ERF of omega-3 fatty acids
|Moderator:Olli (see all)|
What is the exposure-response function (ERF) of omega-3 fatty acids on several health end points?
library(OpasnetUtils) objects.latest("Op_en5830", code_name = "initiate") # [[ERF of omega-3 fatty acids]] ovariables ERF, threshold. oprint(summary(EvalOutput(ERF_omega3)))
|Obs||Exposure agent||Response||Sex||Exposure||Exposure unit||ER function||Scaling||Threshold||ERF||Description|
|1||DHA||Loss in child's IQ points||Maternal intake through placenta||mg /kg bw /day||ERS||BW||0||-0.07 +- 0.01||Cohen et al. 2005; Gradowska 2013; Standard deviation. Negative loss is a benefit.|
|2||DHA||Loss in child's IQ points||Maternal intake through placenta||mg /day||ERS||None||0||-0.0013 (-0.0018 - -0.0008)||Cohen et al. 2005; also according to Zeilmaker 2013|
|3||Omega3||Coronary heart disease mortality||Ingested intake of EPA+DHA from fish||mg /day||RR||None||0||0.9980 +- 0.000396||Mozaffarian and Rimm 2006; Gradowska 2013 slope = -0.002, SD = exp(-0.002)-exp(-0.002+3.97E-4)|
|4||Omega3||CHD arrythmia mortality||Ingested intake of EPA+DHA from fish||mg /day||Relative Hill||None||200||-0.3||Mozaffarian and Rimm 2006|
|5||Omega3||CHD2 mortality||Ingested intake of EPA+DHA from fish||mg /day||Relative Hill||None||47||-0.17 (-0.25 - -0.088)||Cohen et al 2005. "antiarrhythmic effect". No-exposure is <1 serving/mo, therefore ED50 = two servings per month = 1400 mg/mo = 47 mg/d|
|6||Omega3||CHD2 mortality||Ingested intake of EPA+DHA from fish||mg /day||RR||None||0||0.99951 (0.99934 - 0.99989)||Cohen et al 2005 "antiatherosclerotic effect". 1-0.039*0.01|
|7||Omega3||Stroke mortality||Ingested intake of EPA+DHA from fish||mg /day||Relative Hill||None||47||-0.12 (-0.25 - 0.01)||Cohen et al. 2005|
|8||Omega3||Stroke mortality||Ingested intake of EPA+DHA from fish||mg /day||RR||None||0||0.9998 (0.99934 - 1.00027)||1-0.02*0.01, Cohen et al 2005: −2.0% 95% CI: +2.7% to −6.6%|
|9||Fish||Subclinical brain infarct (one or more) prevalence||Ingested intake of tuna/other fish||≥3 times/week vs. <1/month||RR||None||0||0.74 (0.54 - 1.01)||Virtanen et al. 2008; 95% CI|
|10||Fish||Any prevalent subclinical brain infarct prevalence||Ingested intake of tuna/other fish||Each one serving per week||RR||None||0||0.93 (0.88 - 0.994)||Virtanen et al. 2008; 95% CI|
|11||Fish||Subclinical brain infarct (one or more)incidence||ngested intake of tuna/other fish||≥3 times/week vs. <1/month||RR||None||0||0.56 (0.30 - 1.07)||Virtanen et al. 2008; 95% CI|
|12||Fish||Any incident subclinical brain infarct incidence||Ingested intake of tuna/other fish||Each one serving per week||RR||None||0||0.89 (0.78 - 0.993)||Virtanen et al. 2008; 95% CI|
|13||Fish||Status of cerebral white matter grade score||Ingested intake of tuna/other fish||Each one serving per week||ERS||None||0||0.038||Virtanen et al. 2008; 95% CI|
- ERF publications
|Exposure agent||Trait||Response metric||Exposure route||Exposure metric||Exposure unit||ERF parameter||Threshold||ERF||Description|
|DHA||Child´s IQ||Change in IQ points||Placenta||Maternal intake||mg/kg bw/day||ERS||0||0.07(±0.01)||Cohen et al. 2005; Gradowska 2013|
|Omega3||CHD||Δlog(CHD mortality rate)||Ingestion||Intake from fish||mg/day EPA+DHA||ERS||0||-0.002 (±3.97E-4)||Mozaffarian and Rimm 2006; Gradowska 2013|
|DHA and EPA||Stroke||Incidence||Ingestion||Dietary intake of EPA+DHA||520 mg /day vs. 148 mg /day||RR||0||0.72 (0.54-0.96)||Multivariable- and PCB-adjusted RR (95% CI) Bergkvist et al. 2014|
|Fish||Subclinical brain infarct (one or more)||Prevalence||Ingestion||Intake of tuna/other fish||=3 times/week vs. <1/month||RR||0||0.74(0.54-1.01)||Virtanen et al. 2008|
|Fish||Any prevalent subclinical brain infarct||Prevalence||Ingestion||Intake of tuna/other fish||Each one serving per week||Decrease in RR %||0||7(0.6-12)||Virtanen et al. 2008|
|Fish||Subclinical brain infarct (one or more)||Incidence||Ingestion||Intake of tuna/other fish||=3 times/week vs. <1/month||RR||0||0.56(0.30-1.07)||Virtanen et al. 2008|
|Fish||Any incident subclinical brain infarct||Incidence||Ingestion||Intake of tuna/other fish||Each one serving per week||Decrease in RR %||0||11(0.7-22)||Virtanen et al. 2008|
|Fish||Status of cerebral white matter||Grade score||Ingestion||Intake of tuna/other fish||Each one serving per week||Increase in grade score %||0||3.8||Virtanen et al. 2008|
|Fish||Cerebrovascular disease||Incidence||Ingestion||Intake of fish||2-4 versus ≤1 servings a week||RR||0||0.94 (0.90-0.98)||95% CI, Meta-analysis based on 18 and eight studies Chowdhury et al,. 2015|
|Fish||Cerebrovascular disease||Incidence||Ingestion||Intake of fish||≥5 versus ≤1 servings a week||RR||0||0.88 (0.81-0.96)||95% CI, Meta-analysis based on 18 and eight studies Chowdhury et al,. 2015|
|Fish||Cerebrovascular disease||Incidence||Ingestion||Intake of fish||Increment of two servings per week||Reduced risk||0||0.04 (0.01-0.07)||95% CI, Meta-analysis based on 18 and eight studies Chowdhury et al,. 2015|
|Fish||thrombotic infarction||Incidence||Ingestion||Intake of fish||2 or more times per week vs. <1 serving per month||RR||0||0.49 (0.26-0.93)||95% CI, Iso et al. 2001|
Exposure-response of fish oil intake for MI risk in adults is indexed by variable age. It applies to age categories > 18 years.
Elizabeth Pennisi reports several studies about genetic variation of fatty acid metabolism and links to cardiovascular risk. The overall conclusion is that although these issues are not well understood, there seem to be genetic variation about the health benefits of omega-3 fatty acids.
Aung et al. have conducted a meta-analysis of omega-3 supplement trials with more than 77000 individuals. They found only weak, border-marginal cardiovascular benefits and concluded that the study did not support the use of dietary omega-3 supplements.
The US NCCIH says more about these studies and also: "Moderate evidence has emerged about the health benefits of eating seafood. The health benefits of omega-3 dietary supplements are unclear".
EFSA also has looked at the safety of omega-3 fatty acids and concluded that daily supplemental intakes of 5g of long-chain omega-3 fatty acids raise no safety concerns for adults
The study by Cohen et al. 2005  estimates that increasing maternal docosahexaenoic acid (DHA) intake by 100 mg/day increases child's IQ by 0.13 points D↷. This value represents central estimate while the upper and lower bound for this ERF is 0.08 and 0.18. Triangular distribution is used.
In a recent study, 3660 over 65-year-old individuals were monitored for five years, and the change in small brain infarctions was observed by magnetic resonance imageing. The infaction risk was 25 % lower in those who ate at least three portions of omega-3-rich fish meals per week, and 13 % lower in those who ate one meal per week. 
Fernandez-Jarne et al.  examined the relationship between intake of fish and n-3 PUFA and the risk of first acute myocardial infarction (AMI) in a low risk population from Navarre (Spain). They found that the n-3 PUFA intake has a protective effect on AMI. The adjusted odds ratio (OR) for the second and third tertile of n-3 PUFA intake were 0.44 (95% Cl, 0.21-0.91) and 0.47 (95% Cl, 0.22-1.00), respectively. The trend test was not statistically significant. D↷
Mozaffarian and Rimm  estimated that at intakes between 0 and 250 mg/d, the relative risk of coronary heart disease (CHD) death is lower by 14.6% (95% CI: 8% to 21%) per each 100 mg/d of EPA and DHA intake and that at higher intakes ( > 250 mg/d) the risk reduction is 0.0% (95% CI: -0.9% to 0.8%) per each 100 mg/d.
The ERF of omega-3 fatty acids (DHA+EPA) intake from fish (in unit of mg/kg bw-day) on the CHD mortality is estimated based on information provided in . First, the central estimate and the 95% CI for the change (in this case decrease) in natural logarithm of relative risk (RR) of CHD mortality per unit change in omega-3 fatty acids intake (in unit of mg/day) in both intake intervals were derived. In general, the relationship between the percent change in RR (%RR) associated with c-unit increase in omega-3 fatty acids intake and the incremental change in lnRR (beta) per unit change in omega-3 fatty acids intake is beta = (1/c)*ln((%RR/100)+1). Normal distribution was chosen to describe the uncertainty in the parameter of the log-linear model for RR in each intake interval. For intake of EPA+DHA between 0 and 250 mg/day the mean and the standard deviation of parameter distribution are -0.0016 and 0.0004, for higher intakes 0 and 0.0005. Then, the distribution of ERF of omega-3 fatty acids intake from fish in units of mg/kg bw-day was obtained by multiplying ERFs of omega-3 fatty acids intake measured in mg/day by the body weight of adult.
In cohort studies comparing categories of fish intake the pooled relative risk for cerebrovascular disease was demonstrated based ob 26 prospective cohort studies and 12 randomised controlled trials with aggregate data on 794 000 non-overlapping people and 34 817 cerebrovascular outcomes by a review by Chowdhury et al. 2015 
In a study by Iso et al. 2001  a significant inverse associations between fish intake and age- and smoking-adjusted risk of total stroke, ischemic stroke, and thrombotic infarction, specifically lacunar infarction was found. After further adjustment for other cardiovascular and selected dietary variables, the inverse remained significant for thrombotic infarction and lacunar infarction, with a reduced risk of these stroke subtypes among women who ate fish 2 or more times per week. The multivariate RRs were 0.49 (95% CI, 0.26-0.93; P = .03), and 0.28 (95% CI, 0.12-0.67; P = .004), respectively. We found no excess risk of hemorrhagic stroke, either intraparenchymal or subarachnoid hemorrhage, among women who ate fish frequently.
Marckmann and Grønbaek  made a systematic review of eleven prospective cohort studies in 1999. The cohorts counted a total of 116764 individuals. Of four studies judged to be of high quality, the two largest (n = 44895 and 20051) were performed in populations at low risk of coronary heart disease. They found no protective effect of fish consumption. The other two high-quality studies were relatively small (n = 852 and 1822) and included individuals at higher risk. They both found an inverse relationship between fish consumption and coronary heart disease death, suggesting that 40-60 g fish per day (containing 0.6-0.9 g/d of omega-3) is optimal and associated with a risk reduction of 40-60%. Results of four studies of intermediate quality support that fish consumption is inversely associated with coronary heart disease mortality in high-risk populations only.
Secondary prevention trials were reviewed by Din et al. in 2004.
A large part of omega-3 benefit literature is based on studies on cardiac patients, and there is uncertainty about how well this can be extrapolated to the general population.
Hu et al. (2003)  examined prospectively the association between intake of fish and omega-3 fatty acids and risk of CHD and total mortality among 5103 female nurses with diagnosed type 2 diabetes but free of cardiovascular disease or cancer at baseline. Compared with women who seldom consumed fish (<1 serving/mo), the relative risks (RRs) (95% CI) of CHD adjusted for age, smoking, and other established coronary risk factors were 0.70 (0.48 to 1.03) for fish consumption 1 to 3 times per month, 0.60 (0.42 to 0.85) for once per week, 0.64 (0.42 to 0.99) for 2 to 4 times per week, and 0.36 (0.20 to 0.66) for 5 or more times per week (P for trend=0.002). Higher consumption of fish was also associated with a significantly lower total mortality (multivariate RR=0.48 [0.29 to 0.80] for > or =5 times per week [P for trend=0.005]). Higher consumption of long-chain omega-3 fatty acids was associated with a trend toward lower incidence of CHD (RR=0.69 [95% CI 0.47 to 1.03], P for trend=0.10) and total mortality (RR=0.63 [95% CI, 0.45 to 0.88], P for trend=0.02). #: . Not in the tables --Arja (talk) 08:17, 20 October 2016 (UTC) (type: truth; paradigms: science: relevant comment)
A meta-analysis was performed by Chen et al. (2016)  to evaluate the associations of dietary or circulating n-3 long-chain polyunsaturated fatty acids (LCPUFA) with risk of all-cause mortality. Potentially eligible studies were identified by searching PubMed and EMBASE databases. The summary relative risks (RRs) with 95% confidence intervals (CIs) were calculated using the random-effects model. Eleven prospective studies involving 371 965 participants from general populations and 31 185 death events were included. The summary RR of all-cause mortality for high-versus-low n-3 LCPUFA intake was 0.91 (95% CI: 0.84-0.98). The summary RR for eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) intake was 0.83 (95% CI: 0.75-0.92) and 0.81 (95% CI: 0.74-0.95), respectively. In the dose-response analysis, each 0.3 g/d increment in n-3 LCPUFA intake was associated with 6% lower risk of all-cause mortality (RR = 0.94, 95% CI: 0.89-0.99); and each 1% increment in the proportions of circulating EPA and DHA in total fatty acids in blood was associated with 20% (RR = 0.80, 95% CI: 0.65-0.98) and 21% (RR = 0.79, 95% CI: 0.63-0.99) decreased risk of all-cause mortality, respectively. Moderate to high heterogeneity was observed across our anlayses. Our findings suggest that both dietary and circulating LCPUFA are inversely associated with all-cause mortality. #: . Not in the tables yet --Arja (talk) 08:17, 20 October 2016 (UTC) (type: truth; paradigms: science: relevant comment)
Larsson et al. (2012)  conducted a meta-analysis of prospective studies to summarize available evidence regarding the relation between long-chain omega-3 PUFA intake and stroke. Prospective studies that provided relative risks (RRs) with 95 % confidence intervals (CIs) for the association between dietary long-chain omega-3 PUFA intake and stroke were eligible. A random-effects model was used to combine study-specific results. Eight prospective studies, with 5238 stroke events among 242,076 participants, were included in the meta-analysis. The combined RR of total stroke was 0.90 (95 % CI, 0.81-1.01) for the highest versus lowest category of long-chain omega-3 PUFA intake, without heterogeneity among studies (P = 0.32). Results were similar for ischemic (RR, 0.82; 95 % CI, 0.71-0.94) and hemorrhagic stroke (RR, 0.80; 95 % CI, 0.55-1.15). A statistically significant reduction in total stroke risk was observed in women (RR, 0.80; 95 % CI, 0.65-0.99). This meta-analysis showed no overall association between omega-3 PUFA intake and stroke, but suggests that women might benefit from a higher intake of these PUFAs. #: . Not in the tables yet --Arja (talk) 08:17, 20 October 2016 (UTC) (type: truth; paradigms: science: relevant comment)
The association between intake of fish and n-3 polyunsaturated fatty acids (n-3 PUFA) and the risk of breast cancer was investigated in a meta-analysis and systematic review of prospective cohort studies by Zheng at al. (2013) . Twenty six publications, including 20,905 cases of breast cancer and 883,585 participants from 21 independent prospective cohort studies were eligible. Eleven articles (13,323 breast cancer events and 687,770 participants) investigated fish intake, 17 articles investigated marine n-3 PUFA (16,178 breast cancer events and 527,392 participants), and 12 articles investigated alpha linolenic acid (14,284 breast cancer events and 405,592 participants). Marine n-3 PUFA was associated with 14% reduction of risk of breast cancer (relative risk for highest v lowest category 0.86 (95% confidence interval 0.78 to 0.94), I(2)=54), and the relative risk remained similar whether marine n-3 PUFA was measured as dietary intake (0.85, 0.76 to 0.96, I(2)=67%) or as tissue biomarkers (0.86, 0.71 to 1.03, I(2)=8%). Subgroup analyses also indicated that the inverse association between marine n-3 PUFA and risk was more evident in studies that did not adjust for body mass index (BMI) (0.74, 0.64 to 0.86, I(2)=0) than in studies that did adjust for BMI (0.90, 0.80 to 1.01, I(2)=63.2%). Dose-response analysis indicated that risk of breast cancer was reduced by 5% per 0.1g/day (0.95, 0.90 to 1.00, I(2)=52%) or 0.1% energy/day (0.95, 0.90 to 1.00, I(2)=79%) increment of dietary marine n-3 PUFA intake. No significant association was observed for fish intake or exposure to alpha linolenic acid. #: . Not in the tables yet --Arja (talk) 08:17, 20 October 2016 (UTC) (type: truth; paradigms: science: relevant comment)
- lnRR/ 1 (mg/kg bw-day) change in EPA+DHA intake from fish
- Beneris distributions
- For intakes of EPA+DHA from fish between 0 and 250 mg/day: N(-0.0016,0.0004)*BW
- For intakes of EPA+DHA from fish higher than 250 mg/day: N(0,0.0005)*BW
|Health effect||Relationship||Central estimate||Uncertainty|
|Fish consumption and CHD mortality||ΔRR for some fish consumption vs no fish consumption (<1 serving/month)||−17%||95% CI a: −8.8% to −25%|
|ΔRR per additional serving/week||−3.9%||95% CI a: −1.1% to −6.6%|
|Fish consumption and stroke incidence||ΔRR for some fish consumption vs no fish consumption (<1 serving/month)||−12%||95% CI a: +1.0% to −25%|
|ΔRR per additional serving/week||−2.0%||95% CI a: +2.7% to −6.6%|
|MeHg exposure and cognitive development||ΔIQ per μg/g total Hg in maternal hair||−0.7 pts||Bounds: 0 to 1.5 pts|
|DHA intake and cognitive development||ΔIQ per g/day maternal intake of DHA||1.3 pts||Bounds: 0.8 to 1.8 pts|
a 95% CI is based on the distribution for this coefficient calculated from the regression analysis used to develop the dose–response relationship for CHD or stroke. CHD, coronary heart disease; CI, confidence interval; DHA, docosahexaenoic acid; MeHg, methyl mercury; pts, points; RR, relative risk. Serving size was 100 g.
Here we need to convert the serving size to omega-3 intake. It depends on the average omega-3 content in the diets of the patients in the studies, and it is not known to us. Therefore, we may assume that it is higher than in lean fish (0 - 0.5 %) and lower than in fatty fish (1-2 %), i.e. say 0.7 % or 700 mg per serving. Therefore, the published RR changes (per servings/week) must be multiplied by 1 per (servings * 700 mg/serving / (week * 7 d/week) = 0.01 * RR changes per mg/d. CHD2 is used as the trait to prevent double counting with the other ERFs based on Mozaffarian and Rimm.
In addition, Cohen concluded that the ERFs for CHD and stroke are non-linear with a larger reduction in risk between non-consumers and some-consumers (the limit defined as 1 serving per month). In addition, a linear incremental benefit was estimated for intakes more than 1 serving per week. This results in two independent ERFs where the low-dose "antiarrhythmic" effect follows Relative Hill function and the high-dose "antiatherosclerotic" effect follows RR function. Cohen assumed a negative correlation between these, but this is not easy to implement with the current HIA ovariables and therefore we ignore the correlation; this results in an increase of the estimated uncertainty in the model.
In a recent large cohort study by Engeset et al. (2015)  no associations were seen for consumption of total fish, lean, or fatty fish and either total mortality or cause-specific mortality among men; broadly similar results were obtained for women. The statistically significant associations found in the non-calibrated analyses disappeared in the calibrated analyses and bootstrap analyses.
library(OpasnetUtils) d <- opbase.data("Op_en5830") d$Obs <- NULL colnames(d) <- gsub(" ", "_", colnames(d)) d$Result <- ifelse(d$Result == "", "0", as.character(d$Result)) ERF_omega3 <- Ovariable("ERF_omega3", data = d[d$Observation == "ERF", colnames(d) != "Observation"]) threshold_omega3 <- Ovariable("threshold_omega3", data = d[d$Observation == "Threshold", colnames(d) != "Observation"]) objects.store(ERF_omega3, threshold_omega3) cat("Ovariables ERF, threshold stored.\n")
- ERF of methyl mercury
- A press release from the University of Kuopio (in Finnish)
- Reviews by Henna Karvonen in Beneris:
- Impact of fish consumption on nutrient intakes
- Cardiovascular dose-responses of fish consumption
- Mental health dose-responses of fish consumption
- Immunological disease dose-responses of fish consumption
- Diabetes and glucose dose-responses of fish consumption
- Developmental dose-responses of fish consumption
- Cancer dose-responses of fish consumption
- Bone dose-responses of fish consumption
- All-cause mortality dose-response of fish consumption
- Pennisi E. Is fish oil good for you? Depends on your DNA. Science 17 September 2015 
- Aung T, Halsey J, Kromhout D, et al. Cardiovascular Disease Risks. Meta-analysis of 10 Trials Involving 77 917 Individuals. JAMA Cardiol. 2018;3(3):225-233. doi:10.1001/jamacardio.2017.5205 
- EFSA. Scientific Opinion on the Tolerable Upper Intake Level of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA). EFSA Journal 2012;10(7):2815. doi:10.2903/j.efsa.2012.2815 
- Cohen, J.T., PhD, Bellinger, D.C, PhD, W.E., MD, Bennett A., and Shaywitz B.A. 2005b. A Quantitative Analysis of Prenatal Intake of n-3 Polyunsaturated Fatty Acids and Cognitive Development. American Journal of Preventive Medicine 2005;29(4):366–374).
- Fish consumption and risk of subclinical brain abnormalities on MRI in older adults Jyrki K. Virtanen, David S. Siscovick, Will T. Longstreth, Lewis H. Kuller, Dariush Mozaffarian Neurology 2008;71:439–446.
- Fernandez-Jarne E, Garrido FA, Gutierrez AA, Arrillaga CDF, Martinez-Gonzales MA. Dietary intake of n-3 fatty acids and the risk of acute myocardial infarction: a case-control study. (In Spanish) 2002;118:121–5.
- Mozaffarian D., Rimm E.B., Fish intake, contaminants, and human health. Evaluating the risks and the benefits. (Reprinted) JAMA, 2006. Vol 296, No. 15
- Chowdhury et al. 2012 Association between fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review and meta-analysis. 
- Iso et al. 2001. Intake of fish and omega-3 fatty acids and risk of stroke in women. JAMA. 2001;285(3):304-312. doi:10.1001/jama.285.3.304. 
- Marckmann P, Gronbaek M. (1999). Fish consumption and coronary heart disease mortality. A systematic review of prospective cohort studies. European Journal of Clinical Nutrition; 53(8):585-590.
- Din JN, Newby DE, Flapan AD. Science, medicine, and the future - Omega 3 fatty acids and cardiovascular disease - fishing for a natural treatment. British Medical Journal 2004; 328(7430):30-35. Intranet file
- Hu FB, Cho E, Rexrode KM, Albert CM, Manson JE (2003). Fish and long-chain omega-3 fatty acid intake and risk of coronary heart disease and total mortality in diabetic women. Circulation 107(14):1852-7.
- Chen GC, Yang J, Eggersdorfer M, Zhang W, Qin LQ. (2016). N-3 long-chain polyunsaturated fatty acids and risk of all-cause mortality among general populations: a meta-analysis. Sci Rep. 6:28165. doi: 10.1038/srep28165.
- Larsson SC, Orsini N, Wolk A. (2012). Long-chain omega-3 polyunsaturated fatty acids and risk of stroke: a meta-analysis. Eur J Epidemiol. 27(12):895-901. doi: 10.1007/s10654-012-9748-9.
- Zheng JS, Hu XJ, Zhao YM, Yang J, Li D. (2013). Intake of fish and marine n-3 polyunsaturated fatty acids and risk of breast cancer: meta-analysis of data from 21 independent prospective cohort studies. BMJ 27;346. doi: 10.1136/bmj.f3706.
- Cohen JT, Bellinger DC, Shaywitz BA. A quantitative analysis of prenatal methyl mercury exposure and cognitive development. Am J Prev Med. 2005 Nov;29(4):353-65. 
- Engeset et al. 2014: Fish consumption and mortality in the European Prospective Investigation into Cancer and Nutrition cohort