点击显示 收起
1 From the Cardiovascular Nutrition Laboratory, Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston (ATE and AHL); the Department of Clinical Nutrition, University of Kuopio, Kuopio, Finland (ATE); the Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, and the Departments of Epidemiology and Nutrition, Harvard School of Public Health, Boston (DM); and the Department of Internal Medicine/Cardiology, Wake Forest University School of Medicine, Winston-Salem, NC (DMH)
2 Based on work supported by the US Department of Agriculture under agreement no. 58-1950-4-401. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the US Department of Agriculture.
3 Supported by grants from the National Heart, Lung, and Blood Institute (U01-HL-45488); the National Center for Research Resources, General Clinical Research Center (Mo1 rr07122); the US Department of Agriculture (58-1950-9-001); and the Academy of Finland (80232 and 79433).
4 Address reprint requests to AH Lichtenstein, Cardiovascular Nutrition Laboratory, Jean Mayer USDA-HNRCA, Tufts University, 711 Washington Street, Boston, MA 02111. E-mail: alice.lichtenstein{at}tufts.edu.
See corresponding editorial on page 535.
ABSTRACT
Background: Higher intakes of fish and n–3 fatty acids are associated with a reduced risk of cardiovascular events and mortality. However, limited data exist on the effect of fish intake on actual measures of progression of coronary artery atherosclerosis.
Objective: The aim was to examine the association between fish intake and the progression of coronary artery atherosclerosis in women with coronary artery disease.
Design: This was a prospective cohort study of postmenopausal women (n = 229) participating in the Estrogen Replacement and Atherosclerosis trial. Usual fish intake was estimated at baseline with a food-frequency questionnaire. Quantitative coronary angiography was performed at baseline and after 3.2 ± 0.6 (
Conclusions: Consumption of fish is associated with a significantly reduced progression of coronary artery atherosclerosis in women with coronary artery disease.
Key Words: Fish • atherosclerosis • angiography • women • diet • n–3 fatty acids • diabetes mellitus
INTRODUCTION
The consumption of fish and the long-chain n–3 fatty acids eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) is associated with a reduced risk of cardiovascular events and mortality (1–6). Some of the favorable associations between n–3 fatty acid intake and low rates of cardiovascular disease mortality may be related to a decreased risk of sudden death (2, 4, 7, 8). However, several lines of evidence suggest that n–3 fatty acid intake may also have favorable effects on the pathogenesis or progression of atherosclerosis (9–12). Direct evidence linking fish intake to the rate of progression of coronary disease is limited and, to date, is not available in women (13–15). On the basis of observational data, Hu et al (16) recently reported that a high fish intake is associated with a reduced risk of coronary artery disease (CAD) and total mortality in diabetic women. It is unclear whether there is a direct relation between fish intake and the progression of coronary atherosclerosis. The aim of this study was to examine the relation between fish consumption and the progression of angiographically defined coronary atherosclerosis in a group of postmenopausal women undergoing baseline and 3-y follow-up coronary angiography as part of a randomized clinical trial of hormone replacement therapy (HRT) (17).
SUBJECTS AND METHODS
Subjects
The Estrogen Replacement and Atherosclerosis trial was a randomized, double-blind, placebo-controlled trial of HRT. The study design and primary results were reported previously (17), and the results on fat intake and atherosclerosis progression were reported separately (18). Briefly, postmenopausal women with coronary stenoses of 30% of the luminal diameter were randomly assigned into 3 groups and received 1) 0.625 mg conjugated equine estrogen (n = 100), 2) 0.625 mg conjugated equine estrogen plus 2.5 mg medroxyprogesterone acetate (n = 104), or 3) placebo (n = 105). At baseline, the subjects completed questionnaires about their health status, medical history, dietary intake, and cardiovascular disease risk factors and underwent clinical examination and quantitative coronary angiography. The subjects were classified as having diabetes if their fasting glucose concentration was 7.0 mmol/L or their 2-h glucose was 11.1 mmol/L during an oral-glucose-tolerance test, their glycated hemoglobin value was >7% (19), or they reported having diabetes or the use of diet, oral hypoglycemics, or insulin as treatment for diabetes. The mean (±SD) period between angiographic measurements was 3.2 ± 0.6 y. The study protocol was approved by the Institutional Review Board at the participating sites (17) and at the New England Medical Center, and Tufts University, Boston. All subjects gave their informed written consent before participating in the study.
Dietary assessment
Usual dietary intake during the year before the baseline measurements was assessed with a validated, semiquantitative, 126-item food-frequency questionnaire as described previously (20). Dietary information was judged as unreliable and excluded from further analysis if reported energy intakes were <660 or >3500 kcal/d or >11 food item was left blank (n = 8); 282 reliable food-frequency questionnaires were completed. Frequency of fish consumption was calculated by summing the frequency of intake of tuna (84-112 g/serving) dark fish (84-140 g/serving), or other fish (84-140 g/serving). Because tuna and dark fish are the major dietary contributors to the long-chain n–3 fatty acid intake, their intake was calculated by summing these 2 intakes alone.
Laboratory measurements
Serum lipids were analyzed by using standardized enzymatic methods (17). Glucose and glycated hemoglobin were measured as previously described (17, 21). C-reactive protein (CRP) was measured at baseline with the use of a high-sensitivity kit (American Laboratory Products Co, Windham, NH) (22). Interleukin 6 (IL-6) and vascular and intercellular adhesion molecules (VCAM-1 and ICAM-1, respectively) were similarly measured by using standardized enzyme-linked immunoassay techniques.
Outcome measurements
Quantitative coronary angiography was performed by using standardized techniques at baseline and at the end of follow-up in 248 women (
All statistical analyses were performed by using SAS (version 8; SAS Institute, Cary, NC). The normality of continuous variables was checked, and log transformations were applied as needed. Differences in baseline characteristics and nutrient intakes were tested between the different intake categories by using independent-samples t test, Wilcoxon’s rank-sum test, or chi-square test, as appropriate. The association between total fish intake and different types of fish intake was tested by using the test parameters of changes in mean minimum coronary artery diameter and mean percentage stenosis with the use of mixed-model analysis of covariance (ANCOVA). The model allows for differential variation in disease progression across different coronary segments and correlation in segment changes within a given participant. These measurements were adjusted for age, the location of the coronary segment, the time of follow-up, study clinic, coronary artery bypass grafting (CABG), percutaneous transluminal coronary angioplasty (PTCA), race, body mass index (BMI; in kg/m2), smoking, use of cholesterol-lowering medication, HRT, diabetes, and energy intake (model 1). Model 2 included adjustments for the factors in model 1 and energy-adjusted intakes of saturated fat, monounsaturated fat, polyunsaturated fat, cholesterol, fiber, and alcohol. Further adjustments for blood pressure; serum concentrations of total cholesterol, HDL cholesterol, LDL cholesterol, triacylglycerol, and inflammatory markers; education, and strenuous physical activity were also made. Differences in the development of new lesions were tested with ANCOVA [general linear models (GLM)], with adjustment for study clinic, time of follow-up, CABG, PTCA, use of cholesterol-lowering medication, HRT, diabetes, energy intake, and energy-adjusted intakes of saturated fat, monounsaturated fat, polyunsaturated fat, cholesterol, fiber, and alcohol. The interaction between fish intake and diabetes status was tested with ANCOVA, and if the P value for the interaction term was <0.10, ANCOVA (without adjustment for diabetes) was performed according to diabetes status. Spearman correlation coefficients adjusted for BMI were calculated between inflammatory markers and fish intake. A value of P < 0.05 (two-tailed) was considered statistically significant.
RESULTS
Most of the baseline characteristics, including concentrations of serum lipids and inflammatory markers and blood pressure did not differ significantly between women who reported <2 or 2 servings of fish per week (Table 1). Women who ate 2 servings of fish per week had a higher educational level and were engaged more frequently with strenuous physical activity. Women who consumed 2 servings of fish per week reported higher intakes of energy, protein, cholesterol, alcohol, and carotene and lower intakes of carbohydrates (Table 2).
View this table:
TABLE 1. Baseline characteristics according to fish consumption
View this table:
TABLE 2. Composition of the diet according to fish consumption
As previously reported, neither estrogen alone nor estrogen plus medroxyprogesterone acetate compared with placebo affected the progression of atherosclerosis in this population (17); however, the HRT was controlled in the multivariate models that assessed progression. Because of a prior report suggesting a selectively beneficial effect of fish intake in diabetic women on CAD events (16), we analyzed our data on the basis of diabetes. Forty-two percent of the women participating in this study were classified as having diabetes (Table 1). Change in minimum coronary artery diameter was significantly smaller in women who consumed 2 servings of fish per week than in those who consumed <2 fish servings per week (P for fish intake = 0.02) (Table 3). Adjustment for age, location of coronary artery segment, time of follow-up, study clinic, CABG, PTCA, race, BMI, smoking, use of cholesterol-lowering medication, HRT, diabetes, and energy intake (model 1) diminished the relation (P = 0.06). Further adjustment for dietary variables known to affect CAD risk, saturated fat, monounsaturated fat, polyunsaturated fat, cholesterol, fiber, and alcohol (model 2) strengthened the association between fish intake and change in minimum coronary artery diameter (P = 0.006).
View this table:
TABLE 3. Angiographic measures according to fish consumption and diabetes status
Among diabetic women, baseline mean percentage stenosis was greater in those who consumed 2 servings of fish per week (Table 3). Compared with the lower fish intake, the change in percentage stenosis was smaller in women who consumed 2 servings of fish per week (P < 0.001). This association was restricted to diabetic women after adjustments in models 1 and 2. Subsequent adjustments of model 2 for blood pressure; serum concentrations of total cholesterol, HDL cholesterol, LDL cholesterol, triacylglycerol, and inflammatory markers; education; or strenuous physical activity did not appreciably alter the relations between fish intake and changes in minimum coronary artery diameter and percentage stenosis (data not shown).
Women who consumed 2 servings of fish per week had significantly fewer new lesions (P = 0.02, ANCOVA after adjustment for the specified factors) (Table 3). Seventeen (21%) of the women consuming 2 servings of fish per week and 51 (34%) of the women consuming <2 servings of fish per week developed at least one new lesion (chi-square test, P = 0.03).
Tuna and dark fish are the major contributors of long-chain n–3 fatty acids in the American diet. The average content of EPA and DHA in tuna (0.3-1.3 g/serving) and dark fish (0.7-1.8 g/serving) is at least twice that of fish in the "other fish" category (0.1-0.4 g/serving) (23). Therefore, it was of interest to determine whether the type of fish influenced atherosclerosis progression. Women who reported consuming 1 serving of tuna or dark fish per week had a smaller change in minimum coronary artery diameter (P = 0.02) (Table 4). This association was diminished and limited to diabetic women when adjusted for the factors in model 1. The association became significant when adjusted for the factors in model 2, which suggests an independent effect of tuna and dark fish intake (P = 0.02). Among diabetic women, baseline stenosis was greater and changes in percentage stenosis were smaller (models 1 and 2) in those who consumed 1 serving of tuna and dark-meat fish per week. The relations between tuna and dark fish intake and percentage stenosis were not significant in nondiabetic women. Subsequent adjustments of model 2 for blood pressure levels, serum concentrations of lipids and inflammatory markers, education, or strenuous physical activity did not appreciably alter the relations between tuna and dark fish intakes and angiographic measures (data not shown).
View this table:
TABLE 4. Angiographic measures according to different types of fish consumption and diabetes status
Mean baseline minimum coronary artery diameter was smaller and percentage stenosis was greater in the diabetic women who consumed 1 serving of "other fish" than in those who reported lower intakes (Table 4). In nondiabetic women, those reporting 1 serving of "other fish" per week had smaller changes in minimum coronary artery diameter and percentage stenosis. The associations between angiographic changes and consumption of other fish were significant when factors in model 2 were adjusted (P = 0.04 and P = 0.05, respectively); however, the interaction between diabetes status and intake of other fish was not significant. There were also fewer segments with new lesions in women who consumed 1 serving of "other fish" per week.
To determine whether the associations of fish intake were related to markers of inflammation, we evaluated correlations between fish intake and concentrations of inflammatory markers after adjustment for BMI, because CRP concentrations have been suggested to be directly associated with BMI (24). Fish intake was inversely correlated with concentrations of VCAM-1 in all women (Table 5). Concentrations of CRP, IL-6, and ICAM-1 were not significantly correlated with fish consumption.
View this table:
TABLE 5. Spearman correlation coefficients adjusted for BMI between fish consumption and inflammatory markers
DISCUSSION
Fish and n–3 fatty acid intakes have received considerable attention in the recent past as potential dietary factors to reduce the risk of developing CAD (4–6, 25). This growing body of data culminated with a 2002 recommendation by the American Heart Association for persons without documented CAD to eat fish twice or more per week and for patients with documented CAD to consume 1 g EPA + DHA/d (26). The importance of these and other recommendations for diabetic persons was highlighted when the National Cholesterol Education Program (27) designated diabetes as a CAD risk equivalent. This may be particularly important in women, in whom the adverse effects of diabetes may be greater than those in men (28).
There are no previous data on fish intake and atherosclerosis progression in women. The present results extend prior observational data on CAD risk by showing that fish intake is associated with reduced progression of atherosclerosis in postmenopausal women with established CAD within the 3-y observational period of the study. Subsequent analysis indicated that this association was strongest in the subgroup of women with diabetes. The interaction terms between fish intake and diabetes were not significant in all of the analyses. The possibility that this observation was due to limited statistical power cannot be ruled out. Nevertheless, within the 3-y period, new lesions were observed in all women regardless of fish intake. Data from the current study are consistent with recent observational data, suggesting an association with fish and n–3 fatty acid intakes and a decreased incidence of CAD and total mortality in diabetic women (16). These data can be interpreted in 2 ways. The observation of a significant association of fish intake in diabetic, but not in nondiabetic women, may be due to a higher degree of sensitivity of diabetic women to n–3 fatty acids. However, the possibility cannot be excluded that the observation may reflect a more rapid rate of atherosclerosis progression in the diabetic women, hence, a greater likelihood that a positive association would be observed within the observational period.
The lower rate of angiographically documented atherosclerosis progression was significantly associated with tuna and dark fish intake, whereas these associations appeared to be weaker for the intake of "other fish." This finding is likely a reflection of the higher n–3 fatty acid content of tuna and dark fish than of "other fish." Alternatively, it may be related to the method used for preparation of tuna and dark fish relative to that for white fish. The later type of fish is more likely to be fried in fat that is high in either trans or saturated fatty acids (25). The findings of the current study are consistent with those of previous studies, which documented an inverse association between dark fish intake and total mortality and CAD morbidity in diabetic women participating in the Nurses’ Health Study (16). Furthermore, intake of tuna or other broiled or baked fish has been shown to be associated with a reduced risk of cardiac death in female and male subjects participating in the Cardiovascular Health Study (25). Similarly, intake of fatty fish was associated with a lower risk of cardiac death in men participating in the Seven Countries Study (29).
Although fish intake and fish-oil supplements have been associated with a reduced risk of clinical events, the effect of fish or n–3 fatty acid intake on the progression of atherosclerosis has been controversial. Bairati et al (30) reported that dietary n–3 fatty acid intake was associated with a lower frequency of restenosis at 6 mo after PTCA in men, whereas Watts et al (31) reported no association between atherosclerosis progression and dietary intakes of EPA and DHA in men after 3 y of follow-up. Less progression of atherosclerosis has been reported after 1 y of treatment with fish-oil supplements in CABG patients (14) and CAD patients after 2 y of treatment (15). In contrast, Sacks et al (13) reported that fish-oil supplementation did not alter the progression of atherosclerosis in CAD patients over a 28-mo period. Self-reported intake of fish at least twice a week has been shown to be associated with a lower prevalence of myocardial lesions in men at autopsy (32). Fish-oil supplementation has been reported to have little effect on restenosis after PTCA (33–35). However, restenosis is a pathophysiologically distinct process compared with progression of native atherosclerosis. There are no data focusing on fish intake and atherosclerosis progression in women or in diabetic subjects. In most cases, the data sets include 80% male subjects (13–15, 34), and data are not reported separately for female subjects.
Fish and n–3 fatty acids are suggested to have antiinflammatory properties (12) that may contribute to the reduced risk of CAD, especially in diabetic subjects (36). Concentrations of most inflammatory markers, with the exception of VCAM-1, were not associated with fish intake in this study. VCAM-1, which is mainly expressed on atherosclerotic plaques, is suggested to predict CAD risk, especially in subjects with advanced atherosclerosis (36). These data may have been confounded by the relatively high BMI or proportion of diabetic women in the study population. These data are consistent with previous work suggesting that fish-oil supplementation does not affect concentrations of CRP and IL-6 (37) or ICAM-1 and VCAM-1 (38).
Our study had certain limitations. Higher fish intakes were associated with a healthier lifestyle. Although these factors were controlled for during the analysis, residual confounding cannot be ruled out. The results suggest that intake of tuna and dark fish is more strongly associated with a reduction in progression than is the intake of other types of fish. Because the latter category of fish is more likely to not only be fried but also to be consumed with other fried foods (eg, fried potatoes), such a dietary pattern could have contributed to the observed outcome (25). There have been some concerns that fish contaminated with mercury could diminish the benefits of fish (39). We were not able to evaluate the effect of mercury because there was no data collected on mercury exposure.
In conclusion, fish consumption was associated with a significantly reduced progression of coronary atherosclerosis in postmenopausal women with CAD. This relation was strongest in diabetic women. The findings support the view that regular fish consumption is a part of a healthy diet.
ACKNOWLEDGMENTS
DMH contributed to the original planning of the Estrogen Replacement and Atherosclerosis Study. ATE, AHL, and DMH contributed to the study design. ATE was responsible for the statistical analyses and the writing of the first draft. All authors contributed to the data interpretation and revision of the manuscript. There were no conflicts of interest.
REFERENCES