| 31) Tomasallo C (2010) | Cohort name not defined | 2 cohorts were formed:
-2,123 captains/spouses, and anglers.
-1,367 referents.
(USA) | 12
(1995-2006) | See variables | All-cause mortality | Total fish (captains were frequent consumers of fish from the Great Lakes, while referents did not consume fish from the Great Lakes, but ate commercial fish) |
Captains/spouses, and anglers
(177 cases) | Referents
(128 cases) |
HR = 1.00 (0.63-1.58) for the highest vs lowest tertile of consumption.
Amount specific data (Frequency of consumption):
0-1/month: HR = 1.
> 1/month to < 1/week: HR = 1.14 (0.72-1.83).
≥ 1/week: HR = 1.00 (0.63-1.58).
|
HR = 0.60 (0.38-0.95; P = < 0.05) for the highest vs lowest tertile of consumption.
Amount specific data (Frequency of consumption):
0-1/month: HR = 1.
> 1/month to < 1/week: HR = 0.76 (0.49-1.17).
≥ 1/week: HR = 0.60 (0.38-0.95).
|
No significant interaction was found between fish intake and income, or fish intake and BMI.
The size of the fish meal or number of ounces consumed was not obtained.Sex, age, BMI, and income. |
| 30) González S (2008) | No cohort name defined | 288 elderly people aged 60-85 from nursing homes in Asturias (Northern Spain) | 6
(1999-2002 to 2007) | 83 | All-cause mortality | Fish and shellfish (not defined) | RR = 1.141 (0.853-1.525; P = 0.375) per 1 SD increase. | Age, gender, BMI, energy intake, hyperglycemia, hypercholesterolemia, chewing ability, physical activity, smoking, self perceived health, education level and institution. |
| 29) Yamagishi K (2008) | The JACC Study | 57,972 subjects (22,881 men and 35,091 women) aged 40-79, and without heart disease, stroke, or cancer.
(Japan) | 12.7
(1988-90 to 1999-2003) | 7,008 | Total death | Fish (fresh fish, kamaboko [steamed fish paste], dried or salted fish, and deep-fried foods or tempura [a common form of deep-fried fish or shellfish]) |
HR = 0.92 (0.85-1.00; P = 0.08) for the highest vs lowest quintile of consumption.
Amount specific data (g/day):
20.5: HR = 1.
33: HR = 0.96 (0.89-1.04).
45.5: HR = 0.98 (0.90-1.05).
62: HR = 0.96 (0.89-1.04).
85.5: HR = 0.92 (0.85-1.00).
No differences were found by gender in the association. | Energy, history of hypertension and diabetes mellitus, smoking status, alcohol consumption, BMI, mental stress, walking, sports, education, total energy, dietary intakes of cholesterol, saturated and omega 6 polyunsaturated fatty acids, vegetables, and fruits. |
| 29) Iso H (2007) | The JACC Study | 43,918 men, and 59,796 women.
(Japan) | Not defined. | 9,534 men, and
6,444 women. | All-cause mortality | Fresh fish (not defined) |
| Men: | Women: |
HR = 0.94 (0.89-0.99; P = < 0.05) for the highest vs lowest tertile of consumption.
Amount specific data (Frequency of intake):
< 3/w: HR = 1.
3-4/w: HR = 0.93 (0.88-0.98; P = < 0.01).
≥ 5/w: HR = 0.94 (0.89-0.99).
|
HR = 0.94 (0.88-1.00; P = < 0.05) for the highest vs lowest tertile of consumption.
Amount specific data (Frequency of intake):
< 3/w: HR = 1.
3-4/w: HR = 0.97 (0.91-1.03).
≥ 5/w: HR = 0.94 (0.88-1.00).
|
Age and study area. |
| 28) Knoops KT (2006) | The HALE (Healthy Ageing: a Longitudinal study in Europe).
(Including subjects of the SENECA and FINE study) | 2,044 men and, 1,049 women aged 70-90 from 10 European countries.
(The FINE Study included the survivors of 5 cohorts [from Finland, Italy, and The Netherlands] of The Seven Countries Study) | 10
(1988-2000) | See variables | All-cause mortality | Fish (not defined) |
Total population
(1,384? cases) | Northern Europe
(796? cases) | Southern Europe
(588? cases) |
| HR = 0.89 (0.82-0.97) for above median vs below median consumption.
|
HR = 0.89 (0.75-1.04) for above median vs below median consumption.
|
HR = 0.88 (0.78-0.99) for above median vs below median consumption.
|
Age, gender, physical activity, smoking, alcohol use, education, BMI, chronic disease at baseline and study centre. |
| 28) Iestra J (2006) | The HALE
(Including the SENECA study, and the FINE study. The FINE study is an extension of "The Seven Countries Study" beyond the 25 years follow-up) | 462 men and women age ≥ 70 with a history of MI | 10 | Not defined (53% died) | All-cause mortality | Fish (not defined) |
HR = 0.80 (0.70-1.18) for above vs below median consumption.
Stratified by sex:
| Men | Women |
| HR = 0.79 (0.59-1.06) for above vs below median consumption.
|
HR = 1.04 (0.57-1.89) for above vs below median consumption.
|
Stratified by latitude:
| Northern Europe | Southern Europe |
| HR = 0.81 (0.58-1.13) for above vs below median consumption.
|
HR = 1.20 (0.74-1.97) for above vs below median consumption.
|
Study, gender, age, years of education, BMI, history of diabetes or stroke, smoking, physical activity, alcohol |
| 27) Ness AR (2005) | The Boyd Orr Cohort. | 4,028 people (1,995 men, and 2,033 women) from England and Scotland. | average 37
(1948 to 2000) | 1,010 | All-cause mortality | Fish (not defined) |
RR = 0.98 (0.79-1.20; P = 0.8) for the highest vs lowest quartile of consumption.
Amount specific data (g):
0-6.9: RR = 1.
7.0-15.7: RR = 0.92 (0.76-1.12).
15.8-29.2: RR = 0.93 (0.77-1.13).
29.2-148.9: RR = 0.98 (0.79-1.20). | Age, energy, sex, childhood family food expenditure, father's social class, district of residence as a child, period of birth, season when studied as a child, and Townsend score for current address or place of death. |
| 26) Nakamura Y (2005) | The NIPPON DATA80 | 8,879 subjects (3,945 men, and 4,934 women) aged ≥ 30, and without coronary disease, stroke, cancer, or significant comorbidities such as renal insufficiency.
(Japan) | 19
(1980-1999) | 1,745 | Death from all causes | Fish (not defined) |
RR = 1.00 (0.77-1.29; P = 0.98) for the highest vs second quintile of consumption.
Amount specific data (Frequency of consumption):
< 1/wk: RR = 1.12 (0.87-1.44).
1-2/wk: RR = 1 (Reference group).
1/2 day: RR = 0.98 (0.85-1.12).
1/day: RR = 1.03 (0.88-1.20).
≥ 2/day: RR = 1.00 (0.77-1.29).
Additional adjustment for serum total cholesterol did not change the association. | Age, sex, smoking, alcohol drinking, hypertension, BMI, and diabetes. |
| 25) Kelemen LE. (2005) | The Iowa Women's Health study. | 29,017 postmenopausal women aged 55-69.
(USA) | 15 (1986-2000) | 3,978? | All-cause mortality | Fish (fresh fish, canned fish, and seafood) | RR = 0.97 (0.86-1.09; P = 0.34) for the highest vs lowest quintile (median servings per 1,000 kcal between extreme quintiles = 0 and 0.28) of substitution for an isoenergetic amount of all carbohydrate-rich foods*
Amount specific data (quintiles. No amounts specified):
Q1: RR = 1.
Q2: RR = 0.91.
Q3: RR = 0.94.
Q4: RR = 0.91.
Q5: RR = 0.97.
*Carbohydrate rich foods [defined as: A composite of refined carbohydrates (rice, pasta, potatoes, refined cold breakfast cereal, muffins, snack foods, sweetened sodas, pizza, chocolate, candy, cakes, cookies, donuts, pastries, pies) and whole-grain carbohydrates (dark bread, brown rice, oatmeal, whole-grain breakfast cereal, bran, wheat germ, and other grains such as bulgar, kasha, and couscous)]. | Age, total energy, saturated fat, polyunsaturated fat, monounsaturated fat, trans-fat, total fiber, dietary cholesterol, dietary methionine, alcohol, smoking, activity level, BMI, history of hypertension, postmenopausal hormone use, multivitamin use, vitamin E supplement use, education, family history of cancer, servings of fruits and vegetables excluding potatoes, legumes, dairy, eggs, red meats, poultry, and fish. |
| 25) Folsom AR (2004) | The Iowa Women's Health Study | 41,836 women aged 55-69, and free of heart disease or cancer. | 1986-2000 | Women initially free of heart disease: 4,653.
Diabetic women: ?
Women with a history of heart disease: 1,069. | Total mortality | Total fish and seafood (Dark-meat fish [such as mackerel, salmon, sardines, bluefish, or swordfish]; Canned tuna; Other fish;, and Shrimp, lobster, or callops as a main dish) |
| Women initially free of heart disease: | Diabetic women:
(Age, and energy-adjusted) | Women with a history of MI, angina, or other heart disease: |
RR = 0.93 (0.83-1.05; P = 0.15) for the highest vs lowest quintile of consumption.
Amount specific data (servings/week):
< 0.5: RR = 1.
0.5-< 1.0: RR = 0.99 (0.88-1.11).
1.0-1.5: RR = 0.97 (0.88-1.07).
> 1.5-< 2.5: RR = 0.93 (0.83-1.05).
≥ 2.5: RR = 0.93 (083-1.05).
|
RR = 0.92 (P = 0.78) for the highest vs lowest quintile of consumption.
Amount specific data (servings/week):
< 0.5: RR = 1.
0.5-< 1.0: RR = 0.97.
1.0-1.5: RR = 0.90.
> 1.5-< 2.5: RR = 1.05.
≥ 2.5: RR = 0.92.
|
No association (No data shown; P = 0.88).
|
Age, energy intake, education level, physical activity level, alcohol consumption, smoking status, pack-years of cigarette smoking, age at first livebirth, estrogen use, vitamin use, BMI, waist/hip ratio, diabetes, hypertension, intake of whole grains, fruits and vegetables, red meat, cholesterol, and saturated fat. |
| 25) Folsom AR (2004) | The Iowa Women's Health Study | 41,836 women aged 55-69, and free of heart disease or cancer. | 1986-2000 | 4,653 | Total mortality | Omega-3 fatty acids from fish (excluding fish oil supplements) |
RR = 0.96 (0.86-1.06; P = 0.79) for the highest vs lowest quintile of consumption.
Amount specific data (g/day):
≤ 0.05: RR = 1.
0.06-0.10: RR = 0.98 (0.89-1.08).
0.11-0.16: RR = 1.01 (0.91-1.10).
0.17-0.26: RR = 1.04 (0.94-1.14).
≥ 0.27: RR = 0.96 (0.86-1.06).
Diabetic women: No association was found among diabetic and nondiabetic women (results not shown). | Age, energy intake, education level, physical activity level, alcohol consumption, smoking status, pack-years of cigarette smoking, age at first livebirth, estrogen use, vitamin use, BMI, waist/hip ratio, diabetes, hypertension, intake of whole grains, fruits and vegetables, red meat, cholesterol, and saturated fat. |
| 24) Trichopoulou A (2009) | The Greek segment of the EPIC Study | 23,349 men and women aged 20-86, and without cancer, CHD, or diabetes. | 8.5
(1994-97 to 2008) | 1,075? | All-cause mortality | Fish and seafood (not defined) |
Mortality Ratio = 1.078 (0.950-1.224; P = 0.243) for ≥ vs < median consumption (23.67 g/day for men, and 18.89 g/day for women). | Sex, age, education, smoking status, waist:hip ratio, BMI, MET score, total energy, and consumption of vegetables, legumes, fruits and nuts, cereals, monounsaturated/saturated lipids ratio, dairy products, meat and meat products, ethanol. |
| 24) Trichopoulou A (2006) | The EPIC Study. | 1,013 subjects taking drugs for diabetes mellitus at enrolment.
(Greece) | 4.5 | 80? (46 cardiovascular, 19 cancer, 15 other) | Mortality | Fish and seafood (not defined) | HR = 1.06 (0.82-1.37; P = 0.64) per increment of 18 g/day. | Gender, age, education, smoking, waist-to-height, hip circumference, MET score, treatment with insulin, treatment for hypertension, treatment for hypercholesterolaemia, vegetables, legumes/potatoes, fruits/nuts, dairy, cereals, meat products, eggs, sugar/confectionery, soft drinks/juices, tea/coffee, ethanol, and olive oil. |
| 24) Trichopoulou A (2005) | The EPIC-elderly Study | 74,607 men and women aged ≥ 60 without a history of CHD, stroke, or cancer, and from 9 European countries.
(Denmark, France, Germany, Greece, Italy, the Netherlands, Spain, Sweden, UK) | Median 89 months.
(1992-2000 to 1999-2003) | 4,047? | Overall mortality | Fish (not defined) | Mortality ratio: 1.00 (0.97-1.04; No P-value) for an increment of 31 g/day. | Stratified by country and adjusted for sex, age, diabetes mellitus at baseline, BMI, educational achievement, smoking status, physical activity at occupation, physical activity at leisure, alcohol intake, and total energy intake. |
| 24) Trichopoulou A (2005) | The Greek part of the EPIC Study | 1,302 men and women aged 20-86 with CHD (MI and/or angina pectoris), and with or without diabetes mellitus.
(Greece) | 3.78
(1994-97 to 2003) | 131? | All-cause mortality | Fish and seafood (not defined) | MR = 0.97 (0.81-1.15 for an increment of 18 g/d. | Sex, age, previous treatment for hypercholesterolemia, diabetes mellitus at enrollment, years of schooling, smoking status, waist-to-hip ratio, metabolic equivalent score, BMI, and total energy |
| 24) Trichopoulou A (2003) | The Greek component of The EPIC Study | 22,043 adults aged 20-86.
(Greece) | 3.7
(1994-99 to 2002) | 275? | All-cause mortality | Fish (not defined) | HR = 1.02 (0.92-1.12; No P-value) for an increment of 15 g/day. | Sex, age, waist-to-hip ratio, energy-expenditure score, education, smoking status and 5 categories of current smoker, BMI, and total energy. |
| 23) Burr ML (2003) | No cohort name defined | 3,114 men from South Wales aged < 70, with angina, without other serious illness, and who did not eat oily fish twice a week.
(Randomized controlled trial) | 3-9
(1990-96 to 1999) | 440 | Total mortality | Experimental group (Fish advice: to eat ≥ 2 portions of oily fish each week).
Control group (Advice to eat 'sensibly': non-specific advice that did not include intervention with oily fish/fish oil, vegetables, fruits, or oats). | HR = 1.13 (0.94-1.37; P = 0.20) for fish advice vs no fish advice. | Age, smoking, previous MI, history of high blood pressure, diabetes, BMI, serum cholesterol, medication, and fruit advice. |
| 23) Burr ML (2003) | No cohort name defined | 3,114 men from South Wales aged < 70, with angina, without other serious illness, and who did not eat oily fish twice a week.
(Randomized controlled trial) | 3-9
(1990-96 to 1999) | 525? | Total mortality | Experimental group (Fish advice: to eat ≥ 2 portions of oily fish each week, or to take up to 3 g of fish oil ['Maxepa']).
Control group (Advice to eat 'sensibly': non-specific advice that did not include intervention with oily fish/fish oil, vegetables, fruits, or oats). | HR = 1.15 (0.96-1.36; P = 0.13) for fish advice vs no fish advice. | Age, smoking, previous MI, history of high blood pressure, diabetes, BMI, serum cholesterol, medication, and fruit advice. |
| 22) Erkkilä AT (2003) | The Finnish cohort of the EUROASPIRE Study | 400 subjects aged 33-74 with CAD admitted to Kuopio University Hospital. | 5 | 34 | Death | Fish (not defined) |
RR = 0.37 (0.14-1.00; P = 0.059) for the highest vs lowest tertile of consumption.
Amount specific data (g/d):
0: RR = 1.
1-57: RR = 0.50 (0.20-1.28).
> 57: RR = 0.37 (0.14-1.00). | Age, sex, diagnostic category, energy intake, serum cholesterol, serum triaglycerol, diabetes, BMI, and education. |
| 21) Osler M (2003) | No cohort name defined | 4,513 men and 3,984 women aged 30-70.
(Denmark) | (1982-92 to 2000) | See variables | All-cause mortality | Fish (not defined) |
All:
(1,329 cases) | Men:
(826 cases) | Women:
(503 cases) | High-risk individuals:
(562 cases) |
HR = 1.06 (0.88-1.28; P = 0.02) for the highest vs third quartile of consumption.
Amount specific data (Frequency of consumption):
≤ once a month: HR = 0.88 (0.76-1.02).
2/month: HR = 0.84 (0.73-0.96).
1/week: HR = 1 (reference group).
≥ 2/week: HR = 1.06 (0.88-1.28).
|
HR = 1.01 (0.80-1.28; P = 0.08) for the highest vs third quartile of consumption.
Amount specific data (Frequency of consumption):
≤ once a month: HR = 0.80 (0.65-0.90).
2/month: HR = 0.87 (0.73-1.03).
1/week: HR = 1 (reference group).
≥ 2/week: HR = 1.01 (0.80-1.28).
|
HR = 1.21 (0.90-1.63; P = 0.06) for the highest vs third quartile of consumption.
Amount specific data (Frequency of consumption):
≤ once a month: HR = 1.03 (0.82-1.30).
2/month: HR = 0.80 (0.64-1.04).
1/week: HR = 1 (reference group).
≥ 2/week: HR = 1.21 (0.90-1.63).
|
HR = 1.23 (0.94-1.61; P = 0.03) for the highest vs third quartile of consumption.
Amount specific data (Frequency of consumption):
≤ once a month: HR = 0.83 (0.65-1.06).
2/month: HR = 0.84 (0.68-1.03).
1/week: HR = 1 (reference group).
≥ 2/week: HR = 1.23 (0.94-1.61).
|
High risk individuals:
(men over 50 years or women over 60 years, who were current smokers with serum cholesterol over 6 mmol/L or current nonsmokers with a serum cholesterol over 7 mmol/L)Smoking status, physical activity, alcohol, educational status, healty diet score, total cholesterol, and BMI. |
| 20) Barzi F (2003) | The GISSI-Preventione Trial. | 11,246 men and women aged 19-90 with recent (≤ 3 months) myocardial infarction from 172 centres in Italy. | 6.5
(1993-95 to ?) | 1,660 | All-cause mortality | Fish (not defined) | The intakes of the foods were assessed at baseline, 6, 18 and 42 months.
OR = 0.76 (0.62-0.94; P = 0.0003) for the highest vs lowest quartile of consumption.
Amount specific data (frequency of consumption):
Never/almost never: OR = 1.
1 per week: OR = 0.87 (0.75-1.02).
2 per week: OR = 0.81 (0.69-0.94).
> 2 per week: OR = 0.76 (0.62-0.94). | Age, sex, hypertension, HDL-cholesterol, diabetes, smoking, claudication, electrical instability, left ventricular dysfunction, residual myocardial ischaemia, dietary supplementation (vitamin E, n-3 PUFA and the interaction), pharmacological therapies (aspirin, beta-blockers, angiotensin converting enzyme inhibitors), and consumption of fruit, raw vegetables, cooked vegetables, and olive oil. |
| 19) Virtanen JK (2008) | The Health Professionals Follow-up Study | 40,230 men aged 40-75 and free of major chronic disease.
(USA) | 18
(1986-2004) | 9,715 | Major chronic disease incidence (total CVD, total cancer, or other nontraumatic death [e.g., pneumonia, kidney or liver disease]) | Fish (canned tuna fish, dark meat fish [such as mackerel, salmon, sardines, bluefish, and swordfish], and other fish) |
RR = 0.96 (0.86-1.08; P = 0.87) for the highest vs lowest quintile of consumption.
Amount specific data (servings):
< 1/mo: RR = 1.
1-3/mo: RR = 0.96 (0.87-1.05).
1/wk: RR = 0.92 (0.85-1.00).
2-4/wk: RR = 0.94 (0.85-1.03).
≥ 5/wk: RR = 0.96 (0.86-1.08).
No significant effect modification by n-6 fatty acid intake was seen (P for interactions > 0.10). | Age, BMI, smoking, physical activity, history of diabetes, hypertension or hypercholesterolemia, first-degree family history of MI before age 60, first degree family history of colon cancer, aspirin use, multivitamin use, glycemic load, and intakes of protein, fiber, trans fat, saturated fat, n-6 fatty acids, alpha-linolenic acid, red meat, total calories, and alcohol. |
| 19) Virtanen JK (2008) | The Health Professionals Follow-up Study | 40,230 men aged 40-75 and free of major chronic disease.
(USA) | 18
(1986-2004) | 9,715 | Major chronic disease incidence (total CVD, total cancer, or other nontraumatic death [e.g., pneumonia, kidney or liver disease]) | EPA + DHA (from the consumption of all seafood) |
RR = 1.04 (0.92-1.18; P = 0.67) for the highest vs lowest quintile of consumption.
Amount specific data (g/d):
< 0.05: RR = 1.
0.05-< 0.2: RR = 1.00 (0.90-1.11).
0.2-< 0.4: RR = 0.96 (0.86-1.06).
0.4-< 0.6: RR = 0.94 (0.84-1.05).
≥ 0.6: RR = 1.04 (0.92-1.18). | Age, BMI, smoking, physical activity, history of diabetes, hypertension or hypercholesterolemia, first-degree family history of MI before age 60, first degree family history of colon cancer, aspirin use, multivitamin use, glycemic load, and intakes of protein, fiber, trans fat, saturated fat, n-6 fatty acids, alpha-linolenic acid, red meat, total calories, and alcohol. |
| 19) Hu FB (2003) | The Nurses' Health Study | 5,103 women with physician-diagnosed type 2 diabetes mellitus. | 1980-96 | 468 | Total mortality | Fish (dark-meat fish [such as mackerel, salmon, sardines, bluefish, or swordfish]; canned tuna; other fish; and shrimp, lobster, or scallops as main dish) |
RR = 0.49 (0.30-0.81; P = 0.006) for the highest vs lowest quintile of consumption.
Amount specific data (Frequency of intake):
< 1/mo: RR = 1.
1-3/mo: RR = 0.75 (0.53-1.07).
1/wk: RR = 0.67 (0.49-0.93).
2-4/wk: RR = 0.69 (0.47-1.02).
5+/wk: RR = 0.49 (0.30-0.81).
Frequency of consumption was defined in portion units of 6-8 oz. | Age, time intervals, smoking status, BMI, alcohol intake, parental history of MI, menopausal status and postmenopausal hormone use, moderate to vigorous activities, usual aspirin use, multivitamin supplement use, vitamin E supplement use, history of hypertension, hypercholesterolemia, duration of diabetes, and hypoglycemic medication. |
| 19) Hu FB (2002) | The Nurses' Health Study | 84,688 women aged 34-59, and free of CVD and cancer. | 16
(1980-1996) | 4,121 | Death from all causes | Fish ((1) dark-meat fish such as mackerel, salmon, sardines, bluefish, or swordfish (84-140 g [3-5 oz]); (2) canned tuna (84-112 g [3-4 oz]); (3) other fish (84-140 g [3-5 oz]); and (4) shrimp, lobster, or scallops as the main dish (98 g [3.5 oz])) | RR = 0.68 (0.57-0.82; P = < 0.001) for consumption ≥ 5 times/week vs < once/month. | Age, time periods, smoking status, BMI, alcohol intake, menopausal status, vigorous to moderate activity, aspirin, multivitamin use, vitamin E supplement use, and history of hypertension, hypercholesterolemia, and diabetes. |
| 18) Nagata C (2002) | The Takayama Study | 13,355 men and 15,724 women aged 35 or older.
(Japan) | 7
(1992-1999) | 1,163 men, and
899 women | All-cause mortality | Fish (16 items, not defined), and
fish oil |
Fish
| Men | Women |
HR = 0.94 (0.78-1.12; P = 0.50) for the highest vs lowest quintile of consumption.
Amount specific data (median intake in g/day):
46.2: HR = 1.
68.1: HR = 0.92 (0.76-1.11).
86.8: HR = 0.91 (0.75-1.10).
111.9: HR = 0.90 (0.75-1.09).
157.8: HR = 0.94 (0.78-1.12). |
HR = 0.86 (0.70-1.05; P = 0.17) for the highest vs lowest quintile of consumption.
Amount specific data (median intake in g/day):
36.6: HR = 1.
53.9: HR = 0.93 (0.76-1.14).
68.8: HR = 0.96 (0.79-1.18).
88.1: HR = 0.93 (0.76-1.14).
122.4: HR = 0.86 (0.70-1.05). |
Fish oil
| Men | Women |
HR = 0.87 (0.73-1.05; P = 0.38) for the highest vs lowest quintile of consumption.
Amount specific data (median intake in mg/day):
410: HR = 1.
602: HR = 0.82 (0.67-0.99).
788: HR = 0.87 (0.72-1.05).
1,051: HR = 0.88 (0.73-1.06).
1,582: HR = 0.87 (0.73-1.05). |
HR = 0.77 (0.62-0.94; P = 0.01) for the highest vs lowest quintile of consumption.
Amount specific data (median intake in mg/day):
332: HR = 1.
486: HR = 0.92 (0.76-1.13).
635: HR = 0.84 (0.69-1.04).
832: HR = 0.90 (0.73-1.09).
1,253: HR = 0.77 (0.62-0.94).
This association did not change after additional adjustment for dietary factors significantly associated with all-cause mortality (rice/grains/potatoes and vegetables): HR = 0.73 (0.59-0.90). |
Age, total energy, marital status, BMI, smoking status, alcohol intake, exercise, and history of hypertension and diabetes mellitus.
Additionally for men: coffee intake.
Additionally for women: education, age at menarche, and menopausal status. |
| 17) Yuan JM (2001) | The Shanghai Cohort Study | 18,244 men aged 45-64 from Shanghai (China), and without a history of cancer. | 12
(1986-89 to 1998) | 2,134 | Total mortality | Seafood (fresh fish [e.g., carp, beam, pomfret]; salted fish [e.g., yellow croaker and hairtail]; and shellfish [e.g., shrimp and crab]) |
RR = 0.79 (0.69-0.91; P = 0.01) for the highest vs lowest quintile of consumption.
Amount specific data (g/week):
< 50: RR = 1.
50-< 100: RR = 0.79 (0.70-0.89).
100-< 150: RR = 0.76 (0.66-0.87).
150-< 200: RR = 0.86 (0.75-0.99).
≥ 200: RR = 0.79 (0.69-0.91).
One serving = 50 g. | Age, total energy intake, level of education, BMI, current smoking, cigarettes smoked, alcohol, history of diabetes, and history of hypertension. |
| 16) Fortes C (2000) | Cohort name not defined | 161 elderly (52 men, and 109 women) ≥ 65 years. (Mean age 80) in Rome.
(Italy) | 5
(1993-1998) | 53? (21 men, and 32 women) | Overall mortality | Fish in general (not defined) | RR = 0.89 (0.46-1.74; No P-value) for consumption ≥ 1 vs < 1 time/week. | Unadjusted. |
| 15) Gillum RF (2000) | The NHANES I Study | 8,825 white and black men and women aged 24-74.
(USA) | 18.8
(1971-75 to 1992) | See variables | Death from all causes | Fish or shellfish (not defined) |
White men:
(1,236 cases) | Black men:
(277 cases) | White women:
(1,103 cases) | Black women:
(285 cases) |
RR = 0.85 (0.68-1.06) for the highest vs lowest quartile of consumption.
Amount specific data (Frequency of consumption):
Never: RR = 1.
< 1/week: RR = 0.88 (0.73-1.05).
1/week: RR = 0.76 (0.63-0.91; P = ≤ 0.01).
> 1/week: RR = 0.85 (0.68-1.06).
|
RR = 1.11 (0.68-1.81) for the highest vs lowest quartile of consumption.
Amount specific data (Frequency of consumption):
Never: RR = 1.
< 1/week: RR = 1.01 (0.65-1.56).
1/week: RR = 1.05 (0.68-1.63).
> 1/week: RR = 1.11 (0.68-1.81).
|
RR = 0.90 (0.71-1.15) for the highest vs lowest quartile of consumption.
Amount specific data (Frequency of consumption):
Never: RR = 1.
< 1/week: RR = 1.02 (0.84-1.23).
1/week: RR = 1.02 (0.84-1.25).
> 1/week: RR = 0.90 (0.71-1.15).
|
RR = 0.82 (0.52-1.28) for the highest vs lowest quartile of consumption.
Amount specific data (Frequency of consumption):
Never: RR = 1.
< 1/week: RR = 0.77 (0.51-1.16).
1/week: RR = 0.79 (0.53-1.19).
> 1/week: RR = 0.82 (0.52-1.28).
|
No significant interactions of age with fish consumption were seen.Age, smoking, history of diabetes, education, high school graduate, systolic blood pressure, serum cholesterol concentration, BMI, alcohol intake, and physical activity. |
| 14) Jamrozik K (2000) | The Perth Community Stroke Study | 790 controls from a previous case-control analysis with or without a history of CVD.
(Australia) | 4
(1990-91 to 1994) | 198? | Death from any cause | Fish (not defined) | No significant relationship was found (no data shown). | Sex and age. |
| 13) Whiteman D. (1999) | The OXCHECK Study | 10,522 men and women aged 35-64 without a previous history of angina.
(UK) | 9
(1989-1997) | 477 | All-cause mortality | Fresh or frozen fish (not defined) |
RR = 1.06 (0.71-1.59; No P-value) for the highest vs lowest tertile of consumption.
Amount specific data (days/wk):
< 1: RR = 1.
1-3: RR = 1.16 (0.96-1.40).
4-7: RR = 1.06 (0.71-1.59). | Gender, smoking and age. |
| 12) Albert CM (1998) | The Physician's Health Study | 20,551 men aged 40-84, and without a history of MI, stroke, transient ischemic attack, or cancer.
(USA) | 11
(1983-1995) | 1,652 | Total mortality | Fish or shellfish (canned tuna fish; dark meat fish [eg, mackerel, salmon, sardines, bluefish, or swordfish]; other fish; and shrimp, lobster, or scallops as a main dish) |
| Over 5 levels of consumption: | Over 3 levels of consumption: |
RR = 0.73 (0.55-0.96; P = 0.45) for the highest vs lowest quintile of consumption.
Amount specific data (servings):
< 1/mo: RR = 1.
1-3/mo: RR = 0.79 (0.59-1.06).
1-< 2/wk: RR = 0.71 (0.55-0.91).
2-< 5/wk: RR = 0.70 (0.54-0.89).
≥ 5/wk: RR = 0.73 (0.55-0.96).
|
RR = 0.70 (0.55-0.89; P = 0.003) for the highest vs lowest tertile of consumption.
Amount specific data (servings):
< 1/mo: RR = 1.
1-3/mo: RR = 0.79 (0.59-1.06).
≥ 1/wk: RR = 0.70 (0.55-0.89).
|
Age, aspirin and beta carotene treatment assignment, evidence of CVD, BMI, smoking status, history of diabetes, history of hypertension, history of hypercholesterolemia, alcohol, vigorous exercise, vitamine E, vitamin C, and multivitamin use. |
| 11) Mann JI (1997) | The Oxford Vegetarian Study | 10,802 subjects (4102 men and 6700 women) aged 16-79.
(UK) | 13.3
(1980-84 to 1995) | 389 | All-cause mortality | Fish (not defined) |
Death rate ratio = 96 (76-121; P = Not Significant) for the highest vs lowest tertile of consumption.
Amount specific data (per week):
Never eaten: DRR = 100.
< once: DRR = 97 (74-126).
≥ once: DRR = 96 (76-121). | Age, sex, smoking and social class. |
| 10) Salonen JT (1995) | The KIHD Study | 1,833 men aged 42-60.
(Finland) | 1984-89 to 1991 | 78? | Death from all causes | Fish (not defined) |
| Fish intake in g/d: | Fish intake ≥ 30 vs < 30 g/d: |
| RR = 1.002 (0.999-1.006; P = 0.141)
|
RR = 1.10 (0.69-1.74; P = 0.694)
|
The most commonly used fish species were vendance (Coregonus albula, a small local white fish) (17%), rainbow trout (12%), and the Northern pike (10%).Age, examination year, ischemic exercise, ECG, maximal oxygen uptake, family history of CHD, cigarette-years, mean systolic blood pressure, diabetes, socioeconomic status, place of residence, dietary iron intake, and serum apolipoprotein B, HDL-cholesterol, and ferritin concentrations.
Additional adjustment for intakes of energy, saturated fatty acids, dietary polyunsaturated fatty acids, and carotene; the consumption of alcohol and coffee; leisure time physical activity; BMIK; blood leukocyte count; plasma ascorbate and alpha-tocopherol, serum copper, albumin, and triglyceride concentrations had no appreciable effect on the RRs. |
| 9) Kromhout D (1995) | Cohort name not defined | 272 subjects (137 men and 135 women) aged 64-87 from a general practice in Rotterdam, and without cancer.
(The Netherlands) | 17
(1971-1987) | 187 | Total mortality | Fish (mostly lean fish [e.g., cod and plaice], but also minimal amounts of fatty fish [e.g., herring and mackerel], and canned fish [e.g., sardines]) | RR = 0.96 (0.72-1.30) for fish intake vs no fish intake. | Age, gender, prevalence of myocardial infarction and angina pectoris, systolic blood pressure, smoking, and alcohol. |
| 8) Dolecek TA (1992) | The Multiple Risk Factor Intervention Trial (MRFIT) | 6,250 men aged 35-57, and at a high risk of CHD.
(USA) | 10.5
(?-1985) | 522? | Death from all causes | Dietary n-3 fish fatty acids | This was a randomized trial, but only data from the "usual care" group was analyzed.
| Grams | Percentage of total kilocalories |
RR = 0.75 (P = < 0.10) for the highest vs lowest quintile of consumption.
Amount specific data (grams):
0.000: RR = 1.
0.009: RR = 1.09
0.046: RR = 1.02
0.153: RR = 0.85
0.664: RR = 0.75
|
RR = 0.73 (P = < 0.01) for the highest vs lowest quintile of consumption.
Amount specific data (% of calories):
0.000: RR = 1.
0.004: RR = 1.09
0.019: RR = 0.97
0.063: RR = 0.92
0.284: RR = 0.73
|
Age, race, smoking, baseline diastolic blood pressure, HDL, LDL, and alcohol. |
| 8) Dolecek TA (1991) | The Multiple Risk Factor Intervention Trial (MRFIT) | 6,258 men aged 35-57, and at a high risk of developing CHD based upon smoking status, diastolic blood pressure, and serum cholesterol levels.
(USA) | ?-1985 | 439 | All cause mortality | Dietary long-chain omega 3 fatty acids from fish [EPA (20:5), DPA (22:5), and DHA (22:6)] | This was a randomized trial, but only data from the "usual care" group was analyzed.
RR = 0.76 (No 95% CI; P = 0.01) for the highest vs lowest quintile of consumption.
Amount specific data (g):
0.000: RR = 1
0.009: RR = 1.09
0.046: RR = 1.02
0.153: RR = 0.85
0.664: RR = 0.76 | Age, race, baseline smoking, diastolic blood pressure, HDL and LDL levels. |
| 7) Hirayama T (1990) | No cohort name defined | 265,118 adults (122,261 men, 142,857 women) aged ≥ 40 from 6 prefectures in Japan. | 17
(1966-1982) | 55.523? (31.979 men, 23.544 women) | All cause mortality | Fish & shellfish (not defined) |
| Men & women | Men | Women |
| RR = 0.93 (90% CI = 0.92-0.95; No P-value) for consumption ≥ 4 vs ≤ 3 times/wk.
|
RR = 1.35 (90% CI = 1.20-1.52; No P-value). for the lowest vs highest quartile of consumption.
Amount specific data (Frequency of consumption):
≥ 4 times/wk: RR = 1.
1-3 times/wk: RR = 1.05 (1.03-1.07).
1-3 times/mo: RR = 1.09 (1.04-1.14).
None: RR = 1.35 (1.20-1.52). |
RR = 1.25 (90% CI = 1.13-1.39; No P-value) for the lowest vs highest quartile of consumption.
Amount specific data (Frequency of consumption):
≥ 4 times/wk: RR = 1.
1-3 times/wk: RR = 1.10 (1.07-1.12).
1-3 times/mo: RR = 1.16 (1.11-1.22).
None: RR = 1.26 (1.13-1.39). |
Not defined (possibly only age). |
| 6) Ness AR (2002) | The Diet and Reinfarction Trial (DART) | 2,033 men aged < 70 who had survived a MI, and were rectruited from hospitals in England.
(Randomized trial) | 1983-87 to 2000 | 1,083 | All-cause mortality | Experimental group: (Men were encouraged to eat 2 portions of fatty fish a week and as much other fish as they could manage. Those unable to eat this amount of fish were offered fish oil capsules [MaxEPA]).
Control group: (No fish advice) |
Dietary differences:
-At 2 y reported fatty fish intake was 35 g per day in the fish advice group and around 9 g per day in those not given fish advice.
-At the end of the follow-up period in 1999-2000 those allocated to fish advice reported eating more fish (43.6 vs 36.9 g/day; P = < 0.01) and in particular more fatty fish (20.7 vs 13.2 g/day; P = < 0.01) although the differences were substantially reduced compared to those reported at 2 y. In addition, they were more likely to take fish oil supplements (120 [26.9%] vs 83 [19.3%; P = < 0.01]).
Results:
HR = 0.94 (0.84-1.06).
Hazard by follow-up period in years:
0-2: HR = 0.70 (0.54-0.92).
2-5: HR = 1.31 (1.01-1.70).
5-10: HR = 1.02 (0.82-1.27).
10+: HR = 0.84 (0.68-1.06). | Unadjusted.
Adjustment for history of MI, angina, hypertension at baseline; X-ray evidence of cardiomegaly, pulmonary congestion or pulmonary oedema at baseline; and treatment (at entry) with b-blockers, other anti-hypertensives, digoxin/anti-arrhythmics, or anticoagulants did not change the HRs. |
| 6) Burr ML (1989) | The Diet And Reinfarction Trial (DART) | 2,033 men aged < 70, admitted to 21 hospitals, and who had recovered from acute MI (excluding diabetic patients).
(Randomised controlled trial). | 2 | 224 | Death | Experimental group: fish advice (≥ 2 weekly portions [200-400 g] of fatty fish [mackerel, herring, kipper, pilchard, sardine, salmon, or trout]).
Control group: No fish advice. |
RR = 0.71 (0.54-0.92; P = < 0.05) for fish advice vs no fish advice.
This RR remained 0.71 (0.54-0.93; P = < 0.05) after adjustment for history of MI, angina, or hypertension; X-ray evidence of cardiomegaly, pulmonary congestion, or pulmonary oedema; and treatment (at entry) with beta-blockers, other antihypertensives, digoxin/antiarrhytmics, or anticoagulants.
The effect appeared early in the trial.
% of deaths according to fish advice, within other dietary groupings:
| Fish advice group | No fish advice group |
Fat, fibre: 8.0%
Fat, no fibre: 8.9%
No fat, fibre: 12.4%
No fat, no fibre: 7.8%
|
Fat, fibre: 13.5%
Fat, no fibre: 13.2%
No fat, fibre: 14.5%
No fat, no fibre: 9.9%
|
-Men were allocated to receive or not to receive advice on each of 3 dietary factors: (1) fat advice, (2) fish advice, and (3) fibre advice, thus creating 8 possible combinations.
-At 6 months and 2 years, the subjects was administered a detailed dietary questionnaire.
-In the fish advice group 14% took maxepa capsules at 6 months, and 22% at 2 years as a partial substitute for fatty fish.
-The amount of fish in the study was small, supplying 2.5 g EPA weekly (corresponding to about 300 g fatty fish every week), and thus involved no radical change in diet.
-During the first 6 months a small but significant (p = < 0.01) net increase of 2.1% in total cholesterol was attributable to fish advice; however, over the 2 years, the net change attributable to fish advice was not significant.
-On average the weights of the subjects did not change much during the trial: the overall mean weight rose by 0-6 kg.Unadjusted. |
| 5) Nube M (1987) | No cohort name defined | 2,820 Dutch civil servants of Amsterdam and their spouses aged 40-64 (1,394 men, and 1,426). | 25
(1953-54 to ?) | 742? men, and
448? women | Survival | Fish (not defined) |
Survival rates (%). Times/month:
| Men | Women |
A positive correlation for the highest vs lowest tertile of consumption.
Amount specific data:
0: 45.2 +/- 2.2.
1-2: 46.4 +/- 2.7.
4-16: 47.9 +/- 1.7.
|
Amount specific data:
0: 72.9 +/- 1.9.
1-3: 65.1 +/- 2.5.
4-28: 66.8 +/- 1.6.
|
Age. |
| 4) Lapidus L (1986) | Cohort name not defined | 1,462 women in Gothenburg.
(Sweden) | 12
(1968-69 to 1980-81) | 75? | All-cause mortality | Fish (not defined) | No significant correlation was found (no data shown) | Age. |
| 3) Vollset SE (1985) | No cohort name defined | 11,000 men.
(Norway) | 14
(1968-1981) | 2,587 | Total mortality | Fish (not defined) |
Numbers of Observed and Expected deaths for increasing sextiles of consumption (P = 0.44).
Amount specific data (Frequency of consumption per month):
0-4: Observed deaths = 156. Expected deaths = 151.1
5-9: Observed deaths = 565. Expected deaths = 527.2
10-14: Observed deaths = 989. Expected deaths = 1037.6
15-19: Observed deaths = 412. Expected deaths = 405.8
20-24: Observed deaths = 352. Expected deaths = 352.0
≥ 25: Observed deaths = 113. Expected deaths = 113.2 | Stratified by age, urban/rural status, region, and smoking status. |
| 2) Daviglus ML (1997) | The Chicago Western Electric Study | 1,822 men aged 40-55, and free of cardiovascular disease. | 30
(1957-59 to ?) | 1,042 | Death from all causes | Fish (not defined) |
RR = 0.85 (0.64-1.10; P = 0.175) for the highest vs lowest quartile of consumption.
Amount specific data (g/day):
0: RR = 1.
1-17: RR = 1.02 (0.82-1.27).
18-34: RR = 0.98 (0.79-1.22).
≥ 35: RR = 0.85 (0.64-1.10). | Age, education, religion, systolic pressure, serum cholesterol, number of cigarettes smoked, BMI, diabetes, electrocardiographig abnormalities, and intake of energy, cholesterol, saturated, monounsaturated, and polyunsaturated fatty acids, total protein, cabohydrate, alcohol, iron, thiamine, riboflavin, niacin, vitamin C, beta carotene, and retinol. |
| 2) Shekelle RB (1985) | The Western Electric Study | 1,931 middle-aged men.
(Chicago) | 25
(1957-?) | 707 | Death from all causes | Fish (not defined) |
A nonsignificant inverse association was found for the highest vs lowest quartile of consumption (P = 0.051).
Amount specific data (g/day):
0: % of deaths = 41.0
1-17: % of deaths = 38.5
18-34: % of deaths = 34.4
> 35: % of deaths = 34.9.
This association remained after adjustment for age, blood pressure, serum cholesterol, cigarette smoking, BMI, evidence of existing disease (e.g., diabetes mellitus, hypertensive retinopathy, or electrocardiographic abnormalities), intake of dietary cholesterol, % of calories from saturated, monounsaturated fatty acids, and polyunsaturated fatty acids, education, intake of energy/kg weight, and alcohol (P = 0.014). | Unadjusted. |
| 1) Fraser GE (1997) | The Adventist Health Study | Non-Hispanic white Seventh Seventh Day Adventists aged ≥ 84.
At study baseline there were 603 subjects older than 84 years, and during a 12-year follow-up, these and intitially younger subjects contributed 11,828 person-years of observation in the oldest-old age range.
(USA) | 12
(1976-88) | 1,387?
(451 men, and 936 women) | All-cause mortality | Fish (not defined) |
| Total: | Men: | Women: |
| HR = 0.98 (0.76-1.24) for consumption ≥ 1 vs < 1 per week.
|
HR = 0.89 (0.58-1.38) for consumption ≥ 1 vs < 1 per week.
|
HR = 0.99 (0.73-1.33) for consumption ≥ 1 vs < 1 per week.
|
Sex, diabetes, smoking, exercise, nuts, fruit, bread, donuts, sweet desserts, and beef. |
| 1) Kahn HA (1984) | The Adventist Health Study | 27,530 California members of the Seventh-Day Adventist Church aged ≥ 30, including black, white and oriental subjects.
(USA) | 21
(1960-1980) | 5,971 | All-cause mortality | Fish (not defined) |
| OR with consideration of lenght of survival | OR without consideration of lenght of survival |
OR = 1.12 (No 99% CI; No P-value) for the highest vs lowest tertile of consumption.
Amount specific data (days/wk):
None: OR = 1.00.
Occasionally: OR = 1.08.
> or = 1: OR = 1.12. |
OR = 1.14 (99% CI = 1.00-1.28; P = < 0.01) for the highest vs lowest tertile of consumption.
Amount specific data (days/wk):
None: OR = 1.00.
Occasionally: OR = 1.12.
> or = 1: OR = 1.14 (99% CI = 1.00-1.28; P = < 0.01). |
After additional adjustment for meat and eggs the association with fish became nonsignificant (OR = 1.04).
Age (15-year intervals), sex, history of disease (heart disease, stroke, hypertension, diabetes, or cancer), age at initial exposure to the Adventist Church, and smoking history (never vs ever). |
| | | | | | | | | | | | | | | |