fish intake contaminants and human health jama

8/19/2019 Fish Intake Contaminants and Human Health JAMA

1/16

CLINICIAN’S CORNERCLINICAL REVIEW

Fish Intake, Contaminants, and Human HealthEvaluating the Risks and the BenefitsDariush Mozaffarian, MD, DrPH

Eric B. Rimm, ScD

SINCE THE PUBLICATION OF PIO -neering studies demonstratinglow rates of death from coro-nary heart disease (CHD)

amongGreenland Eskimos, 1 fish(usedhereinto refer tofinfish orshellfish) hasbeen considered a healthy food. Dur-ing ensuing years, evidence from sev-eralresearchparadigms—includingani-mal-experimental, observational, andclinicalstudies—further supported thishypothesis and identified 2 long-chain n-3 polyunsaturated fatty acids(n-3 PUFAs), eicosapentaenoic acid(EPA) and docosahexaenoic acid(DHA), as the likely active constitu-ents. 2-20 DHAalsoappears important forneurodevelopmentduringgestationandinfancy. 21-26 Conversely, concern hasarisen over potential harm from mer-

cury, dioxins, and polychlorinatedbiphenyls (PCBs) present in some fishspecies. 27-34 The public is faced withseemingly conflicting reports on therisks andbenefitsof fish intake, result-ing in controversy and confusion overtheroleof fish consumption inahealthydiet. 35,36 To elucidate the relative risksand benefits, we reviewed the scien-tific evidence for adverse and benefi-cial healtheffects of fish consumption.

EVIDENCE ACQUISITIONIdentification of StudiesA myriad of exposures and outcomeshave been related to fish consump-tion; we focused on populations and

topics for which evidence and con-cern are greatest . We searchedMEDLINE,governmental reports, andsystematic reviews and meta-analyses

to identify reports published throughApril2006 evaluating (1) intake of fishor fish oil and risk of cardiovascularevents andmortality, (2) effects of me-thylmercury and fish oil on early neu-rodevelopment, (3) risks of methyl-mercury for cardiovascular andneurologicoutcomes in adults, and (4)healthrisks ofdioxins andPCBs in fish.

MEDLINE search terms were ( Fishor n-3 PUFA or omega-3) and ( coro-nary or cardiac or cardiovascular or mor-

See also Patient Page.

CME available online atwww.jama.com

Author Affiliations: Channing Laboratory, Depart-ment of Medicine, Brigham and Women’s Hospital,andHarvardMedicalSchool; andDepartments of Epi-demiology and Nutrition, Harvard School of PublicHealth, Boston, Mass.Corresponding Author: Dariush Mozaffarian, MD,DrPH,Harvard School of Public Health, 665 Hunting-ton Ave, Bldg 2, Room 315, Boston, MA 02115([email protected]).ClinicalReview Section Editor: Michael S. Lauer, MD.We encourage authors to submit papers for consid-eration as a Clinical Review. Please contact MichaelS. Lauer, MD, at [email protected].

Context Fish (finfish or shellfish) may have health benefits and also contain con-taminants, resulting in confusion over the role of fish consumption in a healthy diet.Evidence Acquisition We searched MEDLINE, governmental reports, and meta-analyses, supplemented by hand reviews of references and direct investigator con-tacts, to identify reports published through April 2006 evaluating (1) intake of fish or fish oil and cardiovascular risk, (2) effects of methylmercury and fish oil on early neu-rodevelopment, (3) risks of methylmercury for cardiovascular and neurologic out-comes in adults, and (4) health risks of dioxins and polychlorinated biphenyls in fish.We concentrated on studies evaluating risk in humans, focusing on evidence, whenavailable, from randomized trials and large prospective studies. When possible, meta-analyses were performed to characterize benefits and risks most precisely.Evidence Synthesis Modest consumption of fish (eg, 1-2 servings/wk), especiallyspecies higher in the n-3 fatty acids eicosapentaenoic acid (EPA) and docosa-hexaenoic acid (DHA), reduces risk of coronary death by 36% (95% confidence in-terval, 20%-50%; P .001) and total mortality by 17% (95% confidence interval,0%-32%; P=.046) and may favorably affect other clinical outcomes. Intake of 250 mg/dof EPA and DHA appears sufficient for primary prevention. DHA appears beneficialfor,andlow-level methylmercurymay adversely affect, early neurodevelopment.Womenof childbearing age and nursing mothers should consume 2 seafood servings/wk, lim-iting intake of selected species. Health effects of low-level methylmercury in adultsare not clearly established; methylmercury may modestly decrease the cardiovascular benefits of fish intake. A variety of seafood should be consumed; individuals with veryhigh consumption ( 5 servings/wk) should limit intake of species highest in mercurylevels. Levels of dioxins and polychlorinated biphenyls in fish are low, and potential

carcinogenic and other effects are outweighed by potential benefits of fish intake andshould have little impact on choices or consumption of seafood (women of childbear-ing age should consult regional advisories for locally caught freshwater fish).Conclusions For major health outcomes among adults, based on both the strengthof the evidence and the potential magnitudes of effect, the benefits of fish intake ex-ceed the potential risks. For women of childbearing age, benefits of modest fish in-take, excepting a few selected species, also outweigh risks. JAMA. 2006;296:1885-1899 www.jama.com

©2006 American Medical Association. All rights reserved. (Reprinted) JAMA, October 18, 2006—Vol 296, No. 15 1885

wnloaded From: http://jama.jamanetwork.com/ by a Oregon State University User on 09/01/2013


2/16


3/16

ders, 41,42 and asthma or inflammatorydisorders 43,44) were not reviewed inthis report.

EVIDENCE SYNTHESISBenefits of Fish Intake

Cardiovascular Outcomes. Deathfrom CHD—ie, documented or sus-pected fatal myocardial infarction—and sudden death—ie, a suddenpulseless condition of presumed car-diac etiology—are clinically definedentities often sharing the final com-mon pathway of ventricular arrhyth-mia, often ischemia-induced ventricu-lar fibrillation. The evidence fromprospective studies and randomizedtrials 2-4,6-17,45-51 suggests that consump-tion of fish or fish oil lowers risk of C H D d e a t h a n d s u d de n d e a t h(FIGURE 1 and FIGURE 2). Across dif-ferent studies (Figure 1), comparedwith little or no intake, modest con-sumption ( 250-500 mg/d of EPAand DHA) lowers relative risk by25% or more. Higher intakes do notsubstantially further lower CHD mor-tality, suggesting a threshold of effect. 52 Pooling all studies, this pat-tern was clearly evident (Figure 2).At intakes up to 250 mg/d, the rela-tive risk of CHD death was 14.6%lower (95% confidence interval [CI],

8% to 21%) per each 100 mg/d of EPA and DHA, for a total risk reduc-tion of 36% (95% CI, 20% to 50%).At higher intakes, little additionalrisk reduction was present (0.0%change per each 100 mg/d; 95% CI,−0.9% to 0.8%). This thresholdeffect explains findings among Japa-nese populations, 17,50 in whom highbackground fish intake (eg, median900 mg/d of EPA and DHA 50) is asso-ciated with very low CHD death rates(eg, 87% lower than comparable

Western populations9,17

), and addi-tional n-3 PUFA intake predicts littlefurther reduction in CHD death;thus, most of the populat ion isalready above the threshold for maxi-mum mortality benefits. Comparingdifferent types of fish, lower riskappears more strongly related tointake of oily fish (eg, salmon, her-

ring, sardines), rather than lean fish(eg, cod, catfish, halibut). 10,15 Fishintake may modestly affect other car-diovascular outcomes, but evidenceis not as robust as for CHD death(TABLE 1).17,50,53-66

n-3 PUFAs influenceseveral cardio-vascular riskfactors. 18,19,43,49,50,60-75,79-84 Ef-fects occur within weeks of intake andmay result from altered membrane flu-idity and receptor responses followingincorporation of n-3 PUFAs into cellmembranes 76,77 anddirectbinding ofn-3PUFAs to intracellular receptors regu-latinggene transcription. 78 The hetero-geneity of the effects of fish or fish oilintake on cardiovascular outcomes islikely related to varying dose and timeresponses of effects on the risk factors

(FIGURE 3). At typical dietary intakes,antiarrhythmiceffects predominate, re-ducing risk of sudden death and CHDdeath within weeks. At higher doses,maximum antiarrhythmic effects havebeen achieved, butother physiologicef-fects maymodestly impact other clini-cal outcomes (possibly requiring yearsto produce clinical benefits). For in-

stance, nonfatal myocardial infarctionmay not be significantly affected bylower doses or shorter durations of in-take but may be modestly reduced byhigher doses or prolonged intake (eg,1.8 g/d for 5 years 17).

Heterogeneity of clinical effectsmayalso be related to differingpathophysi-ologies of theclinical outcomes.For in-stance, disparate pathophysiologies of primary ventricular fibrillation (oftenischemia-induced) vs recurrent ven-triculartachyarrhythmias (ectopicor re-entrant) mayexplain stronger effects of n-3PUFAson the former. Similarly, bio-logical differences in development of atherosclerosisvs acuteplaque rupture/ thrombosis vs arrhythmia would ac-count for heterogeneous effects of n-3

PUFAs on plaque progression vs non-fatal myocardial infarction vs CHDdeath. Consumption of fish may dis-place that of other foods, such as meatsor dairy products, in the diet. How-ever, this likelyaccounts for little of theobservedhealth benefits, because foodsreplaced would be highly variableamong individuals and across cul-

Figure 2. Relationship Between Intake of Fish or Fish Oil and Relative Risks of CHD Death inProspective Cohort Studies and Randomized Clinical Trials

0 500 1000 1500 2000 2500 3000EPA + DHA Intake mg/d

1.2

0.4

0.6

0.8

1.0

0.2

R e

l a t i v e

R i s k ∗

The relationship between intake of fish or fish oil and relative risk of coronary heart disease (CHD) death in apooled analysis of the prospective studies and randomized trials shown in Figure 1, evaluated nonparametri-cally using restricted cubic splines 38,39 and adjusted for each within-study relationship. Given the much higher referencegroupintakes in some studies, thereferencerelative riskwas scaledby 0.7for studies with referencegroup intakes between 150-500 mg/d of eicosapentaenoic acid (EPA) docosahexaenoic acid (DHA) (n=5)and by 0.6 for studies with reference group intakes 500 mg/d (n= 1) based on spline relationships prior toincluding these studies; exclusion of these studies, or of the few groups with intakes 1000 mg/d, had littleeffect on the pooled spline relationship. A significant threshold effect ( P .001) was evident at intake of 250mg/d: between 0 and 250 mg/d, mortality risk was lower by 14.6% (95% confidence interval [CI], 8% to21%) per each 100-mg/d greater intake (total risk reduction, 36%; 95% CI, 20% to 50%; P .001), while athigherintakes, riskwas notfurtherlowered (0.0% change pereach 100mg/d; 95%CI, −0.9% to 0.8%; P=.94).* Relative risks in the control and intervention groups (for randomized trials) or relative risks in the referencegroup and multivariable-adjusted relative risks in the comparison groups (for cohort studies).

FISH INTAKE, CONTAMINANTS, AND HUMAN HEALTH




4/16

tures, andmodest intakeof such foodsis not associated with CHD risk. 85

Total Mortality. n-3 PUFAs moststrongly affect CHD death 5,9,14-16,18 andare unlikely to affect appreciablyothercauses ofmortality. Effectson totalmor-

tality in a population would thereforedependon theproportionofdeathsdueto CHD, ranging from one quarter of deaths in middle-age populations 86 toone half of deaths in populations withestablishedCHD. 9 Thus, givena 36%reduction in CHD death (Figure 2), in-take of fishor fishoil would reduce totalmortality by between 9% (36% re-duction 25% CHD deaths) to 18%(36% reduction 50% CHDdeaths),oran average of 14% in mixedpopula-tions. This is consistent with a meta-analysis of randomized trials through2003 3,9,51,56,57,87-93 that found a nonsig-nificant 14% reduction in total mortal-itywith n-3PUFAs(pooled relative risk,0.86; 95% CI, 0.70 to1.04). 94 Whenweadded additional placebo-controlled,double-blind, randomized trials 60-62 per-formed since 2003, marine n-3 PUFAsreduced total mortalityby 17% (pooledrelativerisk, 0.83; 95% CI,0.68 to1.00;P=.046) ( FIGURE 4). This can be com-pared to effects of statins on total mor-tality—a 15% reduction—in a meta-analysis of randomized trials (pooled

relative risk, 0.85; 95% CI, 0.79 to0.92). 95Neurologic Development. DHA is

preferentially incorporated intothe rap-idlydeveloping brain during gestationandthe first 2 yearsof infancy, concen-trating in gray matter and retinalmem-branes. 26 Infants can convert shorter-chain n-3 fatty acids to DHA, 96 but itis unknown whether such conversionis adequate for the developing brain inthe absence of maternal intake of DHA.22,25

Effects of maternal DHA consump-tion onneurodevelopmenthave been in-vestigated in observational studies andrandomizedtrials,with heterogeneity inassessedoutcomes(visual acuity, globalcognition, specific neurologic do-mains) and timing of DHA intake (ges-tational vs nursing). In a meta-analysisof 14 trials, DHA supplementation

Table 1. Summary of Evidence for Effects of Consumption of Fish or Fish Oilon Cardiovascular Outcomes

OutcomeClinicalEffect

Strengthof Evidence Comment

CHD mortalityCHD deathSudden death

35% decrease50% decrease

StrongStrong

Probable threshold of effect—most risk reduction occurs

with modest intake( 250 mg/d EPA DHA),with little additional benefitwith higher intakes 2-4,6-17,45-51 *

Ischemic stroke 30% decrease Moderate Strong evidence from prospectivecohort studies 53,54 ; no RCTs

Nonfatal CHDNonfatal MI Modest benefit? Equivocal Possible benefits at very high intakes

( 2 g/d n-3 PUFAs) 17,50

Progression of atherosclerosis

Modest benefit? Equivocal Mixed results in cohort studies 55and RCTs 56-58

Postangioplastyrestenosis

Modest benefit? Equivocal Possible benefits in a meta-analysisof RCTs 59

Recurrent ventriculartachyarrhythmias

Modest benefit? Equivocal Mixed results in 3 RCTs 60-62

Atrial fibrillation 30% decrease Limited Mixed results in 2 cohort studies 63,64 ;benefit in 1 RCT 65

Congestive heart failure 30% decrease Limited Benefit in 1 prospective cohort study 66

Abbreviations: CHD, coronary heart disease; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid;MI, myocardial infraction; n-3 PUFA, n-3 polyunsaturated fatty acid; RCT, randomized clinical trial.

* See Figure 1.

Figure 3. Schema of Potential Dose Responses and Time Courses for Altering Clinical Eventsof Physiologic Effects of Fish or Fish Oil Intake

R e

l a t i v e

S t r e n g

t h o

f E f f e c

t

0 500 1500 250020001000

EPA + DHA Intake, mg/d

TYPICAL SUPPLEMENTAL

DOSES

Antiarrhythmia Weeks

Clinical Effect Time Course To Alter Clinical Events

Triglyceride-Lowering Months to YearsHeart Rate–Lowering MonthsBP–Lowering Months to Years

Antithrombosis Weeks

TYPICAL DIETARY

DOSES

Therelativestrengthof effectis estimatedfrom effects of eicosapentaenoic acid (EPA) docosahexaenoic acid(DHA) on each risk factor and on the corresponding impact on cardiovascular risk. 70-72,79-84 For example, doseresponse for antiarrhythmic effects is initially steep with a subsequent plateau, and clinical benefits may occur within weeks, while dose response for triglyceride effects is more gradual and monotonic, and clinical benefitsmay require years of intake. At typical Western levels of intake (eg, 750 mg/d EPA DHA), the physiologiceffects most likely to account for clinical cardiovascular benefits include (1) modulation of myocardial sodiumand calcium ion channels, reducing susceptibility to ischemia-induced arrhythmia; 18,19 and(2) reduced left ven-tricular workload andimprovedmyocardial efficiencyas a resultof reduced heart rate, lower systemic vascular resistance, and improved diastolic filling. 67-72,80 At higher levels of intake seen with fish oil supplementation or in Japanese populations 49,50 ( 750 mg/d EPA DHA), maximum antiarrythmic effects have been achievedand clinically relevant effects occur on levels of serum triglycerides 79 and possibly, at very high doses, throm-bosis.75 Potentially important effects on endothelial, 73 autonomic, 74 and inflammatory 43 responses are notshownbecause dose responsesand timecourses of such effects on clinical risk arenot well established.Effects arenotnecessarily exclusive: eg, antiarrythmic effects may be partly mediated by effects on blood pressure (BP) or heart rate.


1888 JAMA, October 18, 2006—Vol 296, No. 15 (Reprinted) ©2006 American Medical Association. All rights reserved.



5/16

improved visual acuity in a dose-dependent manner. 23 Results for cogni-tive testing are less consistent, possiblydue to differences in neurologic do-mains evaluated 21,25,26 ; a quantitativepooled analysisof8 trials estimated that

increasing maternal intake of DHA by100 mg/d increased child IQ by 0.13points (95% CI, 0.08 to 0.18). 24 Mosttrialsevaluated effects ofmaternalDHAintake during nursing,rather thanpreg-nancy. In a trial among 341 pregnantwomen, treatmentwithcod liveroil fromweek 18 until 3 months postpartum in-creasedDHA levelsin cordbloodby50%and raised mental processing scores, ameasure of intelligence, at age 4 years. 97This is consistent with observationalstudies showingpositiveassociationsbe-tween maternal DHA levels or fish in-takeduringpregnancy andbehavioralat-tention scores, visual recognitionmemory, and language comprehensionin infancy. 98-100 Thus, while dose re-sponses and specific effects require fur-ther investigation, these studies to-gether indicate that maternal intake of DHA is beneficial for early neurodevel-opment.

Risks of MercuryMercury is a reactiveheavymetalemit-ted from natural sources (volcanoes)and human sources (coal-fired elec-tric power plants, gold mining, insti-tutional boilers, chlorine production,

andwaste incineration).101

Fromthe at-mosphere, mercury cycles from rain-water into lakes and oceans, where itis converted by microbial activity intoorganicmethylmercury. Inorganic mer-cury is poorly absorbed following in-gestion, and elemental mercury doesnot readily cross tissue barriers. Incontrast, methylmercury is readily ab-sorbed andactivelytransported intotis-sues. 27 Thus, methylmercury bioaccu-mulates in aquatic food chainsandhasgreater potential toxicity than inor-ganicmercury. 27,28,30 Concentrationsof methylmercury in aquatic species de-pend on levels of environmental con-tamination and on the predatory na-ture andlifespan of thespecies. Larger,longer-living predators (eg, sword-fish, shark)have higher tissueconcen-trations, while smaller or shorter-livedspecies (eg, shellfish, salmon) havevery low concentrations ( TABLE 2).122

Preparation methods have little im-pact on methylmercury content. 27

Health effects of very high mercuryexposure followingoccupational or in-dustrial accidents are well docu-mented, including paresthesias, ataxia,

andsensory abnormalitiesin adults,anddelayed cognitive and neuromusculardevelopment following in utero expo-sure. 27,131 Toxicity appears related tobinding of methylmercury to sulfhy-dryl groups of enzymes, ion channels,andreceptors, resulting in inhibition of antioxidant systemsand production of free radicals and reactive oxygen spe-cies.27,29 Health effects of chronic low-level mercury exposure—ie, that seenwith fish consumption—are less wellestablished. The public is aware of thepotential harm from mercury in fish butlacks clear understanding of who is atrisk or which seafood species containmercury. 35,36 Wereviewtheevidence forhealth effects below.

Methylmercury andNeurodevelopmentMethylmercurycrosses theplacenta, andfetal exposure correlateswith maternal

Figure 4. Risk of Total Mortality Due to Intake of Fish or Fish Oil in Randomized Clinical Trials

5.01.00.2

Relative Risk (95% CI)

Source

Brouwer et al, 62 2006Brox et al, 87 2001Burr et al, 3 1989Burr et al, 51 2003Eritsland et al, 88 1996Gruppo Italiano, 9 1999Johansen et al, 89 1999Kaul et al,90 1992Leaf et al,91 1994Leaf et al,61 2005Nilsen et al,92 2001Raitt et al,60 2005Sacks et al, 56 1995Singh et al, 93 1997von Schacky et al, 57 1999

Overall

% Weight3.90.3

18.724.4

2.926.0

0.70.30.44.64.62.30.39.90.6

100.0

Relative Risk(95% CI)

0.57 (0.24-1.38)0.17 (0.01-4.05)0.71 (0.55-0.92)1.15 (0.99-1.34)1.23 (0.43-3.51)0.86 (0.76-0.97)0.33 (0.03-3.18)0.28 (0.01-6.78)0.20 (0.01-4.18)1.09 (0.49-2.46)1.00 (0.45-2.24)0.40 (0.12-1.32)0.32 (0.01-7.57)0.56 (0.34-0.91)0.50 (0.05-5.39)

0.83 (0.68-1.00)

The size of the shaded squares indicates each trial’s contribution (inverse-variance weight) to the pooled estimate (dotted line) and 95% confidence interval (CI; dia-mond), determined by random effects meta-analysis. 37 Intake of fish or fish oil reduced total mortality by 17% ( P=.046), with evidence for heterogeneity betweentrials (P=.04 for heterogeneity). If 2 trials with methodologic concerns 51,93 were excluded, the pooled relative risk was 0.83 (95% CI, 0.74-0.92; P .001) with littleevidence for heterogeneity ( P=.75). A recently reported trial of fish oil among Japanese individuals 17 was not included in the primary analysis due to very high fishintake in the reference group (estimated eicosapentaenoic acid docosahexaenoic acid intake, 900 mg/d) which would obviate mortality benefits of additional fish oilintake. When this trial was added to the secondary analysis, the pooled relative risk was 0.87 (95% CI, 0.76-0.99; P=.048; P=.29 for heterogeneity).





6/16

exposure. 132 Marked neurodevelop-mental abnormalities occur in childrenfollowing very high gestational expo-sure, 27,131 such as from maternal con-sumption of highly contaminated fish(10-30 ppm mercury) from industri-

ally pollutedMinimataBay, Japan, in the1950s, or of contaminated grain in Iraqin 1971 (maternal intake, 710-5700ug/kg per day; 18-598 ppm mercury inmaternal hair). More typical methylmercury exposures are substantially

lower: among US women of childbear-ing age, median (10th-95th percen-tiles) levels ofmercury inhair were0.19(0.04-1.73) ppmoveralland0.34(0.09-2.75)ppmamongwomen consuming3or more servings of fish per month. 133

Table 2. Levels of n-3 Fatty Acids and Contaminants in Commonly Consumed Fish, Shellfish, and Other Foods *

EPA DHA,mg/serving

(Serving Size†)

EPA DHA,mg/100 g

(3.5 oz)Selenium,

µg/g (ppm)Mercury,

µg/g (ppm)PCBs,

ng/g (ppb)Dioxins, TEQpg/g (ppt)‡

FDA action level 33,102 NA NA NA 1.0 2000 None§

Fish

Anchovy 1165 (2 oz) 2055 0.68 0.05 0.35 (1997-1998) 103

Catfish, farmed 253 (5 oz) 177 0.15 0.05 50 (1997) 104 0.53 (1995-1997) 1050.51 (1996) 1062.09 (1995-1996) 1071.65 (1995) 108

Cod, Atlantic 284 (6.3 oz) 158 0.38 0.10 0.05 (1995-1997) 1050.15 (1995-1996) 107

Fish burger, fast food 337 (2.2 oz) 546 0.17‡ 0.05 8 (2001) 109 0.01 (2001) 1100.11 (2001) 109

Fish sticks, frozen 193 (3.2 oz) 214 0.17 0.05 0.04 (2001) 110

Golden bass (tilefish), Gulf of Mexico 1358 (5.3 oz) 905 0.52 1.45

Golden bass (tilefish), Atlantic 1358 (5.3 oz) 905 0.52 0.14

Halibut 740 (5.6 oz) 465 0.47 0.25 1.00 (1995-1997) 105

Herring, Atlantic 1712 (3 oz) 2014 0.47 0.05 0.97 (1995-1998) 105

Mackerel, Atlantic 1059 (3.1 oz) 1203 0.52 0.05 0.87 (1997-1998) 1030.32 (1995-1998) 105

Mackerel, King 618 (5.4 oz) 401 0.47 0.73

Mahimahi 221 (5.6 oz) 139 0.47 0.15

Pollock, Alaskan 281 (2.1 oz) 468 0.43

0.05 0.01 (1998)105

0.24 (1998) 111

Salmon, farmed 4504 (6 oz) 2648 0.41 0.05 21 (2001-2003) 112 0.50 (2001-2003) 11215 (2002) 113 0.87 (2002) 11440 (2002) 115 ¶ 0.45 (2002) 11526 (2001) 110 0.33 (2001) 11025 (2001) 116 0.50 (1997) 10551 (1999-2000) 117 ¶38 (1999) 116

Salmon, wild 1774 (6 oz) 1043 0.46 0.05 3 (2002) 115 ¶ 0.03 (2002) 1150.5 (2002) 113 0.34 (2002) 114

5 (2000) 117 ¶

Sardines 556 (2 oz) 982 0.53 0.05 57 (2001-2003) 112 0.44 (2001-2003) 11222 (2002) 118 0.18 (2002) 118

0.60 (1995) 105

Shark 585 (3 oz) 689 0.34 0.99

Snapper 546 (6 oz) 321 0.49 0.19

Swordfish 868 (3.7 oz) 819 0.62 0.98

Trout 581 (2.2 oz) 935 0.15 0.07 11 (2002) 113 0.56 (2002) 113 #0.32 (2002) 1140.74 (1998-2000) 1190.35 (1998) 105

Tuna, light (skipjack) 228 (3 oz) 270 0.80 0.12 45 (2001) 110 0.02 (1995-1998) 105

Tuna, white (albacore) 733 (3 oz) 862 0.66 0.35 100 (2001-2003) 112 0.23 (2001-2003) 112

(continued)





7/16

Table 2. Levels of n-3 Fatty Acids and Contaminants in Commonly Consumed Fish, Shellfish, and Other Foods * (cont)EPA DHA,mg/serving

(Serving Size†)

EPA DHA,mg/100 g

(3.5 oz)Selenium,

µg/g (ppm)Mercury,

µg/g (ppm)PCBs,

ng/g (ppb)Dioxins, TEQpg/g (ppt)‡

Shellfish

Clams 241 (3 oz) 284 0.64 0.05 3 (2001-2003) 112 0.05 (2001-2003) 112

2 (2002) 118 0.05 (2002) 1180.10 (1997-1998) 103

Crab 351 (3 oz) 413 0.40 0.09 6 (2002) 113 0.55 (2002) 113 #1.05 (1998) 111

Lobster 71 (3 oz) 84 0.43 0.31 0.69 (1998) 1110.12 (1997-1998) 103

Mussels 665 (3 oz) 782 0.90 0.15 7 (2001-2003) 112 0.09 (2001-2003) 1120.8 (2002) 113 0.11 (2002) 113 #

2 (2002) 118 0.07 (2002) 1180.39 (1998) 1050.45 (1995-1996) 107

Oysters 585 (3 oz) 688 0.77 0.05 17 (2001-2003) 112 0.46 (2001-2003) 1120.8 (2002) 113 0.19 (2002) 113 #

Scallops 310 (3 oz) 365 0.28 0.05 0.16 (1998) 111

Shrimp 267 (3 oz) 315 0.40 0.05 2 (2002) 118 0.06 (2002) 113 #0.2 (2002) 113 0.11 (2002) 118

0.06 (2001) 1100.19 (1995-1997) 1050.08 (1995-1996) 107

Other Foods

Beef 0 0 0.19 0 22 (2001) 110 0.13 (2001) 1100.27 (1995) 120

Bologna 0 0 0.14 0 0.16 (2001) 1100.29 (1995) 120

Butter, regular 0 0 0.05 0 70 (2001) 110 0.22 (2001) 1100.31 (1995-1996) 1070.66 (1995) 120

Cheese 0 0 0.22 0 0.25 (2001)110

0.77 (1998) 1110.34 (1995) 120

Chicken 0 0 0.23 0 32 (2001) 110 0.02 (2001) 1100.20 (1995) 120

Eggs 22 (1 egg) 43 0.23 0 19 (2001) 110 0.05 (2001) 1100.52 (1998) 1110.31 (1995) 120

Milk, whole 0 0 0.02 0 0.01 (2001) 1100.12 (1995-1996) 1070.13 (1995) 120

Pork 0 0 0.34 0 18 (2001) 110 0.10 (2001) 1100.23 (1995) 120

Abbreviations: DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; FDA, US Food and Drug Administration; NA, not applicable; PCB, polychlorinated biphenyl; ppb, partsper billion; ppm, parts per million; ppt, parts per trillion; TEQ, toxic equivalence.

*

Based on data from US Department of Agriculture (USDA),121

Food and Drug Administration (FDA),110

Environmental Protection Agency,122

and other103-109,111-120,123-126

sources. These values may vary due to methodologic, geographic, temporal, and fish-to-fish differences. Levels of PCBs and dioxins may overestimate current levels because contami-nant levels in most foods, including fish species, are decreasing over time 33,110,112,127,128 (eg, TEQs decreased by 33%-81% in meats 127 and 66%-77% in salmon and tuna fish 112between 1995 and 2003); year of sampling is given in parenthesis.

†Based on USDA serving sizes: 2 oz anchovies or sardines; 1 fillet catfish, cod, mackerel, mahimahi, snapper, or trout; 1 ⁄ 2 fillet halibut, king mackerel, pollock, or golden bass;6 oz salmon; 3 oz herring, shark, shellfish, or tuna; 1 piece (3.75 oz) swordfish. 121

‡The sum of dibenzodioxins (PCDDs) debenzofurans (PCDFs) (nondetects = 1/2 LOD when multiple estimates available).§Due to “numerous questions and uncertainties regarding scientific data on and analysis of dioxin risk.” 129

Forthe same specific species, there areminimaldifferencesin nutritional or contaminant content of cannedvs fresh salmonor tuna. However, differentspeciesare typicallycannedvs sold fresh. For salmon, differences between species are small compared with differences between farmed and wild salmon. For tuna, canned light (skipjack) tuna and freshyellowfin/ahi tuna are more similar overall, while canned white (albacore) tuna and fresh bluefin tuna are more similar overall.

¶Measured including the fish skin; levels may be lower in the edible portion. 130#Includes dioxin-like PCBs.





8/16

These exposure levels do not producesymptomatic neurodevelopmentaldefi-cits, butseveralprospective studieshaveevaluatedwhethersubclinicaleffects,de-tectable with specialized testing, mightoccur. 98,100,134-140 Among children fromthe Faroe Islands, 134,135 New Zea-land, 136,137 and Poland, 138 higher gesta-tional exposure to mercurywas associ-ated with lower scores on someneurologic tests(eg,finger tapping,nam-ing tests) but not others. In contrast,higher gestational exposure to mer-cury was associated with higher scoreson some neurologic tests among Sey-

chellois children.139,140

In a US cohort,gestationalmaternalfishintakewasposi-tivelyassociated with, butmercurylev-els in hair were negatively associatedwith,visual recognition memory scoresin infancy, 98 indicating possible oppos-ing effects of overall fish consumption(ie, providing DHA) and methylmer-cury exposure. In a British cohort, ges-tational mercury exposure was not as-sociated with, but maternal and infantfish intake was associated with, im-proved neurodevelopmental scores. 100

Other studies did not detect consistentassociations between gestational expo-sure to mercury and neurologic testscores during childhood. 141

Comparisonsacross studies are lim-ited by heterogeneity of study designs(prospective vs cross-sectional), mer-cury assessment methods, neurologictests used, timing of assessment (in-

fancy vs childhood), and statisticalmethods. Some analyses are also lim-ited by multiple statistical testing (eg,

30 neurologic variables) or incom-plete adjustment for other potential riskfactors. Randomized trials to test ef-fects of reducing low-level methylm-ercury exposure during gestation havenot been performed. Nevertheless,givenassociationswith some lowerneu-rologic test scores in some studies, andclinical neurotoxicity of methylmer-curyfollowing high-level accidental ex-posures, it is prudent to conclude thatsubclinical neurodevelopmental defi-

cits mayoccurat lower exposure levels.Based on this, the EnvironmentalProtection Agency determined a refer-ence dose, ie, theallowable upper limitof daily intake, for methylmercury of 0.1 ug/kg per day ( 50 µg/wk for a70-kg woman, calculated from thelower 95% confidence limit at whichgestational exposure to mercury mayproduce abnormal neurologic testscores, multiplied by a 10-fold uncer-tainty factor) 132 and published a fo-cusedadvisory forwomenof childbear-

ing age, nursing mothers, and youngchildren. 142 Theadvisoryspecificallyad-vises such individuals to avoid shark,swordfish, goldenbass, andking mack-erel (each containing 50 µg meth-ylmercuryperserving) (Table2); to eatup to 12 oz/wk (2 average meals) of avariety of fish and shellfish lower inmercury, including up to 6 oz/wk of al-

bacore tuna (30 µg methylmercuryperserving); and to consult local adviso-ries for locally caught freshwater fish.This advisory was not intended for thegeneral population,because the impor-tance of this reference dose to healtheffects in adults wasunclear. 143 We re-view the evidence for such effects be-low.

Health Effects of Methylmercuryin AdultsCardiovascular Disease. Several stud-ies144-148 have evaluated the relation-ship between mercuryexposureandin-

cidence of cardiovascular disease(FIGURE 5). The conflicting results pro-vide inconclusive evidence for cardio-vascular toxicity of mercury. Notably,in the 2 studies observing higher riskwith higher mercury levels, the neteffect of fish consumption wasstill ben-eficial: greater mercury exposure less-ened the benefit associated with con-sumption of fish or n-3 PUFAs but didnot increase overall risk. 146,148,150 Thus,the principal question may not bewhether consumption of mercury-

containing fish increases cardiovascu-lar risk but whether consumption of such fish woulddecreaserisk even fur-ther if mercury were not present. Thiswould be most true for oily fish spe-cies containing higher amounts of n-3PUFAs (ie, most mercury-containingocean fish), compared with lean fresh-water fish. This is an important public

Figure 5. Multivariate Risk of Incident Coronary Heart Disease (CHD) With Higher Levels of Mercury Exposure

No. of Events8778

684470

282

Study DesignProspectiveProspectiveRetrospectiveProspective

Prospective

Source Ahlqwist et al, 144 1999Hallgren et al,145 2001Guallar et al,146 2002

Yoshizawa et al, 147 2002

Virtanen et al, 148 2005

Overall

Relative Risk(95% CI)

0.71 (0.4-1.26)0.51 (0.21-1.24)2.16 (1.09-4.29)1.03 (0.65-1.65)

1.66 (1.2-2.29)

1.12 (0.71-1.75)

5.01.00.2

Relative Risk (95% CI)

Relative risk and 95% confidence intervals (CIs) are shown comparing the highest to the lowest quantile of mercury exposure after adjustment for other risk factors.In 2 studies in Sweden, higher mercury levels were associated with trends toward lower risk, 144,145 but findings may have been limited by relatively few numbers ofevents. In 2 larger European studies, positive associations between mercury levels and CHD risk were reported. 146,148 A large US study observed no association, 147 butmost participants were dentists, in whom mercury levels in part represented occupational exposure to inorganic mercury, 149 which may be less toxic than methylmer-cury in fish.27,28,30 The overall pooled relative risk (dotted line) and 95% CI (diamond), estimated using inverse-variancerandom-effects meta-analysis, 37 was1.12 (95%CI, 0.71-1.75; P=.62), with significant heterogeneity between studies ( P=.008).





9/16

health issue, which requires balancingpotentially attenuated benefits of fishintake dueto presence of mercury withthe costs and practicality of reducingmercury contamination in fish spe-cies. Nevertheless, this should not ob-

scure evidence for net cardiovascularbenefits of fish consumption, particu-larly fish richer in n-3 PUFAs.

NeurologicOutcomes. Veryhigh me-thylmercury exposure from accidents(eg,Minimata) 27,151 orprolongedhigh in-takesofmercury-containingfish (eg, 1-2fish servings/d, including species highin mercury, for 10 years 152) can pro-ducesensorimotor symptomsin adults,most commonly paresthesias, whichare often reversible when mercuryexposure is reduced. Whether lowerexposures produce neurologic abnor-malities in adults is not clear. Cross-sectional studies have evaluatedassociations between mercury levelsin hair or blood and subclinical neuro-logic function in adults. Among Ama-zon basin and Quebec Cree individu-als,bothpositiveand inverseassociationswere seen between mercury levels andsomeneurologicmeasures, 153-155 butfind-ings were limited by minimal adjust-ment forother risk factors andmultipletesting(typically 20-30neurologictestsor participant subgroups). Among US

adults, mercury levels were associatedwith lower visual memory scores(P=.01) but better motor and manualdexterity scores ( P=.02) among 20 dif-ferent outcomes evaluated. 156 Amongelderly Swedish adults, no associationswere foundbetween mercurylevels andcognitivefunction. 157 Thus, it isunclearwhetherlow-levelmethylmercuryaffectssubclinical neurologic outcomes inadultsand, ifso,whatquantitiesor dura-tions of exposure are necessary. Con-versely, a growing body of evidence

suggests that fish consumption mayfavorablyaffect clinical neurologic out-comes in adults, including ischemicstroke, 53 cognitive decline and demen-tia,40 anddepression and other neurop-sychiatric disorders. 41,42

P o s s i b l e M e r c u r y - S e l e n i u mInteraction. Health effects of mercurymaypartly result from selenoprotein in-

activation,which might bemitigated byadequate intakeof selenium,an essen-tial dietary trace element. 158-161 Sele-nium also may reduce tissue accumu-lation of mercury in fish 1 6 2 an dhumans. 163 Seafood species are rich di-

etary sources of selenium.121

A protec-tive effect of selenium may partly ac-count for conflicting results of studiesof mercury exposure and neurodevel-opmental indices in children 160 and of mercury exposure and risk of CHD. 164A potential selenium-mercury interac-tion would have important publichealth implications, and additional in-vestigation is warranted.

Risks of PCBs and DioxinsPCBsare syntheticorganochlorine com-pounds previously used in industrialandcommercial processes. 34 Dioxins—commonly referring to dibenzodiox-ins and dibenzofurans—are orga-nochlorine by-products of wasteincineration, paper bleaching, pesti-cide production, and production of polyvinyl chlorideplastics. 33 Manufac-ture and processing of PCBs was pro-hibited in 1977, 34 and regulatory andindustry efforts have reduced dioxinemissions by more than 90% since1987. 33 Nevertheless, these contami-nantspersist for long periods in the en-

vironment, and thus while levels aresteadily declining, 33,110,112,127,128 PCBs anddioxins continue to be present in lowconcentrationsinmany foods (Table2).

Cancer Risks. Animal experimentsand some evidence in humans indi-cate that PCBs and dioxins are carci-nogenic, possibly related to effects onthe aryl hydrocarbon receptor, a tran-scription factor affecting gene expres-sion. 32,165 Multiple congeners (struc-tural variants) of PCBs and dioxinsexist. Potential toxicities of foods are

calculated using toxic equivalence(TEQ):thesumof each congener’s levelin the food multiplied by that conge-ner’s toxic equivalency factor (stan-dardized against 2,3,7,8-tetrachlorod-ibenzo-p-dioxin). In the United States,PCBs comprise 28% and dioxins 72%of total TEQexposure. 120 Amongadults,major dietary sources of PCBs and di-

oxinsare beef,chicken,andpork (34%of total TEQ); dairy products (30%);vegetables (22%); fish and shellfish(9%); andeggs (5%). 120 Dietarysourcesare similar for children. 120

Although major sources of expo-

sure to PCBs and dioxins are meats,dairy products, andvegetables,consid-erable attention has been given to fishsources (Table 2). When PCBs and di-oxins were measured in farmed andwildsalmon, 115,166 levelswere similar tothose in several other foods (Table 2).Farmed and wild salmon also con-tained substantial levels of n-3 PU-FAs: 4504 and 1774 mg of EPA andDHA per 6 oz, respectively. 166 Cancerrisks and CHD benefits were evalu-atedin a quantitative risk-benefitanaly-sis, assuming regular farmed or wildsalmon intake to provide 1000mg/d of EPA and DHA over a 70-year life-time. 167,168 Per100000individuals,con-sumption of farmed vs wild salmonwould result in 24 vs 8 excess cancerdeaths, respectively, while consump-tion of either farmed or wild salmonwould result in 7125 fewer CHDdeaths. 167 We further evaluated age-specific estimates, based on allocationof lifetime cancer risks 167 (adjusted forcompeting risks) by age-specific can-cer mortality 169 and 25% reduction in

age-specific CHD mortality.169

For allages evaluated (25-34 to 85 years),CHD benefits outweighed cancer risksby 100- to 370-fold for farmed salmonand by 300- to more than 1000-fold forwild salmon.

Notably, estimated CHDbenefits arebased on prospective studies and ran-domized trials in humans (Figures 1and 2); estimated cancer risks includea 10-fold safety factor andare basedonanimal-experimental data and limitedstudies inhumansat highdoses. 168 Can-

cer estimates also assumed lifetimesalmon consumption to provide 1000mg/d of EPA and DHA (eg, four 6-ozservings of wild salmon every week for70 years). However, CHD mortality re-duction maybeachieved with lower in-take: 250 mg/d (Figures 1 and 2), orone6-oz wild salmon servingperweek.At this intake, CHD benefits would be





10/16

largely unchanged ( 7125 fewer CHDdeaths), while lifetime cancer riskwould decrease by 75% (6 and 2 es-timated deathsper100 000 lifetimes forfarmed and wild salmon, respec-tively). Consistent with these very lowcancer risks, prospectivestudies in hu-mans have seen little evidence for ef-fects of fish intake on cancer risk. 170

Other Risks. PCBs and dioxinsmay have noncancer risks in adults,

such as immune system or neuro-logic effects. 32-34 Conversely, fish con-sumption may also have other ben-efits, possibly lowering risk of othercardiovascular outcomes (Table 1),dementia, 40 neuropsychiatric disor-ders, 41,42 and inflammatory disor-ders. 43,44 If present, such additionalpossible risks would have to exceedadditional possible benefits by morethan 100-fold to meaningfully alterthe present estimates of risks vs ben-efits. PCB content in fish can be

reduced 12% to 40% by trimmingbelly and back fat during filletingand by not consuming the skin. 130Also, contaminant levels are typicallymeasured in unprepared foods, andcooking may reduce PCB and dioxincontent. 106

Prenatal (but not postnatal) expo-sure to PCBs and dioxins has been

associated with childhood neurode-velopmental deficits in several, 171-177though not all, 178,179 studies. Becausemost exposure ( 90%) generallycomes from meat, dairy, and veg-etable sources, 120,180 this concern isnot specific to fish consumption, par-ticularly since fish also containspotentially beneficial DHA. However,w o m e n c o n s u m i n g 1 o r m o r eservings/d of commercial freshwater

fish or consuming locally caughtfreshwater fish from highly contami-nated inland sources may be moregreatly exposed to PCBs and diox-ins 180 and should consult regionaladvisories.

Related ConsiderationsCosts. We evaluated potential costs of consuming 250 mg/d of EPA andDHA from fish (FIGURE 6). The dailycost was as low as 9 cents, or 63cents/wk. For combinations of differ-

ent types of salmon; salmon andtuna; or salmon, tuna, anchovies,and sardines, the average cost was 37cents/d ($2.59/wk) or less. Actual(net) costs would be lower becauseintake of fish would replace intake of other foods.

Supplements. Fish oilcapsules con-tain 20% to 80% of EPA and DHA by

weight (200-800 mg/g 185,186 ), little tonomercury, 187 and variable levels of PCBs(0-450 ng/g, 116,188 ) and dioxins (0.2-11TEQ pg/g114,189 ). Given small amountsof fish oil consumed (1-3 g/d), expo-sure to PCBs and dioxins from fish oil

intakeis low. “Functionalfoods” supple-mented with EPA and DHA (eg, dairyproducts, salad dressings, cereals) canalso provide reasonable intake to indi-viduals not consumingseafood. 190 Com-paredwith supplements, fishintakealsoprovides potentially beneficial protein,vitamin D, and selenium. 121

Commercial Preparation. Commer-cially-preparedfried fishmeals from fastfood restaurants or supermarket fro-zen sections 123,124 are often made us-ing white-meat fish (lower in n-3 PU-FAs) 27,123 and prepared with partiallyhydrogenated oils (containing transfats) or oils reused for multiple fryingcyc les ( in t roduc ing oxida t ive/ deteriorative products 191). Higher car-diovascular risk seen with fried fish in-take 1 5 , 5 4 , 6 3 , 6 6 may re la te to th i sunfavorable balance of benefit vs harm(lower levels of EPA and DHA; higherlevels of trans fats/deteriorative prod-ucts) or to residual confounding fromother lifestyle factors. While further re-search isneeded, it appearsunlikely thatmost commercially prepared fried fish

meals lower cardiovascular risk.n6:n3 Ratio. Ecologic studies andlimited animal-experimental data sug-gest that linoleic acid (18:2n-6) maycounteract potential benefitsofn-3fattyacids, 192,193 but this hypothesis has notbeen supportedby clinical trials or pro-spective studiesin humans. 16,194 A muchgreater changein thedietary ratioofn-6fatty acids ton-3 fatty acids can beprac-tically achievedby increasing intake of n-3s (eg, going from no intake of oilyfish to 1 serving/wk) compared with

lowering intake of n-6s (which arewidelyconsumed in cooking oils, saladdressings, and prepared foods). Thus,for most populations, attention to rela-tive intakes ofn-6 vsn-3 fatty acids maybe less important than simply increas-ing n-3 intake.

Aquaculture. Concerns exist aboutsustainability of some aquaculture and

Figure 6. Estimated Costs of Consuming the equivalent of 250 mg/d EPA DHA From Fish

Fish Preparation

0 10 20 30 40oz/wk

Ounces per Week to AchieveIntake of 250 mg/d EPA + DHA

FilletFarmed Atlantic SalmonCannedWild Pink Salmon

FilletWild King Salmon

FilletWild Silver SalmonCanned Albacore TunaCannedLight TunaCanned AnchovyCannedSardineFrozenShrimpFrozenScallops

FilletCatfishFilletCod

0 0.50 1.00 1.50 2.00 2.50Cost, $

Cost to Consume250 mg/d of EPA + DHA

Costs were calculated for commonly consumed seafood species, based on retail prices (averaging the mostcommonly sold items in each of 6 US cities in the east, midwest, and south from a national online grocerystore 181 or,for wildking andsilver salmon, from onlineretailers 182-184 ) and on species-specific eicosapentaenoicacid (EPA) docosahexaenoic acid (DHA) content. 121 Least expensive was canned pink salmon (9 cents/250mg of EPA DHA); the average cost per 250 mg of EPA DHA for these 12 types of seafood was 92 cents.

The corresponding ounces per week needed to achieve 250 mg/d of EPA DHA is also shown.





11/16

commercial fishingpractices. 195-197 Con-versely, aquaculture contributes toglobal fish production, 198 and sustain-ability concerns are not unique toaquaculture or fishing but also exist foragricultural, forestry, freshwater, atmo-

spheric, and energy resources.195,199,200

Some progress has been made, such aschanges in fish feeds to reduce depen-dence on fish meal or oil. 195 Given theimportance of n-3 PUFAs for health,balance must be achieved between en-vironmental andeconomic concerns toallow sustainable, financially viableaquaculture and commercial fish-ing.195,196,199

Plant Sources. Alpha-linolenic acid(ALA)(18:3n-3) isann-3fattyacid pre-sent in flaxseed, canola, soybeans, andwalnuts. 121 In humans, ALA is con-verted to EPA in small quantities (inwomen more than men); further con-version to DHA is very limited. 201 Con-sumption of ALA (eg, 2-3 g/d) may re-duce cardiovascular risk 202 or affectneurodevelopment, butbenefitsare lessestablished compared with those forEPA and DHA.

Optimal IntakesOptimal intake of n-3PUFAs may varydepending on population and out-come of interest. In the general popu-

lation, 250 mg/d of EPA and DHA is areasonable target intaketo reduceCHDmortality. Because dietary n-3 PUFAspersist for weeks in tissue mem-branes, 203 this can be converted to aweekly intake of 1500-2000mg.Thiscorresponds to one 6-oz serving/wk of wild salmon or similar oily fish,ormorefrequent intake of smaller or less n-3PUFA–rich servings (Table 2). For in-dividualswith CHD, 1000 mg/d of EPAandDHA iscurrentlyrecommended toreduceCHDmortality. 204,205 Ouranaly-

sis suggests that lower doses may besuf-ficient, but given this population’shigher risk andthat most data are fromprimaryprevention studies, a target in-take of 500 to 1000 mg/d—consistentwith the largest secondary preventiontrial todate 9—appears reasonable. Thiscould be approximated by one 6-ozserving/wkof fish richest inn-3 PUFAs

(eg, farmed salmon, anchovies, her-ring), more frequent consumption of other fish (Table 2), or supplements.Optimal intake levels for other clini-cal outcomes are not well established.

Theeffects, if any, of low-levelmeth-

ylmercuryexposure inadultsare not es-tablished; mercury may modestly re-duce the cardiovascular benefits of fishintake. One can minimize concerns bychoosing fish higher in n-3 PUFAs andlowerin mercury or bysimplyconsum-ing a variety of different seafood. Indi-vidualswith high consumption( 5serv-ings/wk) should limit intakeof selectedspecies highest in mercury (Table 2).

DHAappears important forearly neu-rodevelopment.Womenwhoare ormaybecome pregnant and nursing mothersshould avoid selected species (shark,swordfish, goldenbass, andking mack-erel; locally caught fish per local advi-sories) andlimit intakeofalbacore tuna(6 oz/wk) to minimize methylmercuryexposure. 31,142 However, emphasis mustalso be placed on adequate consump-tion—12 oz/wk—ofother fish andshell-fish to provide reasonable amounts of DHA31,142 andavoid further decreases inalready low seafood intake amongwomen (74% of women of childbear-ing age and 85% of pregnant womenconsume 6 oz/wk). 206,207

Continued efforts to limit environ-mental contaminationfromorganochlo-rine compounds are appropriate. How-ever, levels of PCBs and dioxins in fisharelow, similar to those in several otherfoods, and the magnitudes of possiblerisks in adults are greatly exceeded bybenefits of fish intake and should havelittle impact on individual decisions re-garding fish consumption (for locallycaught freshwater fish, women ofchild-bearing ageshouldconsult regional ad-visories).

CONCLUSIONSPotential risks of fish intake must beconsidered in the context of potentialbenefits. Based on strength of evi-dence and potential magnitudes of effect, the benefits of modest fish con-sumption (1-2 servings/wk) outweighthe risks among adults and, excepting

a few selected fish species, amongwomen of childbearing age. Avoid-ance of modest fish consumption dueto confusion regarding risks and ben-efits could result in thousands of ex-cess CHD deaths annually and subop-

timal neurodevelopment in children.Author Contributions: Dr Mozaffarian had full ac-cess to all of the data in the study and takes respon-sibility for the integrity of the data and the accuracyof the data analysis. Study concept and design; acquisition of data; draft-ing of the manuscript; critical revision of the manuscript for important intellectualcontent: Mozaffarian,Rimm.Analysis and interpretation of data; statistical analysis; obtained funding: Mozaffarian.Administrative, technical, or material support; study supervision: Rimm.Financial Disclosures: Dr Rimm reports that he hasreceived research funding from Merck, Pfizer, andGlaxoSmithKline, and that he has received paymentor honoraria for presentations about food and dietsfromthe US Environmental Protection Agency,the USFood andDrug Administration, theInstituteof Medi-

cine, the Culinary Institute of America, the Interna-tional Chefs Association, and academic conferencesfunded by Bunge and Unilever. Dr Mozaffarian re-ported no financial disclosures.Funding/Support: This study was supported by Na-tional Heart, Lung, and Blood Institute, National In-stitutes of Health grant K08-HL-075628.Role of the Sponsor: The National Heart, Lung, andBlood Institute had no role in the design and conductof thestudy; thecollection, management, analysis, andinterpretation of the data; or the preparation, re-view, or approval of the manuscript.Acknowledgment: We thank IngeborgBrouwer, PhD,Wageningen Centre for Food Sciences, Wagenin-gen, the Netherlands, for providing mortality resultsfrom the Study on Omega-3 Fatty Acids and Ven-tricular Arrythmia trial; Calle Bengtsson, MD, PhD,CeciliaBjö rkelund, MD,PhD, LaurenLissner, PhD,andValter Sundh,BSc, TheSahlgrenska Academyat Göte-

borg University, Gothenburg, Sweden, for providingresults from their analysis of mercury and coronaryheart disease in Sweden;andWalter Willett, MD,DrPH,Meir Stampfer,MD, DrPH, andRuss Hauser, MD,ScD,Harvard School of Public Health, Boston, Mass, for in-sightful comments. None of those acknowledged re-ceived compensation for their contributions.

REFERENCES

1. Bang HO,DyerbergJ. Lipid metabolism andische-micheartdisease in GreenlandEskimos.In: DraperH,ed. Advances in Nutrition Research. New York, NY:Plenum Press; 1980:1-22.2. Kromhout D, Bosschieter EB,de LezenneCoulander C. The inverse relation between fishconsumptionand20-year mortality fromcoronaryheart disease. N Engl J Med . 1985;312:1205-1209.3. Burr ML, Fehily AM, Gilbert JF, et al. Effects ofchanges in fat, fish, and fibre intakes on death andmyocardial reinfarction: diet and reinfarction trial(DART). Lancet . 1989;2:757-761.4. Dolecek TA, Granditis G. Dietary polyunsaturatedfatty acids and mortality in the Multiple Risk Factor InterventionTrial (MRFIT). World RevNutrDiet . 1991;66:205-216.5. Siscovick DS, Raghunathan TE, King I, et al. Di-etary intake and cell membrane levels of long-chainn-3 polyunsaturated fatty acids and the risk of pri-mary cardiac arrest. JAMA. 1995;274:1363-1367.6. Kromhout D, Feskens EJ, Bowles CH. The protec-





12/16

tive effectof a small amount of fish on coronary heartdisease mortality in an elderly population. Int JEpidemiol . 1995;24:340-345.7. Daviglus ML,StamlerJ, Orencia AJ,et al.Fish con-sumption and the 30-year risk of fatal myocardialinfarction. N Engl J Med . 1997;336:1046-1053.8. Albert CM,Hennekens CH,O’DonnellCJ, et al.Fishconsumptionand riskof sudden cardiac death. JAMA.1998;279:23-28.

9. Gruppo Italiano per lo Studio della Sopravvivenzanell’Infarto miocardico. Dietary supplementationwith n-3 polyunsaturated fattyacids andvitamin E af-ter myocardial infarction: results of the GISSI-Prevenzione trial. Lancet . 1999;354:447-455.10. OomenCM, Feskens EJ,Rasanen L, et al.Fish con-sumption and coronary heart disease mortality in Fin-land,Italy, andThe Netherlands. Am J Epidemiol . 2000;151:999-1006.11. YuanJM, RossRK, Gao YT,Yu MC.Fishandshell-fish consumption in relation to death from myocar-dial infarction among men in Shanghai, China. Am JEpidemiol . 2001;154:809-816.12. Hu FB, Bronner L, Willett WC, et al. Fish andomega-3 fatty acid intake and risk of coronary heartdisease in women. JAMA. 2002;287:1815-1821.13. AlbertCM, CamposH, Stampfer MJ,et al.Bloodlevelsof long-chain n-3fatty acids and theriskof sud-den death. N Engl J Med . 2002;346:1113-1118.

14. Lemaitre RN, King IB, Mozaffarian D, Kuller LH,Tracy RP,SiscovickDS. n-3Polyunsaturated fatty ac-ids, fatal ischemic heart disease, and nonfatal myo-cardial infarction in older adults: the Cardiovascular Health Study. Am J Clin Nutr . 2003;77:319-325.15. MozaffarianD, LemaitreRN, Kuller LH, Burke GL,Tracy RP, Siscovick DS. Cardiac benefits of fish con-sumption may depend on the type of fish meal con-sumed: the Cardiovascular HealthStudy. Circulation .2003;107:1372-1377.16. Mozaffarian D, Ascherio A, Hu FB, et al. Inter-play between different polyunsaturated fattyacidsandriskof coronaryheartdiseasein men. Circulation . 2005;111:157-164.17. Yokoyama M, Origasu H, MatsuzakiM, et al.Ef-fects of eicosapentaenoic acid (EPA) on major cardio-vascular events in hypercholesterolemic patients: theJapan EPA Lipid Intervention Study (JELIS). Pre-sented at: American Heart Association Scientific Ses-

sions; November 17, 2005; Dallas, Tex.18. McLennan PL. Myocardial membrane fatty ac-ids and the antiarrhythmic actions of dietary fish oilin animal models. Lipids. 2001;36(suppl):S111-S114.19. Leaf A, Kang JX,XiaoYF, Billman GE.Clinicalpre-vention of sudden cardiac death by n-3 polyunsatu-rated fatty acids and mechanism of prevention of ar-rhythmiasby n-3fish oils. Circulation . 2003;107:2646-2652.20. Wang C, Harris WS, Chung M, et al. n-3 Fattyacids fromfish or fish-oilsupplements,but not{alpha}-linolenic acid, benefit cardiovascular disease out-comes in primary- and secondary-prevention stud-ies: a systematic review. Am J Clin Nutr . 2006;84:5-17.21. Simmer K. Longchain polyunsaturated fatty acidsupplementation in infantsbornat term. Cochrane Da-tabase Syst Rev . 2001;(4):CD000376.22. Food and Nutrition Board, Institute of Medicine.

Dietary Reference Intakes for Energy, Carbohydrate,Fiber,Fat, FattyAcids, Cholesterol, Protein, andAminoAcids (Macronutrients) . Washington, DC: The Na-tional Academies Press; 2002/2005.23. Uauy R, Hoffman DR, Mena P, Llanos A, BirchEE.Term infantstudies of DHA andARA supplemen-tation on neurodevelopment: results of randomizedcontrolled trials. J Pediatr . 2003;143:S17-S25.24. CohenJT, Bellinger DC, ConnorWE, Shaywitz BA.A quantitativeanalysisof prenatal intakeof n-3poly-unsaturated fatty acids and cognitive development.Am J Prev Med . 2005;29:366-374.

25. McCann JC, Ames BN. Is docosahexaenoic acid,an n-3 long-chain polyunsaturated fatty acid, re-quired for development of normal brain function? anoverview of evidence from cognitive and behavioraltests in humans andanimals. AmJ Clin Nutr . 2005;82:281-295.26. Lewin GA, Schachter HM, Yuen D, Merchant P,Mamaladze V, Tsertsvadze A; Agency for HealthcareResearchand Quality (AHRQ).Effectsof omega-3fatty

acids on child and maternal health. Evid Rep TechnolAssess (Summ) . August 2005;(118):1-11.27. US Environmental Protection Agency. MercuryStudy report to Congress. http://www.epa.gov/mercury/report.htm. Accessed January 24, 2006.28. USGeological Survey. Mercury in theenvironment.http://www.usgs.gov/themes/factsheet/146-00/. Ac-cessed October 25, 2005.29. Committee on the Toxicological Effects of Me-thylmercury, Board on Environmental Studies andToxicology; Commission on LifeSciences,National Re-searchCouncil. ToxicologicalEffectsof Methylmercury.Washington, DC: National Academies Press; 2000.30. Risk Assessment Information System. Toxicitysummary for mercury. http://risk.lsd.ornl.gov/tox/profiles/mercury_f_V1.shtml. Accessed January 24,2006.31. Centerfor Food Safetyand Applied Nutrition, USFood and Drug Administration. Seafood information

and resources. http://www.cfsan.fda.gov/seafood1.html. Accessed January 30, 2006.32. World HealthOrganization (WHO). Assessmentof the health risk of dioxins: re-evaluation of the Tol-erableDailyIntake(TDI).WHO Consultation; May25-29, 1998; Geneva, Switzerland.33. National Center for Environmental Assessment,US Environmental Protection Agency. Dioxin and re-lated compounds. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55264. Accessed March 14,2006.34. US Environmental Protection Agency. Polychlo-rinatedbiphenyls(PCBs). http://www.epa.gov/opptintr /pcb/. Accessed March 14, 2006.35. Verbeke W, Sioen I, Pieniak Z, Van Camp J, DeHenauw S. Consumer perception versus scientificevidence about health benefits and safety risks fromfish consumption. Public Health Nutr . 2005;8:422-429.

36. Center for FoodNutrition and Agriculture Policy,University of Maryland. Real mercury facts. http://www.realmercuryfacts.org/index.htm. AccessedMarch 23, 2006.37. DerSimonian R, Laird N. Meta-analysis in clinicaltrials. Control Clin Trials . 1986;7:177-188.38. Smith PL. Splines as a useful and convenient sta-tistical tool. Am Stat . 1979;33:57-62.39. Durrleman S, Simon R. Flexible regression mod-els with cubic splines. Stat Med . 1989;8:551-561.40. Morris MC, Evans DA, Tangney CC, Bienias JL,WilsonRS. Fishconsumption andcognitivedecline withage in a large community study. Arch Neurol . 2005;62:1849-1853.41. Peet M, StokesC. Omega-3 fatty acidsin thetreat-ment of psychiatric disorders. Drugs. 2005;65:1051-1059.42. Young G, Conquer J. Omega-3 fatty acids andneuropsychiatric disorders. ReprodNutr Dev . 2005;45:

1-28.43. Mori TA, Beilin LJ. Omega-3 fatty acids andinflammation. CurrAtheroscler Rep . 2004;6:461-467.44. Mickleborough TD, Lindley MR, Ionescu AA, FlyAD. Protective effect of fish oil supplementation onexercise-inducedbronchoconstriction in asthma. Chest .2006;129:39-49.45. Fraser GE, Sabate J, Beeson WL, Strahan TM. Apossible protective effect of nut consumption on riskof coronaryheart disease:the AdventistHealth Study.Arch Intern Med . 1992;152:1416-1424.46. Mann JI,Appleby PN,Key TJ,Thorogood M. Di-

etary determinants of ischaemic heart diseasein healthconscious individuals. Heart . 1997;78:450-455.47. Osler M, Andreasen AH, Hoidrup S. No inverseassociationbetweenfishconsumptionand risk of deathfrom all-causes, and incidence of coronary heart dis-ease in middle-aged, Danishadults. J Clin Epidemiol .2003;56:274-279.48. Folsom AR, Demissie Z. Fish intake, marineomega-3fattyacids, and mortalityin a cohort ofpost-

menopausalwomen. AmJ Epidemiol . 2004;160:1005-1010.49. Nakamura Y, Ueshima H, Okamura T, et al. As-sociation betweenfish consumptionand all-causeandcause-specific mortality in Japan: NIPPON DATA80,1980-99. Am J Med . 2005;118:239-245.50. IsoH, KobayashiM, Ishihara J, et al.Intake of fishand n3 fatty acids and risk of coronary heart diseaseamong Japanese: the Japan Public Health Center-Based ( JPHC) Study CohortI. Circulation . 2006;113:195-202.51. Burr ML,Ashfield-Watt PA,DunstanFD, etal. Lackof benefit of dietary advice to men with angina: re-sults of a controlledtrial. Eur J ClinNutr . 2003;57:193-200.52. Siscovick DS, Lemaitre RN, Mozaffarian D. Thefish story: a diet-heart hypothesis with clinical impli-cations: n-3 polyunsaturated fatty acids, myocardialvulnerability, andsudden death. Circulation . 2003;107:

2632-2634.53. He K, Song Y, Daviglus ML, et al. Fish consump-tion and incidence of stroke: a meta-analysis of co-hort studies. Stroke . 2004;35:1538-1542.54. Mozaffarian D, Longstreth WT Jr, Lemaitre RN,et al. Fish consumption and stroke risk in elderly in-dividuals: the cardiovascular health study. Arch In-tern Med . 2005;165:200-206.55. Erkkila AT, Lichtenstein AH, MozaffarianD, Her-rington DM. Fish intake is associated with a reducedprogression of coronaryartery atherosclerosis in post-menopausal women with coronaryartery disease. Am J Clin Nutr . 2004;80:626-632.56. Sacks FM, Stone PH, Gibson CM, Silverman DI,RosnerB, PasternakRC; HARP ResearchGroup. Con-trolled trial of fish oil for regression of human coro-naryatherosclerosis. J Am CollCardiol . 1995;25:1492-1498.57. von Schacky C, Angerer P, KothnyW, Theisen K,

Mudra H. The effect of dietary omega-3 fatty acidson coronary atherosclerosis: a randomized, double-blind, placebo-controlled trial. AnnIntern Med . 1999;130:554-562.58. Angerer P, Kothny W, Stork S, von Schacky C.Effect of dietary supplementation with omega-3 fattyacids on progression of atherosclerosis in carotidarteries. Cardiovasc Res . 2002;54:183-190.59. Balk EM, Lichtenstein AH, Chung M, KupelnickB, Chew P, Lau J. Effects of omega-3 fatty acids oncoronary restenosis, intima-media thickness, and ex-ercise tolerance: a systematicreview. Atherosclerosis .2006;184:237-246.60. Raitt MH, Connor WE, Morris C, et al. Fish oilsupplementationand riskof ventriculartachycardia andventricular fibrillation in patients with implantabledefibrillators: a randomized controlled trial. JAMA.2005;293:2884-2891.61. Leaf A, Albert CM, Josephson M, et al. Preven-

tion of fatal arrhythmias in high-risk subjects by fishoiln-3 fatty acid intake. Circulation . 2005;112:2762-2768.62. Brouwer IA, Zock PL, Camm AJ, et al. Effect offish oil on ventricular tachyarrhythmia and death inpatients with implantable cardioverter defibrillators:theStudy on Omega-3Fatty Acids andVentricularAr-rhythmia (SOFA)randomized trial. JAMA. 2006;295:2613-2619.63. MozaffarianD, Psaty BM,Rimm EB,et al.Fish in-take andrisk of incident atrialfibrillation. Circulation .2004;110:368-373.





13/16

64. Frost L, Vestergaard P. n-3 Fatty acids con-sumed from fish and risk of atrial fibrillation or flut-ter: the Danish Diet, Cancer, and Health Study. Am JClin Nutr . 2005;81:50-54.65. Calo L, Bianconi L, Colivicchi F, et al. N-3 Fattyacidsfor the prevention of atrial fibrillation aftercoro-nary artery bypass surgery: a randomized, controlledtrial. J Am Coll Cardiol . 2005;45:1723-1728.66. Mozaffarian D, Bryson CL, Lemaitre RN, Burke

GL,Siscovick DS.Fish intakeand risk of incident heartfailure. J Am Coll Cardiol . 2005;45:2015-2021.67. Charnock JS, McLennan PL, Abeywardena MY.Dietary modulation of lipid metabolismand mechani-calperformance of theheart. Mol Cell Biochem . 1992;116:19-25.68. Kenny D, Warltier DC, Pleuss JA, Hoffmann RG,Goodfriend TL,Egan BM.Effect of omega-3 fatty ac-idson thevascularresponseto angiotensin in normo-tensive men. Am J Cardiol . 1992;70:1347-1352.69. Chin JP,Gust AP,Nestel PJ,Dart AM.Marine oilsdose-dependently inhibit vasoconstriction of fore-armresistancevessels in humans. Hypertension . 1993;21:22-28.70. GeleijnseJM, Giltay EJ, Grobbee DE, Donders AR,Kok FJ. Blood pressureresponse to fishoil supplemen-tation: metaregression analysis of randomized trials. J Hypertens . 2002;20:1493-1499.71. Mozaffarian D, Geelen A, Brouwer IA, Geleijnse

JM, ZockPL,Katan MB. Effectof fish oil onheartratein humans: a meta-analysis of randomized controlledtrials. Circulation . 2005;112:1945-1952.72. Mozaffarian D, Gottdiener JS, Siscovick DS. In-take of tuna or other broiled or baked fish vs. friedfishand cardiacstructure,function, andhemodynamics.Am J Cardiol . 2006;97:216-222.73. Nestel PJ.Fish oil andcardiovascular disease: lip-idsandarterial function. AmJ ClinNutr . 2000;71:228S-231S.74. ChristensenJH. n-3fattyacids and theriskof sud-den cardiac death: emphasis on heart rate variability.Dan Med Bull . 2003;50:347-367.75. KristensenSD, Iversen AM, Schmidt EB. n-3 poly-unsaturatedfattyacidsandcoronary thrombosis. Lipids.2001;36(suppl):S79-S82.76. Clandinin MT,CheemaS, Field CJ, GargML, Ven-katraman J, ClandininTR. Dietary fat: exogenousde-termination of membrane structure and cell function.

FASEB J. 1991;5:2761-2769.77. Feller SE, Gawrisch K. Properties of docosa-hexaenoic-acid-containing lipids and their influenceon the function of rhodopsin. Curr Opin Struct Biol .2005;15:416-422.78. Vanden Heuvel JP. Diet, fatty acids, and regula-tion of genes important for heart disease. Curr Ath-eroscler Rep . 2004;6:432-440.79. Harris WS. n-3Fatty acids andserum lipoproteins:human studies. AM J Clin Nutr . 1997;65(5 suppl):1645S-1654S.80. Dallongeville J, Yarnell J, Ducimetiere P, etal. Fishconsumption is associated with lower heart rates.Circulation . 2003;108:820-825.81. Wilson PW,D’AgostinoRB, Levy D, Belanger AM,SilberschatzH, Kannel WB. Predictionof coronaryheartdisease usingrisk factor categories. Circulation . 1998;97:1837-1847.82. Austin MA, Hokanson JE, Edwards KL. Hyper-

triglyceridemia as a cardiovascular risk factor. Am JCardiol . 1998;81:7B-12B.83. KannelWB, KannelC, PaffenbargerRS Jr,CupplesLA. Heart rate and cardiovascular mortality: theFramingham Study. Am Heart J . 1987;113:1489-1494.84. Jouven X, Zureik M, Desnos M, Guerot C, Duci-metiere P. Resting heart rate as a predictive risk fac-tor for sudden death in middle-aged men. Cardio-vasc Res. 2001;50:373-378.85. Hu FB,Stampfer MJ,MansonJE, etal. Dietary satu-rated fats andtheirfood sources in relation to therisk

of coronary heart disease in women. Am J Clin Nutr .1999;70:1001-1008.86. Anderson RN, Smith LB; Division of Vital Statis-tics, Centers for Disease Control and Prevention. Na-tional Vital Statistics Reports: deaths: leading causesfor 2002. http://www.cdc.gov/nchs/data/nvsr/nvsr53/nvsr53_17.pdf. Accessed March 29, 2006.87. Brox J, Olaussen K, Osterud B, et al. A long-termseal- and cod-liver-oil supplementation in hypercho-

lesterolemic subjects. Lipids. 2001;36:7-13.88. Eritsland J, Arnesen H, Gronseth K, Fjeld NB,Ab-delnoorM. Effect of dietarysupplementationwith n-3fatty acids on coronary artery bypass graft patency.Am J Cardiol . 1996;77:31-36.89. Johansen O, Brekke M, Seljeflot I, Abdelnoor M,ArnesenH; CoronaryAngioplasty RestenosisTrial. N-3fatty acids do not prevent restenosis after coronaryangioplasty: results from the CART study. J Am CollCardiol . 1999;33:1619-1626.90. Kaul U, Sanghvi S, Bahl VK,Dev V, Wasir HS.Fishoilsupplements forpreventionof restenosisafter coro-nary angioplasty. Int J Cardiol . 1992;35:87-93.91. Leaf A, Jorgensen MB, Jacobs AK, et al. Do fishoils prevent restenosis after coronary angioplasty?Circulation . 1994;90:2248-2257.92. Nilsen DW, Albrektsen G, Landmark K, Moen S,Aarsland T, Woie L. Effects of a high-dose concen-trate of n-3fattyacids or corn oilintroduced early af-

ter an acute myocardial infarction on serum triacyl-glycerol and HDL cholesterol. Am J Clin Nutr . 2001;74:50-56.93. Singh RB, Niaz MA, Sharma JP, Kumar R, Ras-togiV, Moshiri M. Randomized, double-blind, placebo-controlled trial of fish oil and mustard oil in patientswith suspected acute myocardial infarction: the In-dian experiment of infarct survival—4. Cardiovasc Drugs Ther . 1997;11:485-491.94. HooperL, ThompsonRL, HarrisonRA, etal. Risksandbenefits of omega 3 fats formortality, cardiovas-cular disease, and cancer: systematic review. BMJ .2006;332:752-760.95. CheungBM, LauderIJ, LauCP, KumanaCR. Meta-analysisof largerandomizedcontrolled trialsto evalu-atethe impactof statins on cardiovascular outcomes.Br J Clin Pharmacol . 2004;57:640-651.96. Uauy R, MenaP, WegherB, NietoS, SalemN Jr.Longchainpolyunsaturated fattyacid formationin neo-

nates: effect of gestational ageand intrauterinegrowth.Pediatr Res . 2000;47:127-135.97. Helland IB, Smith L, Saarem K, Saugstad OD,Drevon CA. Maternal supplementation with very-long-chain n-3 fatty acids during pregnancy and lac-tation augments children’s IQ at 4 years of age.Pediatrics . 2003;111:e39-e44.98. Oken E, Wright RO, Kleinman KP, et al. Mater-nal fish consumption, hair mercury, and infant cog-nition in a U.S.Cohort. EnvironHealth Perspect . 2005;113:1376-1380.99. Colombo J, Kannass KN, Shaddy DJ, et al. Ma-ternal DHA and the development of attention in in-fancy and toddlerhood. Child Dev . 2004;75:1254-1267.100. DanielsJL, Longnecker MP,RowlandAS, Gold-ing J. Fish intake during pregnancy and early cogni-tivedevelopment of offspring. Epidemiology . 2004;15:394-402.

101. US Environmental Protection Agency. Control-ling power plant emissions: emissions progress. http://www.epa.gov/mercury/control_emissions/emissions.htm. Accessed March 29, 2006, 2006.102. Officeof Regulatory Affairs, US Food AndDrugAdministration. Compliancepolicy guides. http://www.fda.gov/ora/compliance_ref/cpg/cpgfod/default.htm#sc540. Accessed February 2, 2006, 2006.103. Bayarri S, Baldassarri LT, Iacovella N, Ferrara F,di Domenico A. PCDDs,PCDFs, PCBs andDDE in ed-iblemarinespecies fromthe AdriaticSea. Chemosphere .2001;43:601-610.

104. Schmitt CJ, Hinck JE, Blazer VS, et al. Environ-mental contaminants and biomarker responsesin fishfrom the Rio Grande and its U.S. tributaries: spatialand temporal trends. Sci TotalEnviron . 2005;350:161-193.105. Karl H, Ruoff U, Bluthgen A. Levels of dioxinsin fish and fishery products on the German market.Chemosphere . 2002;49:765-773.106. Schecter A, Dellarco M, PapkeO, OlsonJ. A com-

parison of dioxins, dibenzofurans and coplanar PCBsin uncooked and broiled ground beef, catfish andbacon. Chemosphere . 1998;37:1723-1730.107. Jensen E, Bolger PM.Exposure assessment of di-oxins/furans consumed in dairy foods and fish. Food Addit Contam . 2001;18:395-403.108. Fiedler H, CooperK, BergekS, et al.PCDD,PCDF,andPCB in farm-raised catfish from southeastUnitedStates—concentrations,sources,and CYP1Ainduction.Chemosphere . 1998;37:1645-1656.109. FocantJF, PirardC, DePauwE. Levels ofPCDDs,PCDFs andPCBs in Belgian andinternational fastfoodsamples. Chemosphere . 2004;54:137-142.110. US Food And Drug Administration. Food andDrugAdministration totaldiet study. http://vm.cfsan.fda.gov/~comm/tds-toc.html. Accessed February 1,2006.111. Hayward DG, Holcomb J, Glidden R, Wilson P,Harris M, Spencer V. Quadrupole ion storage

tandem mass spectrometry and high-resolution massspectrometry: complementaryapplication in the mea-surement of 2,3,7,8-chlorine substituted dibenzo-p-dioxins anddibenzofurans in US foods. Chemosphere .2001;43:407-415.112. Gomara B, Bordajandi LR,Fernandez MA,et al.Levels andtrends of polychlorinated dibenzo-p-dioxins/furans (PCDD/Fs) and dioxin-like polychlorinated bi-phenyls (PCBs) in Spanish commercial fish and shell-fish products, 1995-2003. J Agric Food Chem . 2005;53:8406-8413.113. Rawn DF,Forsyth DS,RyanJJ, et al.PCB,PCDDandPCDFresidues in finand non-fin fishproducts fromthe Canadian retail market 2002. Sci Total Environ .2006;359:101-110.114. Food Safety Authority of Ireland. Summary ofinvestigation of dioxins, furans, and PCBs in farmedsalmon, wildsalmon, farmedtroutand fishoil capsules.March 2002. http://www.fsai.ie/surveillance/food

/surveillance_food_summarydioxins.asp. AccessedMarch 31, 2006.115. HitesRA, ForanJA, CarpenterDO, Hamilton MC,Knuth BA, Schwager SJ. Global assessment of or-ganiccontaminants in farmed salmon. Science. 2004;303:226-229.116. Jacobs MN, Covaci A, Schepens P. Investiga-tionof selectedpersistentorganicpollutantsin farmedAtlantic salmon(Salmo salar), salmonaquaculture feed,and fish oil components of the feed. Environ SciTechnol . 2002;36:2797-2805.117. Easton MD, Luszniak D, Von der GE. Prelimi-nary examination of contaminant loadings in farmedsalmon, wild salmon and commercial salmon feed.Chemosphere . 2002;46:1053-1074.118. BordajandiLR, Martin I, AbadE, Rivera J, Gonza-lez MJ. Organochlorine compounds(PCBs,PCDDs andPCDFs) in seafish and seafood from the Spanish At-lanticSouthwest Coast. Chemosphere . 2006;64:1450-

1457.119. Kiviranta H, Hallikainen A, Ovaskainen ML,Kum-pulainen J, Vartiainen T. Dietary intakes of polychlo-rinated dibenzo-p-dioxins, dibenzofurans and poly-chlorinated biphenylsin Finland. FoodAddit Contam .2001;18:945-953.120. Schecter A, Cramer P, Boggess K, et al. Intakeof dioxins and related compounds from food in theU.S. population. J Toxicol EnvironHealth A . 2001;63:1-18.121. Agricultural ResearchService, US Department ofAgriculture. USDA National Nutrient Database for





14/16


15/16

exposure to background levels of polychlorinated bi-phenyls and cognitive functioning among school-agechildren. Am J Epidemiol . 2005;162:17-26.180. Judd N, Griffith WC, Faustman EM. Contribu-tionof PCBexposurefrom fishconsumptionto total di-oxin-like dietary exposure. Regul Toxicol Pharmacol .2004;40:125-135.181. Peapod by Stop & Shop. http://www.peapod.com/ Accessed July 11, 2006.

182. Great Alaska Seafood. Fresh wildAlaska salmon.http://www.great-alaska-seafood.com/fresh-alaska-salmon.htm#alaska-king-salmon. Accessed July 11,2006.183. Ed’s Kasilof Seafoods. Alaska wild salmon. http://www.kasilofseafoods.com/seafood-gifts/wild-salmon.htm. Accessed July 12, 2006.184. WildPacificSalmon. Wildsalmonproducts.http://www.wildpacificsalmon.com/site/680079/page/45031. Accessed July 12, 2006.185. Chee KM,Gong JX,ReesDM,et al.Fatty acid con-tent of marine oil capsules. Lipids. 1990;25:523-528.186. Centerfor DrugEvaluationand Research, USFoodAnd Drug Administration. Omacor:consumer drug in-formation sheet—approval label. http://www.fda.gov/cder/foi/label/2004/21654lbl.pdf. Accessed April 5,2006.187. ForanSE, Flood JG, Lewandrowski KB. Measure-ment of mercury levels in concentrated over-the-counterfish oilpreparations: is fishoil healthier thanfish?Arch Pathol Lab Med . 2003;127:1603-1605.188. Storelli MM, Storelli A, Marcotrigiano GO. Poly-chlorinated biphenyls, hexachlorobenzene, hexachlo-rocyclohexaneisomers,and pesticide organochlorine resi-dues in cod-liver oil dietary supplements. J Food Prot .2004;67:1787-1791.

189. Jimenez B, Wright C, Kelly M, Startin JR. LevelsofPCDDs, PCDFs andnon-orthoPCBs in dietarysupple-ment fish oilobtained inSpain. Chemosphere . 1996;32:461-467.190. Patch CS, Tapsell LC, Mori TA, et al. The use ofnovel foods enriched withlong-chain n-3fatty acids toincrease dietary intake: a comparison of methodolo-giesassessingnutrient intake. J AmDietAssoc . 2005;105:1918-1926.

191. Warner K. Impact of high-temperaturefood pro-cessing on fats and oils. Adv Exp Med Biol. 1999;459:67-77.192. Simopoulos AP. Essential fatty acids in healthandchronicdisease. AmJClinNutr . 1999;70:560S-569S.193. Kris-EthertonPM, TaylorDS, Yu-PothS, etal.Poly-unsaturated fatty acids in the food chain in the UnitedStates. A m J C li n N u tr . 2000 ;71 :179S-188S.194. Hu FB,StampferMJ, MansonJE, et al.Dietaryin-take ofalpha-linolenicacidand riskof fatal ischemicheartdiseaseamong women. AmJ Clin Nutr . 1999;69:890-897.195. NaylorRL, GoldburgRJ, PrimaveraJH, et al.Effectofaquacultureonworldfish supplies. Nature . 2000;405:1017-1024.196. Pauly D, WatsonR, Alder J. Globaltrends inworldfisheries: impacts on marine ecosystems and foodsecurity. Philos TransR Soc LondB BiolSci . 2005;360:5-12.197. Devine JA,Baker KD,HaedrichRL. Fisheries: deep-sea fishesqualify as endangered. Nature . 2006;439:29.198. National MarineFisheriesService. Fisheriesof theUnitedStates, 2004. SilverSpring, Md: USDeptof Com-merce; 2005.199. Garcia SM, Grainger RJ. Gloom and doom? the

future of marine capture fisheries. Philos Trans R Soc Lond B Biol Sci. 2005;360:21-46.200. World Resources Institute. Millennium Ecosys-temAssessment:Ecosystemsand Human Well-Being— Synthesis Report . Washington, DC: Island Press;2005.201. Williams CM, Burdge G. Long-chain n-3 PUFA:plant v. marine sources. Proc Nutr Soc . 2006;65:42-50.

202. Mozaffarian D. Does alpha-linolenic acid intakereduce the risk of coronary heart disease? a review oftheevidence. AlternTherHealth Med . 2005;11:24-30,31, 79.203. BrownAJ, PangE, Roberts DC. Persistentchangesin the fatty acid composition of erythrocyte mem-branes aftermoderateintake ofn-3polyunsaturatedfattyacids: studydesign implications. AmJ Clin Nutr . 1991;54:668-673.204. Kris-Etherton PM, Harris WS, Appel LJ. Fishconsumption, fish oil, omega-3 fatty acids, andcardiovascular disease. Circulation . 2002;106:2747-2757.205. Van de Werf F, Ardissino D, Betriu A, et al; TaskForce on theManagement of Acute Myocardial Infarc-tion of the European Society of Cardiology. Manage-mentof acutemyocardial infarction in patients present-ingwith ST-segmentelevation. EurHeartJ . 2003;24:28-66.206. Schober SE, Sinks TH, Jones RL,et al.Blood mer-cury levels in US children and women of childbearingage, 1999-2000. JAMA. 2003;289:1667-1674.207. Oken E, Kleinman KP, Berland WE, Simon SR,Rich-Edwards JW, Gillman MW. Decline in fishconsumption among pregnant women after a nationalmercury advisory. Obstet Gynecol . 2003;102:346-351.

Author in the Room Teleconference Join Dr Mozaffarian, an author of this article, on Wednesday, No-vember 15, 2006, from 2 to 3 PM EDT for “Author in the Room,”an interactive teleconference aimed at closing the gap betweenknowledge—what is published in this article—and action—howmuch of this knowledge canbe put into your actual practice. Thisteleconference, facilitated by clinical experts, should help read-ers answer their questions and consider the implications of thearticle for their practice.Author in the Room is brought to you by JAMA and the Institutefor Healthcare Improvement. To register for “Author in theRoom,”please visit http://www.ihi.org/authorintheroom. You can listento past conferences or to subscribe to the podcast at http://jama

.ama-assn.org/authorintheroom/authorindex.dtl.





16/16

prevented the development of protective immunity. In an-other murine model, protective immunity was also inhib-ited by azithromycin. 9 Brunham et al 10 observed that whilechlamydial sexually transmitted infections in Vancouver de-creased substantially over a few years after an azithromy-cin treatment program began, they estimated that annual

risk of re-infection increased by 4.6% thereafter.Personal hygiene and environmental improvements havealready eliminated blinding trachoma in developedand somedeveloping countries. Emphasis should be placedon allSAFEcomponents with further evaluation of the antibiotic com-ponent, longitudinalassessments of efficacy, andvaccine de-velopment for sustainability.Deborah Dean, MD, [email protected] Atik, MD, MPHCenter for Immunobiology and Vaccine DevelopmentChildren’s Hospital Oakland Research InstituteOakland, Calif Ton Thi Kim Thanh, MD

Vu Quoc Luong, MD, PhDNational Institute of OphthalmologyMinistry of HealthStephan Lagree, PhDGroup Research in Technology ExchangeHanoi, VietnamFinancial Disclosures: None reported.

1. West SK, Munoz B, Mkocha H, et al. Infection with Chlamydia trachomatisafter mass treatment of a trachoma hyperendemic community in Tanzania: a lon-gitudinal study. Lancet . 2005;366:1296-1300.2. SchachterJ, West S, Mabey D, et al.Azithromycin in control oftrachoma: effectof communitywide treatmenton Chlamydia trachomatis infection. Lancet . 1999;354:630-635.3. Chidambaram JD, Alemayehu W, Melese M, et al. Effect of a single mass an-tibiotic distributionon theprevalenceof infectioustrachoma. JAMA. 2006;295:1142-1146.

4. Solomon AW, Holland MJ, Alexander ND, et al. Mass treatment with single-dose azithromycin for trachoma. N Engl J Med . 2004;351:1962-1971.5. World Health Organization. Report of the Second Meeting of the WHO Alli-ance for the Global Elimination of Trachoma. Geneva, Switzerland:World HealthOrganization;1998. Publication WHO/PBL/GET/98.2.6. Somani J, Bhullar VB, Workowski KA, Farshy CE, Black CM. Multiple drug–resistant Chlamydia trachomatis associatedwith clinical treatment failure. J Infect Dis. 2000;181:1421-1427.7. Bailey R, Duong T, Carpenter R, Whittle H, Mabey D. The duration of human

ocular chlamydial infection is agedependent. EpidemiolInfect . 1999;123:479-486.8. Su H, Morrison R, Messer R, Whitmire W, Hughes S, Caldwell HD. The effectof doxycycline treatment on the development of protective immunity in a murinemodel of chlamydial genital infection. J Infect Dis. 1999;180:1252-1258.9. FernandezAD, ElmoreMK, Metzger DW. Azithromycinmodulates murineim-mune responses to pneumococcal conjugate vaccine and inhibits nasal clearanceof bacteria. J Infect Dis. 2004;190:1762-1766.10. Brunham RC, Pourbohloul B, Mak S, White R, Rekart ML. The unexpectedimpact of a Chlamydia trachomatis infection control program on susceptibility toreinfection. J Infect Dis. 2005;192:1836-1844.

CORRECTIONS

Citation Error: In the Original Contribution entitled “Impact of Annual TargetedTreatment on Infectious Trachoma and Susceptibility to Reinfection” published inthe September 27, 2006, issue of JAMA (2006;296:1488-1497) page 1493 con-

tained an error in the use of a citation. The sentence “Since the immune responseto C trachomatis is usually sustained for only 1 to 4 months, 24 we reasoned thatindividuals with resolved infection (conversion of PCR-positive result to negativeat a subsequent timepoint) would be susceptibleto infection at thenext timepoint,6 months later” should read “Since the duration of C trachomatis infection is re-duced in older agegroups, presumably as a resultof acquired immunity, 24 we rea-soned that if the immune response is usually sustained for only 1 to 4 months,individuals with resolved infection (conversion of PCR-positive result to negativeat a subsequent timepoint) would be susceptible to infection at thenext timepoint,6 months later.”

Omitted Financial Disclosure Information: In the Clinical Review entitled “FishIntake, Contaminants, and Human Health” published in the October 18, 2006issue of JAMA (2006;296:1885-1899),financial disclosure information was omit-ted. Thedisclosure forDr Mozaffarian shouldhave read: Dr Mozaffarian reportedthat he has received honoraria for presentations or articles about fish or trans fatconsumption & cardiovascularhealth from the Institute of FoodTechnologists, theDanish Nutrition Council, the American Oil Chemists’ Society, Project Syndicate,and several academic medical centers.

LETTERS

590 JAMA, February 14, 2007—Vol 297, No. 6 (Reprinted) ©2007 American Medical Association. All rights reserved.

fish intake contaminants and human health jama

Documents