contaminants in the farmed fish fillet.
The current national dietary guideline is to eat fish as dinner meals 2-3 times per week for all age groups, representing 300-450 g fish per week for adults, including at least 200 g fatty fish, such as salmon, trout, mackerel and herring. Fish is also recommended as bread spread. Further, a daily supplement of vitamin D to infants from 4 weeks of age is recommended, and if this supplement is taken as cod liver oil it will in addition ensure an adequate supply of EPA and DHA.
The present benefit-risk assessment is comprised of three elements, i.e. benefit assessment, risk assessment and benefit-risk comparison. This methodology is in accordance with the guidance given by EFSA in 2010.
Fish Consumption in Norway and Comparison with National Dietary Guidelines
VKM has used information about fish consumption from more recent national dietary surveys among 2-year-olds (Småbarnskost 2007) and adults at 18-70 years of age (Norkost 3, 2010/2011), as well as information for pregnant women who answered the Norwegian Mother and Child Cohort Study (MoBa2, 2002-2008) food frequency questionnaire. The national food consumption survey Ungkost 2000, which covers the age groups 4-, 9-, and 13-year-old children, was considered too old to be used and it is therefore not known if their fish consumption patterns have changed, neither in amount consumed nor type of fish consumed.
Even though there are methodological differences between the dietary surveys used in 2006 and 2014, the amount of fish consumed appears to be unchanged for all population groups. Furthermore, in 2014, lean fish and fatty fish contribute with about 60 and 40 percent, respectively, of the total fish consumption, which is similar to 2006.
Given a portion size of 150 g fish, the average adult eats fish equivalent to 2-3 dinner servings per week and the average pregnant woman eats fish equivalent to 1-2 dinner servings per week, while the average two-year-old eats fish equivalent to 1-2 dinner servings per week given a portion size of 75 g. The table below describes fish intake in the selected populations.
Fish consumption (expressed as raw fish), mean grams (g) per week in 2-year-olds (Småbarnskost 2007, n=1674), adults (Norkost 3, n=1787) and pregnant women (MoBa, n=86277)
VKM concludes that of the different population groups, only adults (18-70 years of age) with an average or higher fish consumption reach the national food based dietary guidelines for total fish consumption. Mean total fish consumption and fatty fish consumption in children (2-year-olds) and pregnant women, as well as the mean fatty fish consumption in adults are lower than recommended. In pregnant women and 2-year-olds, fish consumption is too low to meet the food based dietary guidelines.
Health Effects of Fish Consumption
VKM is of the opinion that according to epidemiological studies, the net effects of the present average fish consumption in Norway for adults including pregnant women is beneficial for specific cardiovascular diseases (particularly cardiac mortality, but also with regard to ischaemic stroke, non-fatal coronary heart disease events, congestive heart failure and atrial fibrillation), as well as for optimal neurodevelopment of foetus and infants. Furthermore, VKM is of the opinion that those with fish consumption less than one dinner serving per week may miss these beneficial effects.
The health benefit of fish consumption is reported from 1-2 dinner servings per week and up to 3-4 dinner servings per week. For higher fish intake per week, the limited number of consumers in epidemiological studies does not allow for drawing firm conclusions about the actual balance of risk and benefit. More knowledge is needed to reveal the beneficial mechanisms of fish consumption.
Benefit Characterisation of Nutrients in Fish
VKM is of the opinion that there has been minor or no changes of the composition and concentrations of nutrients in wild caught fish since 2006.
Due to replacement of fish oil and fish protein with plant proteins and vegetable oils in feed for farmed fish, the concentrations in farmed Atlantic salmon with regard to EPA, DPA and DHA, and selenium are about 50 and 40% respectively, of the corresponding levels in 2006, while the concentration of vitamin D appears unchanged. The level of iodine in farmed Atlantic salmon was low in 2006, and is still low compared to lean fish. The level of n-6 fatty acids is about 4-fold higher than in 2006.
VKM has estimated the contribution from fish to the recommended daily intakes of certain nutrients. Fish is the major source of EPA+DPA+DHA, but for vitamin D, iodine and selenium, other sources in addition to fish are needed in order to meet the recommendation. Fish is not a major dietary source of n-6 fatty acids. The contribution of dietary n-6 fatty acids from farmed salmon compared to the overall dietary intake of n-6 fatty acids is low (less than 3%).
VKM concludes that with current average consumption of fish, the contribution of EPA and DHA from fish will reach the European recommended intake of EPA+DHA for adults and 2-year-olds. For pregnant women the average EPA+DHA intake is insufficient to meet the European recommendation for this group. However, the average intake of DHA is sufficient to meet the national intake recommendation for pregnant women.
For vitamin D, current average fish consumption contributes approximately 20% of the national recommended intakes for adults but less for pregnant women and 2-year-olds.
Furthermore, with current average fish consumption, low intakes of selenium and iodine from fish relative to the national recommended values may be complemented by intake from other dietary sources.
VKM has made various scenarios to foresee how possible changes in fish consumption pattern and amounts will affect the contribution from fish to recommended intakes of specific important nutrients. According to the scenarios, increasing the consumption of fatty fish will increase the intakes particularly of vitamin D, EPA+DPA+DHA, while increasing consumption of lean fish will increase the intakes particularly of iodine. Furthermore, VKM notes that the choice of fatty fish species, e.g. farmed Atlantic salmon, mackerel and herring is also of importance for nutrient intake due to differences in nutrient content.
Risk Characterisation of Undesirable Substances in Fish
The available concentration data of contaminants in wild fish is not suitable for time-trend analyses. A rough comparison of contaminant concentrations between 2006 and 2014 indicates minor or no changes in concentrations of mercury, dioxins and dl-PCBs in wild fish species. However, for dioxins and dl-PCBs, a decreasing environmental time-trend is expected to be reflected also in wild fish species.
Due to replacement of fish oil and fish protein with plant proteins and vegetable oils in farmed fish feed, the concentrations of dioxins and dioxin-like PCBs, and mercury have changed in farmed Atlantic salmon. VKM concludes that the current concentrations of dioxins and dl-PCBs, and mercury in farmed Atlantic salmon are reduced to about 30 and 50%, respectively, of the corresponding levels in 2006.
VKM has estimated the dietary exposure to contaminants from fish based on mean levels in different fish species and compared the exposure levels with the relevant health based guidance levels, tolerable weekly intakes (TWIs). A tolerable intake is the amount of a substance, or substance group, which can be consumed safely throughout a person's lifetime without appreciable risk of adverse health effects. Tolerable intakes incorporate safety margins, in order to protect all parts of the population.
VKM concludes that with the present mean concentration of mercury in fish on the Norwegian market and the present fish consumption in Norway, the methylmercury exposure from fish is below the tolerable weekly intake (TWI) of 1.3 µg/kg bw/ week for more than 95% of the population of 2-year-olds, adults and pregnant women. This exposure represents a negligible risk and is of no concern.
With the present mean level of dioxins and dl-PCBs in fish on the Norwegian market and the present fish consumption in Norway, high fish consumption (the 95th percentile) contributes with up to 50%, 19%, 67% of the TWI of 14 pg TEQ/kg bw/ week for adults, pregnant women and 2-year-olds respectively. Daily consumption of cod liver oil or fish oil (which is common in all population groups) in amounts as suggested on the product will in addition contribute with 0.8 to 16% of the TWI, depending on the body weight. With the present TWI and taking into consideration that fish and fish products are significant sources to dioxins and dl-PCBs in the Norwegian diet, VKM concludes that the exposure from fish to dioxins and dl-PCBs represents negligible risk and is of no concern.
VKM is of the opinion that the present exposure to residues of veterinary medicinal products including residues of antibiotics in farmed fish in the Norwegian diet is of no concern since the levels are very low and often not detectable even with sensitive analytical methods.
For new contaminants in fish feed like the pesticide endosulfan, polyaromatic hydrocarbon (PAHs) and mycotoxins, VKM is of the opinion that the concentrations in farmed fish in the Norwegian diet are likely not a food safety issue since the concentrations are very low and often not detectable even with sensitive analytical methods.
Regarding the environmental contaminants brominated flame retardants; VKM refers to the conclusions in a risk assessment from EFSA in 2011 that the health risk associated with the current exposure to these compounds is low. The amount of fluorinated compounds such as PFOS and PFOA in the Norwegian diet is much lower than what is tolerable according to an EFSA assessment in 2008.
VKM has made various scenarios to foresee how possible changes in fish consumption pattern and amounts will affect the exposure from fish to TWIs of methylmercury, and dioxins and dl-PCBs. Fish is the only source for methylmercury exposure from foods, whereas exposure to dioxins and dl-PCBs also comes from other foods than fish. Based on these scenarios, where only exposure to dioxins and dl-PCBs from fish were taken into consideration, VKM is of the opinion that fish consumption in line with the food-based dietary guideline of 300-450 g fish, hereof 200 g fatty fish per week, does not lead to exposures to dioxins and dl-PCBs or methylmercury from either fatty or lean fish exceeding the respective TWIs, and is therefore, from a contaminant exposure perspective, of no concern.
However, since there are other food sources in the Norwegian diet that contribute to exposure to these contaminants, VKM performed a simple model estimate of weekly intake of dioxins and dl-PCBs in adults from various amounts of farmed salmon and other foods. Based on this scenario, VKM is of the opinion that there is negligible risk associated with eating farmed Atlantic salmon with the present mean concentrations of dioxins and dl-PCBs. The TWI is not exceeded when consuming amounts equivalent to 1400 g farmed salmon weekly for adults (representing 9 weekly dinner servings). Neither is the TWI exceeded when exposures to dioxins and dl-PCBs from other foods and cod liver oil are taken into consideration. In comparison, an adult can consume about 800 g mackerel weekly (representing 5 weekly dinner servings) with current mean concentration of dioxins and dl-PCBs without exceeding TWI. From a contaminant exposure perspective consumption of farmed salmon is of no concern. This also applies for commercially available wild caught fish like mackerel.
Benefit – Risk Comparison
Following a comprehensive assessment of the scientific literature on the positive health effects of fish consumption and the contribution from fish to intake of beneficial compounds as well as exposure to hazardous contaminants in Norway, VKM concludes that the benefits clearly outweighs the negligible risk presented by current levels of contaminants and other known undesirable substances in fish. Furthermore, adults including pregnant women with fish consumption less than one serving per week may miss the beneficial effects on cardiovascular diseases and optimal neurodevelopment in the foetuses and infants. In contrast to the conclusion in 2006, VKM concludes that there is no reason for specific dietary limitations on fatty fish consumption for pregnant women.
This benefit-risk assessment is composed of several different parts. Various databases are used, including data on levels of nutrients and contaminants in fish feed and fish which may all contain uncertainties which in turn may influence the overall assessment. Furthermore, there may be uncertainties in the estimated fish consumption data retrieved from the dietary food surveys and there may be weaknesses in the epidemiological studies about health effects of fish consumption. Despite some limitations in assessing the fish consumption and the uncertainties related to the estimated intakes of nutrients and exposures to contaminants from fish and fish products, VKM concludes that the intake and exposure estimates presented in this opinion are within realistic ranges for each study population. VKM compared intakes of nutrients with national recommended intake values and exposures to contaminants with internationally recognised health based guidance values (tolerable intakes). Likewise, the benefits for health associated with fish consumption were also evaluated by international bodies, and the uncertainties in these assessments were not evaluated by VKM. VKM considers the overall uncertainty in the outcome of the present assessment on benefit and risk of fish consumption in Norway to be low.