Intake of Tuna or Other Broiled or Baked Fish Versus Fried Fish and Cardiac Structure, Function, and Hemodynamics

Transcription

1 Intake of Tuna or Other Broiled or Baked Fish Versus Fried Fish and Cardiac Structure, Function, and Hemodynamics Dariush Mozaffarian, MD, MPH a, *, John S. Gottdiener, MD b, and David S. Siscovick, MD, MPH c Fish intake is associated with improved cardiovascular health, including a lower risk of arrhythmic death, atrial fibrillation, and heart failure. However, the physiologic effects that may produce these cardiovascular benefits are not well-established. We investigated the cross-sectional associations between a usual dietary intake of fish during the previous year and cardiac structure, function, and hemodynamics as determined by physical examination and 2-dimensional, Doppler, and M-mode transthoracic echocardiography among 5,073 older adults enrolled in the Cardiovascular Health Study. On multivariate-adjusted analyses, consumption of tuna or other broiled or baked fish was associated with a lower heart rate (p <0.001), lower systemic vascular resistance (p 0.002), and greater stroke volume (p <0.001). Tuna/other fish intake was also associated with a higher E/A ratio (p 0.004), a measure of more normal diastolic function. In contrast, fried fish or fish sandwich (fish burger) intake was associated with left ventricular wall motion abnormalities (p 0.02), a reduced ejection fraction (p <0.001), lower cardiac output (p 0.04), a trend toward a larger left ventricular diastolic dimension (p 0.07), and higher systemic vascular resistance (p 0.003). In conclusion, in this large population-based study, the intake of tuna or other broiled or baked fish was associated with improved cardiac hemodynamics, but fried fish intake was associated with structural abnormalities indicative of systolic dysfunction and potential coronary atherosclerosis. These findings suggest potential specific physiologic mechanisms that may, in part, account for the effects of fish intake on cardiovascular health Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97: ) Fatty fish intake is associated with a lower risk of several cardiovascular outcomes, including fatal coronary heart disease and arrhythmic death, 1 3 atrial fibrillation, 4 stroke, 5 and congestive heart failure. 6 However, the physiologic effects accounting for these potential benefits are not well-established. Experimental studies have suggested that fish oil may have direct antiarrhythmic effects, 7,8 but this has not been clearly established in humans. Furthermore, such effects would not reduce the incidence of ischemic stroke or heart failure. In experimental trials, fish oil supplementation improved cardiac hemodynamics, influencing heart rate Methods Design and population: CHS was a prospective, population-based, multicenter cohort study of determinants of cardiovascular risk among older adults. 14 In 1989 to 1990 and 1992, 5,888 men and women aged 65 years were randomly selected and enrolled from Medicare eligibility lists in 4 United States communities. The baseline evaluation included health status, medical history, physical exama Channing Laboratory, Department of Medicine, Brigham and Women s Hospital and Harvard Medical School, and Departments of Epidemiology and Nutrition, Harvard School of Public Health, Boston, Massachusetts; b Division of Cardiology, University of Maryland School of Medicine, Baltimore, Maryland; and c Cardiovascular Health Research Unit, Departments of Medicine and Epidemiology, University of Washington School of Medicine, Seattle, Washington. Manuscript received June 10, 2005; revised manuscript received and accepted August 5, This study was supported by contracts N01-HC through N01- HC-85086, N01-HC-35129, and N01-HC and a Mentored Clinical Scientist Award (K08-HL ) to Dr. Mozaffarian from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. * Corresponding author: Tel: ; fax: address: dmozaffa@hsph.harvard.edu (D. Mozaffarian). (HR), blood pressure (BP), left ventricular (LV) diastolic function, and vasoconstrictive responses However, whether these effects obtained using high doses of fish oil in short-term studies are applicable to usual dietary fish intake is not well understood. Also, the relations between fish intake and other important cardiovascular indexes in humans such as LV dimension, left atrial dimension, stroke volume, cardiac output, and systemic vascular resistance are unknown. The presence or absence of such relations would elucidate potential physiologic effects of dietary fish intake and support a potential causal role of fish intake on cardiovascular outcomes. We investigated the associations between usual dietary fish intake and cardiac structure, function, and hemodynamics among 5,073 men and women enrolled in the Cardiovascular Health Study (CHS) /06/$ see front matter 2006 Elsevier Inc. All rights reserved. doi: /j.amjcard

2 Preventive Cardiology/Fish Intake, Cardiac Structure, and Hemodynamics 217 Table 1 Relations between fish intake and other baseline characteristics among 5,073 older adults Tuna/Other Fish Fried Fish (n 2,439) (n 1,634) 1/wk (n 1,000)* Age (yrs) Men (%) Education (% high school) Body mass index (kg/m 2 ) Diabetes mellitus (%) Coronary heart disease (%) Clinical valvular disease (%) Current smoking (%) Smoking history (pack-yrs) Leisure-time activity (kcal/wks) 1,594 2,108 1,785 2,128 1,875 2,100 1, ,759 1,978 1,862 2,134 1,893 2,266 Aspirin use (%) Blocker use (%) Dietary EPA DHA (mg/d) Values are presented as means SD (continuous variables) or percentages (categorical variables); relations of fish intake with other demographic, clinical, laboratory, and dietary variables in the Cardiovascular Health Study have been previously reported. 3 6 * The few participants consuming fried fish 3/weeks (n 74) were combined with 1/week. p for trend Estimated dietary intake from these fish meals based on United States commercial landings data and estimated content of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in each serving (3 to 5 oz). 6 ination, electrocardiography, echocardiography, carotid ultrasonography, pulmonary function testing, and laboratory testing We performed a cross-sectional analysis among participants at baseline using information on usual dietary intake in the previous year, as assessed by a food frequency questionnaire, and cardiac structure, function, and hemodynamics, as determined by physical examination and transthoracic echocardiography. We excluded 687 patients enrolled in 1992 (a food frequency questionnaire was not administered in 1992), 27 patients in whom echocardiography was not completed, and 101 patients with incomplete data on fish consumption, resulting in 5,073 participants in this analysis. Each center s institutional review committee approved the study, and all subjects gave informed consent. Dietary assessment: Each participant s usual dietary intake was assessed using a picture-sort version of the National Cancer Institute food frequency questionnaire. 17 Participants were asked to indicate how often, on average, they had consumed various specified foods during the previous year, including intake of tuna fish, other fish (broiled or baked), and fried fish or fish sandwiches (fried fish burgers). For each food, participants chose among 5 possible response categories ranging from 4/year to 5 /week. For most nutrients, the midpoint of each category was used to sum intakes of different items. Using information from a detailed assessment of fish consumption performed in a separate cohort, 1 the median fish intakes for the 2 highest response categories were calculated to be slightly less than the category midpoint (1.5/week for the category 1 to 4/week and 5/week for the category 5 /week). This had little effect on the relative ranking of participants or on the estimates of relations with outcomes among participants, but likely better estimated the absolute fish intake for each participant. As a biomarker of fish oil content, we have shown that plasma phospholipid levels of eicosapentaenoic acid and docosahexaenoic acid correlate with the intake of tuna fish (Spearman correlation [r] 0.35, p 0.01), other fish (r 0.59, p 0.001), and combined tuna/other fish (r 0.55, p 0.001), but not fried fish (r 0.04, p 0.78). 3 Macronutrient measures, such as saturated fat intake, were estimated from the questionnaire responses and adjusted for total energy using regression analyses. 18 Assessment of cardiac structure, function, and hemodynamics: Each participant s seated HR at rest was measured for 30 seconds, and the seated BP at rest was measured in the right arm using an appropriately sized cuff and a Hawksley random-zero sphygmomanometer (model 7076, Hawksley and Sons Limited, Sussex, United Kingdom). The mean BP was calculated as the diastolic BP plus 1/3 the difference between the systolic and diastolic BP. The methods for 2-dimensional, Doppler, and M-mode transthoracic echocardiography in the CHS have been previously described. 16 Studies were evaluated at a centralized echocardiography center by readers blinded to the participants clinical information. Quality control measures included standardized training of echocardiography technicians and readers, observation of technicians by a trained echocardiographer, periodic blind duplicate readings, phantom studies, and quality control adults. 16 LV and left atrial chamber dimensions were determined by 2-dimensionally directed M-mode images according to American Society of Echocardiography standards. 19 Global LV systolic function was qualitatively assessed on 2-dimensional echocardiography as normal (ejection fraction 0.55), borderline (ejection

3 218 The American Journal of Cardiology ( Table 2 Relations between fish intake and measures of cardiac structure and systolic function* Measure (n) Tuna/Other Fish Intake (n) Fried Fish Intake (n) (n 2,439) (n 1,634) 1/wk (n 1,000) Septal thickness, cm (3,344) Posterior wall thickness, cm (3,344) LV diastolic dimension, mm (3,344) LA dimension, mm (4,915) Echo LV mass, g (3,344) ECG LV mass, g (4,832) Fractional shortening, % (3,298) Reduced EF ( 55), % (5,052) LV WMA, % (5,038) Values are presented as means (continuous variables) or percentages (categorical variables), adjusted for age (yrs), gender (male/female), race (white/ non-white), enrollment site (4 sites), education (less than high school, high school, more than high school), smoking (never, former, current), leisure time physical activity (kcal/d), body mass index (kg/m 2 ), diabetes mellitus (yes/no), coronary heart disease (yes/no), and intake of tuna/other fish (4 categories), fried fish (3 categories), alcohol (drinks per week), and total calories (kcal/day). * For each outcome measure, those for whom the outcome measure was not available were not included in the analysis. p 0.05 and p compared with intake nth. ECG electrocardiographic; EF ejection fraction; LA left atrial; WMA wall motion abnormalities. fraction 0.45 to 0.54), or impaired (ejection fraction 0.45), with inter-reader agreement of 95% ( 0.32) and intrareader agreement of 99% ( 0.92). 20,21 LV mass was estimated using both M-mode echocardiography and 12- lead electrocardiography, as previously described. 20,22 The LV outflow tract velocity time integral and diastolic flow measures, such as mitral E- and A-wave peak velocities, were assessed using 2-dimensional directed Doppler echocardiography. 23 Valvular regurgitation and stenosis were assessed as previously described. 16 Stroke volume was estimated by multiplying the velocity time integral by LV outflow tract area, 24 assuming a LV outflow tract diameter of 2 cm and excluding patients with moderate or severe aortic stenosis (n 17) and adjusting for body mass index in multivariate models. In sensitivity analyses, we excluded participants with moderate or severe aortic regurgitation (n 416) and further standardized the assumed LV outflow tract area to the measured heel-to-knee length (a measure of body size not confounded by obesity or height loss). Cardiac output was calculated by multiplying the stroke volume by HR, and the cardiac index was calculated by dividing the cardiac output by the body surface area. Systemic vascular resistance (SVR) was calculated in Woods units as the mean BP minus an assumed central venous pressure of 8 mm Hg divided by the cardiac output; the SVR index was calculated by dividing the SVR by the body surface area. Statistical analysis: Associations were assessed using linear regression analysis for continuous outcomes and logistic regression analysis for binary outcomes, with variable transformation to approximate normality as appropriate. Fish intake was evaluated as categorical (indicator) variables. For tests for trend, participants were assigned the median intake of their respective category, and this variable was evaluated as a continuous variable. As would be expected, not all echocardiographic measures were obtainable on every participant due to suboptimal tissue penetration, limited acoustic windows, or underlying rhythm or structural irregularities; therefore, for each measure, analyses were performed on patients without missing values. Potential confounders were evaluated based on clinical relevance as factors that might influence exposures and outcomes, previously published associations, or associations with exposures or outcomes in the current data set. The final model included age, gender, race, enrollment site, education, smoking, physical activity, body mass index, diabetes, coronary heart disease, alcohol use, and total calories. For parsimony in model construction, other covariates not materially altering the relations between fish intake and the outcome measures were excluded from the final model, including annual income, treated hypertension, prevalent heart failure, and exercise intensity; use of aspirin, blockers, lipid-lowering medication, fish oil, and estrogen; and estimated intake of total fat, saturated fat, linolenic acid, carbohydrates, protein, fiber, beef/ pork, fruits, vegetables, and wine. Missing covariate values (typically 1%) were imputed using age, race, gender, diabetes, and prevalent cardiovascular disease. Analyses using the population median or excluding missing data were not appreciably different. Potential effect modification by gender, treated hypertension, and prevalent coronary heart disease was assessed in stratified analyses and by likelihood ratio testing with multiplica-

4 Preventive Cardiology/Fish Intake, Cardiac Structure, and Hemodynamics 219 Table 3 Relations between fish intake and measures of cardiac hemodynamics* Measure (n) Tuna/Other Fish Intake Fried Fish Intake (n 2,439) (n 1,634) 1/wk (n 1,000) HR (beats/min) (5,063) Systolic BP (mm Hg) (5,073) Diastolic BP (mm Hg) (5,073) Mean BP (mm Hg) (5,073) Stroke volume (ml) (3,134) E/A ratio (4,602) Cardiac output (L/min) (3,131) SVR (Woods U) (3,126) Values are presented as means (continuous variables) or percentages (categorical variables), with adjustments as in Table 2 footnote. * For each outcome measure, those for whom the outcome measure was not available were not included in the analysis. p 0.01, p 0.05 and p 0.01 compared with intake nth. Excluding 309 subjects with E/A ratio 1.5 (possible restrictive filling pattern). tive (exposure times covariate) interaction terms. All p values were 2-tailed ( 0.05). Analyses were performed using Stata, version 8.2 (College Station, Texas). Results Bivariate (unadjusted) associations between fish intake and selected participant characteristics are listed in Table 1. The relations with other participant characteristics in CHS have been previously reported. 3 6 Tuna/other fish consumption was associated with a slightly younger age, women, and higher education, and fried fish consumption was associated with men and a lower education level. Tuna/other fish and fried fish consumption were associated with a slightly higher body mass index. Fried fish intake was associated with a slightly higher prevalence of diabetes and coronary heart disease. Tuna/other fish intake was not associated with prevalent cardiovascular disease, but was inversely associated with smoking. Tuna/other fish and fried fish intake were each associated with slightly higher leisure time physical activity, but these associations were not statistically significant. The estimated dietary intake of eicosapentaenoic acid plus docosahexaenoic acid from these fish meals is also shown. 6 Cardiac structure and systolic function: Multivariateadjusted associations between fish intake and measures of cardiac structure and systolic function are listed in Table 2. Tuna/other fish intake was generally not associated with measures of cardiac structure or systolic function, except for trends toward a lower LV mass as assessed by electrocardiography (p 0.07) and a lower prevalence of reduced ejection fractions ( 55%, p 0.05). Conversely, fried fish intake was associated with several measures of LV systolic dysfunction, including reduced ejection fraction (p 0.001), the presence of LV wall motion abnormalities (p 0.02), and a trend toward a larger LV diastolic dimension (p 0.07). Cardiac hemodynamics: Tuna/other fish intake was strongly associated with measures of cardiac hemodynamics (Table 3), including lower HR (p 0.001); lower systolic, diastolic, and mean BP (p for each); greater LV stroke volume (p 0.001); and lower SVR (p 0.002). Fried fish intake was not associated with HR and was associated with less substantial differences in BP (Table 3). Fried fish intake was also associated with lower cardiac output (p 0.04) and higher SVR (p 0.003), consistent with measures of LV systolic dysfunction (Table 2). Abnormal diastolic filling can be characterized by either a low E/A ratio (seen with impaired relaxation) or a high E/A ratio (e.g., 1.5 in elderly patients, seen with restrictive filling). Consistent with this, in previous analyses in the CHS, a U-shaped relation was observed between the E/A ratio and the risk of incident heart failure. 25 Among the 4,602 patients with an E/A ratio of 1.5, tuna/other fish intake was associated with a higher E/A ratio (p 0.004; Table 3). In contrast, tuna/other fish intake was not associated with the E/A ratio among the 309 patients with an E/A ratio of 1.5 (p 0.29). No associations were seen between fried fish intake and the E/A ratio (Table 3), and fish intake was not associated with E- or A-wave peak velocities (data not shown). Additional adjustments: Exclusion of participants with moderate or severe aortic regurgitation and standardization of the assumed LV outflow tract area using heel-to-knee length had little effect on the positive relation between tuna/other fish intake and stroke volume: 63.9, 66.0, 66.7, and 67.7 ml across increasing categories of intake (p trend 0.004). Little evidence was present for confounding by use of medications. For example, tuna/other fish intake was still associated with a lower HR (p 0.001) after excluding patients on blockers (n 656) and with a lower SVR (p 0.004) after excluding patients on antihypertensives (n 2,295). Fish consumption was not associated with LV

5 220 The American Journal of Cardiology ( Table 4 Relations between tuna/other fish intake and cardiac measures after further adjustment for potentially interrelated factors Measure (n) Frequency (n)* Adjusted for HR Stroke volume (ml) (3,131) Mean BP (mm Hg) (5,063) SVR (Woods U) (3,126) Adjusted for HR and stroke volume Mean BP (mm Hg) (3,131) SVR (Woods U) (3,126) Adjusted for SVR HR (beats/min) (3,126) Stroke volume (ml) (3,126) Adjusted for HR and SVR Stroke volume (ml) (3,126) Values are presented as means, with other adjustments as in Table 2 footnote. * For each outcome measure, those for whom outcome measure was not available were not included in analysis. p 0.01, p 0.001, and p 0.01 compared with intake nth. relative wall thickness, and relations with the cardiac index and SVR index were similar to the relations with cardiac output and SVR (data not shown). Because different hemodynamic indexes may influence each other, we performed analyses simultaneously adjusting for potentially inter-related factors (Table 4). After adjustment for HR, tuna/other fish intake was still associated with stroke volume (p 0.002), BP (p 0.001), and SVR (p 0.001). After adjustment for HR and stroke volume, tuna/other fish intake was still associated with BP (p 0.001) and SVR (p 0.001). After adjustment for SVR, tuna/other fish intake was still associated with HR (p 0.001). The association of tuna/other fish intake with stroke volume was partly attenuated but still significant (p 0.004) after adjustment for SVR and fully attenuated and no longer significant (p 0.53) after adjustment for both HR and SVR. Other potential confounders: Additional adjustments for other participant characteristics (see Methods ) did not materially affect the results. For example, after further adjustment for annual income (3 categories), treated hypertension (yes/no), and prevalent heart failure (yes/no), the positive association between tuna/other fish intake and stroke volume was not materially altered (64.0, 65.4, 66.9, and 68.2 ml across increasing categories of intake, p trend 0.001). The association of fried fish intake with a reduced ejection fraction was not materially altered by simultaneous adjustment for LV diastolic dimension and SVR (p 0.002). Exclusion of patients with impaired or borderline ejection fraction ( 55%, n 458) did not materially alter the relations of fried fish intake with greater LV diastolic dimension (p 0.07) or higher SVR (p 0.02). Potential effect modifiers: Little evidence was found that the relations between fish consumption and these measures of cardiac structure, systolic function, and hemodynamics varied according to gender, treated hypertension, or the presence or absence of coronary heart disease (data not shown). The results were also not materially different if we excluded participants with prevalent heart failure (n 225) or those taking fish oil supplements (n 186, data not shown). Discussion In this large, population-based study, the consumption of tuna or other broiled or baked fish was associated with several hemodynamic parameters, including lower HR, lower SVR, and greater stroke volume. These findings represent unique data on the relations among fish intake, physical examination findings, and echocardiographic measures in 5,000 adults. The associations of tuna/other fish intake with HR and SVR each appeared independent of the other measures. In contrast, the relations of tuna/other fish intake with stroke volume and diastolic filling appeared at least partly mediated by the effects on HR and SVR. When the findings are considered together, a plausible physiologic hypothesis emerges. Fatty fish intake may lower HR and reduce SVR, improving ventricular efficiency (as manifested by higher stroke volume) and possibly diastolic filling. With time, a lower SVR could also reduce (or prevent an increase) in LV mass. In contrast to these hemodynamic measures, tuna/other fish intake was not strongly associated with other measures of cardiac structure or systolic function, suggesting that fatty fish intake may have lesser effects on these parameters. In previous reports, we have shown tuna/other fish intake to be associated with a lower risk of arrhythmic death, atrial fibrillation, ischemic stroke, and congestive heart failure. 3 6 The present findings suggest plausible physiologic mechanisms that might, in part, account for such clinical benefits.

6 Preventive Cardiology/Fish Intake, Cardiac Structure, and Hemodynamics 221 For instance, afterload reduction (lower SVR) and slowing of HR, with consequent improved stroke volume and reduced LV mass, could plausibly contribute to a lower incidence of heart failure. Beneficial effects on HR, afterload, diastolic filling, and LV mass would also reduce the risk of atrial fibrillation, and lower vascular resistance and BP would reduce the risk of stroke. Although fish oil has direct antiarrhythmic effects in animal experimental studies, 7,8 afterload reduction, reduced HR, and reduced LV mass would also contribute to a decreased risk of fatal arrhythmias. The observed relations could have resulted from other factors associated with fish intake. However, the associations were present after adjustment for a variety of sociodemographic, clinical, and lifestyle factors. The specificity of the relations was also notable: tuna/other fish intake was strongly associated with favorable cardiac hemodynamics, but not with cardiac structure or systolic function. It is biologically plausible that such effects are related to marine n-3 fatty acids in fish. Several findings in this populationbased study of dietary fish intake are consistent with the results of short-term experimental fish oil trials. In rats, fish oil consumption reduced SVR 26 and reduced LV hypertrophy in hypertensive animals. 27 In nonhuman primates, fish oil lowered HR at rest, improved myocardial efficiency in response to sympathetic stimulation, and enhanced LV diastolic filling. 28,29 In short-term trials in humans, fish oil supplementation attenuated peripheral vasoconstrictive responses to norepinephrine and angiotensin II 9,10 and improved LV diastolic filling. 11 In a meta-analysis of placebocontrolled trials (median duration 8 weeks), fish oil intake reduced systolic and diastolic BP by 3.5 and 2.4 mm Hg, respectively, among adults aged 45 years (p 0.01 for each) 12 compared with 5- and 2-mm Hg differences seen with dietary fish intake in the present study. In another meta-analysis, in fish oil trials with a duration of 12 weeks, fish oil intake reduced HR by 2.5 beats/min (p 0.001) 13 compared with a difference of 3 beats/min seen with dietary fish intake in the present study. In these 2 meta-analyses, little evidence was found that very high doses of fish oil had higher effects than modest doses, 12,13 and modest doses are achievable with regular dietary fish intake. Thus, the consistency between our findings and experimental fish oil trials supports a causal effect of fish intake on these hemodynamic parameters. Investigation of the molecular mechanisms whereby fish intake may influence these cardiac parameters and the relevance of such effects to cardiovascular health is warranted. Fried fish intake was associated with LV wall motion abnormalities, reduced ejection fraction, reduced cardiac output, and, possibly, larger LV diastolic dimension. The confluence of these abnormalities suggests LV systolic dysfunction and probable coronary atherosclerosis. Fried fish intake was also associated with a higher mean BP and SVR, even when the ejection fraction was normal. In previous reports, we have observed positive associations or trends toward positive associations between fried fish intake and new-onset nonfatal myocardial infarction, atrial fibrillation, and congestive heart failure. 3,4,6 In each case, adjustment for ejection fraction and BP or hypertension in part attenuated the observed higher risk, 3,4,6 suggesting potential mediation by these factors. The present findings provide additional evidence that associations of fried fish intake with cardiovascular risk relate to LV systolic dysfunction, possibly related to greater severity of coronary atherosclerosis and/or increased afterload. Whether these results are due to the preparation method or to residual confounding from other lifestyle factors requires additional study. This study had potential limitations. Missing data on echocardiographic measures might not have been missing at random, but rather related to conditions such as obesity, pulmonary disease, or underlying cardiac abnormalities. Multiple exposure-outcome relations were evaluated, and associations of borderline statistical significance should be interpreted in this context. Although the diet questionnaire assessed the usual dietary intake in the year before the echocardiogram, a cross-sectional analysis cannot establish temporality. Also, these associations were observed in older, predominantly white patients participating in a cohort study and might not be generalizable to other populations. More detailed information on the fish species consumed or other preparation methods was not available. Although a range of covariates was available in the CHS and were evaluated as potential confounders, residual confounding from unknown or incompletely measured factors could not be excluded. However, the magnitude and specificity of the relations and, for several findings, the consistency with short-term experimental trials of fish oil intake, make it unlikely that residual confounding accounted entirely for the observed relations. The assessments of fish intake and cardiac parameters were imperfect, producing bias toward the null and resulting in an underestimation of the extent of the associations. This would be especially true for outcomes such as stroke volume, cardiac output, and SVR, which represent a combination of 2 measures or assumptions. Therefore, the magnitude of the true associations between fish intake and these cardiac parameters are likely larger than observed. Acknowledgment: The investigators are grateful to the CHS participants, without whom this research would not be possible. For a full list of CHS investigators and institutions, see About CHS-Principal Investigators and Study Sites at 1. Siscovick DS, Raghunathan TE, King I, Weinmann S, Wicklund KG, Albright J, Bovbjerg V, Arbogast P, Smith H, Kushi LH, et al. Dietary intake and cell membrane levels of long-chain n-3 polyunsaturated fatty acids and risk of primary cardiac arrest. JAMA 1995;274: Albert CM, Hennekens CH, O Donnell CJ, Ajani UA, Carey VJ, Willett WC, Ruskin JN, Manson JE. Fish consumption and risk of sudden cardiac death. JAMA 1998;279:23 28.

7 222 The American Journal of Cardiology ( 3. Mozaffarian D, Lemaitre RN, Kuller LH, Burke GL, Tracy RP, Siscovick DS. Cardiac benefits of fish consumption may depend on the type of fish meal consumed: the Cardiovascular Health Study. Circulation 2003;107: Mozaffarian D, Psaty BM, Rimm EB, Lemaitre RN, Burke GL, Lyles MF, Lefkowitz D, Siscovick DS. Fish intake and risk of incident atrial fibrillation. Circulation 2004;110: Mozaffarian D, Longstreth WT, Lemaitre RN, Manolio T, Kuller LH, Burke GL, Siscovick DS. Fish consumption and stroke risk in elderly individuals: the Cardiovascular Health Study. Arch Intern Med 2005; 165: Mozaffarian D, Bryson CL, Lemaitre CL, Burke GL, Siscovick DS. Fish intake and risk of incident heart failure. J Am Coll Cardiol 2005;45: Kang JX, Leaf A. Prevention of fatal cardiac arrhythmias by polyunsaturated fatty acids. Am J Clin Nutr 2000;71:202S 207S. 8. McLennan PL. Myocardial membrane fatty acids and antiarrhythmic actions of dietary fish oil in animal models. Lipids 2001;36:111S 114S. 9. Kenny D, Warltier DC, Pleuss JA, Hoffmann RG, Goodfriend TL, Egan BM. Effect of omega-3 fatty acids on the vascular response to angiotensin in normotensive men. Am J Cardiol 1992;70: Chin JP, Gust AP, Nestel PJ, Dart AM. Marine oils dose-dependently inhibit vasoconstriction of forearm resistance vessels in humans. Hypertension 1993;21: Grimsgaard S, Bonaa KH, Hansen JB, Myhre ES. Effects of highly purified eicosapentaenoic acid and docosahexaenoic acid on hemodynamics in humans. Am J Clin Nutr 1998;68: Geleijnse JM, Giltay EJ, Grobbee DE, Donders AR, Kok FJ. BP response to fish oil supplementation: metaregression analysis of randomized trials. J Hypertens 2002;20: Mozaffarian D, Geelen A, Brouwer IA, Geleijnse JM, Zock PL, Katan MB. The effect of fish oil on HR in humans: a meta-analysis of randomized controlled trials. Circulation 2005;112: Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, Kuller LH, Manolio TA, Mittelmark MB, Newman A, et al. The Cardiovascular Health Study: design and rationale. Ann Epidemiol 1991;1: Psaty BM, Kuller LH, Bild D, Burke GL, Kittner SJ, Mittelmark M, Price TR, Rautaharju PM, Robbins J. Methods of assessing prevalent cardiovascular disease in CHS. Ann Epidemiol 1995;5: Gardin JM, Wong ND, Bommer W, Klopfenstein HS, Smith VE, Tabatznik B, Siscovick D, Lobodzinski S, Anton-Culver H, Manolio TA. Echocardiographic design of a multicenter investigation of freeliving elderly subjects: CHS. J Am Soc Echocardiogr 1992;5: Kumanyika SK, Tell GS, Shemanski L, Martel J, Chinchilli VM. Dietary assessment using a picture-sort approach. Am J Clin Nutr 1997;65:1123S 1129S. 18. Willett W, Stampfer MJ. Total energy intake-implications for epidemiologic analyses. Am J Epidemiol 1986;124: Sahn DJ, DeMaria A, Kisslo J, Weyman A, for the Committee on M-mode Standardization of the American Society of Echocardiography. Recommendations regarding quantitation in M-mode echocardiography results of a survey of echocardiographic methods. Circulation 1978;58: Gardin JM, Siscovick D, Anton-Culver H, Lynch JC, Smith VE, Klopfenstein HS, Bommer WJ, Fried L, O Leary D, Manolio TA. Sex, age, and disease affect echocardiographic left ventricular mass and systolic function in the free-living elderly: the Cardiovascular Health Study. Circulation 1995;91: Gottdiener JS, McClelland RL, Marshall R, Shemanski L, Furberg CD, Kitzman DW, Cushman M, Polak J, Gardin JM, Gersh BJ, Aurigemma GP, Manolio TA. Outcome of congestive heart failure in elderly persons: influence of left ventricular systolic function the Cardiovascular Health Study. Ann Intern Med 2002;137: Rautaharju PM, Manolio TA, Siscovick D, Zhou SH, Gardin JM, Kronmal R, Furberg CD, Borhani NO, Newman A. Utility of new electrocardiographic models for left ventricular mass in older adults. Hypertension 1996;28: Rokey R, Kuo LC, Zoghbi WA, Limacher MC, Quinones MA. Determination of parameters of left ventricular diastolic filling with pulsed Doppler echocardiography: comparison with cineangiography. Circulation 1985;71: Oh JK, Seward JB, Tajik AJ, editors. The Echo Manual. 2nd ed. New York: Lippincott Williams & Wilkins, Aurigemma GP, Gottdiener JS, Shemanski L, Gardin J, Kitzman D. Predictive value of systolic and diastolic function for incident congestive heart failure in the elderly: the Cardiovascular Health Study. JAm Coll Cardiol 2001;37: Demaison L, Blet J, Sergiel JP, Gregoire S, Argaud D. Effect of dietary polyunsaturated fatty acids on contractile function of hearts isolated from sedentary and trained rats. Reprod Nutr Dev 2000;40: von Au D, Brandle M, Rupp H, Jacob R. Influence of a diet rich in fish oil on BP, body weight and cardiac hypertrophy in spontaneously hypertensive rats. Eur J Appl Physiol Occup Physiol 1988;58: Charnock JS, McLennan PL, Abeywardena MY. Dietary modulation of lipid metabolism and mechanical performance of the heart. Mol Cell Biochem 1992;116: McLennan PL, Barnden LR, Bridle TM, Abeywardena MY, Charnock JS. Dietary fat modulation of left ventricular ejection fraction in the marmoset due to enhanced filling. Cardiovasc Res 1992;26: Warner K. Impact of high-temperature food processing on fats and oils. Adv Exp Med Biol 1999;459: Ammu K, Raghunath MR, Sankar TV, Lalitha KV, Devadasan K. Repeated use of oil for frying fish: effects of feeding the fried fish to rats. Nahrung 2000;44: Al-Saghir S, Thurner K, Wagner KH, Frisch G, Luf W, Razzazi-Fazeli E, Elmadfa I. Effects of different cooking procedures on lipid quality and cholesterol oxidation of farmed salmon fish. J Agric Food Chem 2004;52: