Total fish consumption and coronary heart disease risk (CHD). Effect modification.

  • Level of consumption: Effects probably differ between different levels of intake. For this analysis, four different levels of consumption were considered, relative to nonconsumption: < 1, 1, 2, and ≥ 3 servings/week.
    For CHD risk, data from 24 cohorts with 13,451 cases could be included (Table 1). Data from Hirayama T [10] was not included for reasons mentioned in the methods. And data from Gillum RF [17] could not be included because no amounts of cases were defined for the specified cohorts. Average RR's were identical for intakes < 1, and 1 serving/week (RR = 0.95). And RR's were similar for intakes 2, and ≥ 3 servings/week (RR = 0.89, and 0.90, respectively).
    For CHD mortality, RRs were similar to the RR's for CHD for the highest levels of intake. RR = 1.00, 0.93, 0.90, and 0.90 for < 1, 1, 2, and ≥ 3 servings/wk, respectively.
    For nonfatal MI, data from 7 cohorts, with 2,787 cases could be included (Table 3). RR's were similar for lower intakes (RR = 0.88, and 0.87 for < 1, and 1 serving/wk, respectively), and RR's were similar for higher intakes (RR = 0.80, and 0.78 for 2, and ≥ 3 servings/wk, respectively).
  • Gender: For CHD risk, effects did not consistently seem to differ between men and women (Table 4). Average RR's were 0.86, and 0.90 for men and women, respectively. These became 0.90 and 0.84 after excluding the debatable results from Hirayama T [10]). For men, the effect size attenuated even further after including data about coronary-artery bypass grafting (Ascherio A [14]) in the definition for CHD risk: RR = 0.95.
    Other authors did not describe data stratified by gender, but stated that effects were similar for men and women (Norell SE [5], Mozaffarian D [24], de Goede J [34].
  • Geographic area: For CHD risk, protective effects were stronger among Asian cohorts (8 cohorts: RR = 0.85), and US cohorts (11 cohorts: RR = 0.86), than among European cohorts (17 cohorts: RR = 0.95) (Tables 5-7).
  • Serum cholesterol: In one cohort, risk seemed to increase (RR = 2.5) among subjects with hypercholesterolemia, while it seemed to decrease (RR = 0.7) among subjects without hypercholesterolemia (Morris MC [15]). In another cohort, fish consumption seemed to increase risk among subjects with serum cholesterol levels > 7 mmol/L, while consumption seemed to decrease risk among subjects with serum cholesterol levels < 5 mmol/L (Osler M [26]).
    Authors from a third cohort stated that no significant inverse association was seen in any of the subgroups defined according to risk factors for CHD. These factors included a history of hypercholesterolemia. Subjects in all 3 cohorts mentioned, included men only.
    Results from a fourth cohort, including men and women, showed that fish consumption decreased risk, regardless of serum cholesterol levels (Gartside PS [17]).
  • Other possible effect modifiers: No consistent effect modification was found by:
    -age (Streppel MT [1], Daviglus ML [2], Vollset SE [3], Ascherio A [14], Morris MC [15], Gillum RF [17], Mozaffarian D [24], de Goede J [34]),
    -smoking habits (Rodriguez BL [4], Feskens EJ [12], Ascherio A [14], Morris MC [15], Gartside PS [17], Mozaffarian D [24]),
    -blood presssure (Feskens EJ [12], Ascherio A [14], Morris MC [15]),
    -aspirin use (Hu FB 2002 [14], Morris MC [15]), or
    -educational level (Gartside PS [17], Osler M [26]).
  • No effect modification was found by:
    -alcohol consumption (Ascherio A [14], Morris MC [15], Gartside PS [17]),
    -intakes of n-6 polyunsaturated fat (Mozaffarian D [14]), fish oil capsules (Bjerregaard LJ [34]), bread, cheese and dessert (Gartside PS [17]).
    -BMI (Ascherio A [14], Gartside PS [17], Mozaffarian D [24]),
    -race (Gillum RF [17]), family history of MI, profession (Ascherio A [14]), geographic region, or physical activity (Gartside PS [17]).

Table 1.
Prospective studies of total fish and coronary heart disease risk. Statified by intake in servings fish/week:
AuthorCohort nameCases< 1 serving/week1 serving/week2 servings/week≥ 3 servings/week
46) Manger MS (2010)The WENBIT Study210HR = 1HR = 1HR = 1HR = 1.00
45) Tomasallo C (2010)No cohort name defined34 captains, and

26 referents
Captains: HR = 1.48

Referents: HR = 0.74
Captains: HR = 1.81

Referents: HR = 0.31
Captains: HR = 1.81

Referents: HR = 0.31
Captains: HR = 1.81

Referents: HR = 0.31
38) Kaushik S (2008)The Blue Mountains Eye Study184HR = 1HR = 0.88HR = 0.91HR = 0.91
36) Yamagishi K (2008)The JACC Study419 IHD,

107 cardiac arrest, and

307 heart failure
IHD: HR = 1

Cardiac arrest: HR = 1

Heart failure: HR = 1
IHD: HR = 1

Cardiac arrest: HR = 1

Heart failure: HR = 1
IHD: HR = 1.00

Cardiac arrest: HR = 1.22

Heart failure: HR = 0.92
IHD: HR = 0.98

Cardiac arrest: HR = 1.04

Heart failure: HR = 0.70
34) de Goede J (2010)The Dutch part of the EPIC Study82 CHD mortality, and

252 nonfatal MI
CHD mortality: HR = 0.81

Nonfatal MI: HR = 1.01
CHD mortality: HR = 0.52

Nonfatal MI: HR = 1.01
CHD mortality: HR = 0.52

Nonfatal MI: HR = 1.01
CHD mortality: HR = 0.52

Nonfatal MI: HR = 1.01
34) Bjerregaard LJ (2010)The Danish part of the EPIC Study854 men, and

268 women
Men: HR = 1

Women: HR = 1
Men: HR = 1

Women: HR = 1
Men: HR = 0.96

Women: HR = 1.06
Men: HR = 0.94

Women: HR = 1.05
34) Buckland G (2009)The Spanish part of the EPIC Study480 men, and

126 women
Men: HR = 1

Women: HR = 1
Men: HR = 1

Women: HR = 1
Men: HR = 1

Women: HR = 1
Men: HR = 0.84

Women: HR = 0.95
33) Iso H (2006)The JPHC Study258HR = 1HR = 1HR = 0.86HR = 0.82
32) Nakamura Y (2005)The NIPPON DATA80124RR = 1*RR = 0.69RR = 0.69RR = 0.66
31) Ness AR (2005)The Boyd Orr Cohort298RR = 1RR = 0.99RR = 1.18RR = 1.18
30) Chang-Claude J (2005)The German Vegetarian Study60RR = 1.61RR = 2.11RR = 2.11RR = 2.11
29) Folsom AR (2004)The Iowa Women's Health Study922RR = 1.06RR = 0.86RR = 0.75RR = 1.04
28) Erkkilä AT (2003)The EUROASPIRE Study34RR = 1.00RR = 1.00RR = 1.00RR = 0.75
26) Osler M (2003)No cohort name defined491HR = 0.92*HR = 0.98HR = 0.91HR = 0.91
24) Mozaffarian D (2003)The Cardiovascular Health Study247 IHD death, and

363 nonfatal MI
IHD death: HR = 0.89

Nonfatal MI: HR = 0.91
IHD death: HR = 0.77

Nonfatal MI: HR = 0.71
IHD death: HR = 0.53

Nonfatal MI: HR = 0.75
IHD death: HR = 0.47

Nonfatal MI: HR = 0.67
22) Wennberg M (2011)The NSHDS263OR = 0.88OR = 1.09OR = 1.09OR = 1.21
21) Whiteman D (1999)The OXCHECK Study93RR = 1RR = 1.38RR = 1.38RR = 1.37
20) Mann JI (1997)The Oxford Vegetarian Study64DRR = 121DRR = 123DRR = 123DRR = 123
19) Yuan JM (2001)The Shanghai Cohort Study113 MI, and

74 other IHD
MI: RR = 1

Other IHD: RR = 1
MI: RR = 1

Other IHD: RR = 1
MI: RR = 0.54

Other IHD: RR = 1.08
MI: RR = 0.56

Other IHD: RR = 0.84
15) Albert CM (1998)The Physician's Health Study737RR = 0.96RR = 0.99RR = 1.03RR = 1.02
14) Bernstein AM (2010)The Nurses' Health Study3,162RR = 0.88RR = 0.94RR = 0.79RR = 0.81
14) Ascherio A (1995)The Health Professionals Follow-up Study811RR = 0.66RR = 0.82RR = 0.69RR = 0.75
13) Järvinen R (2006)The Finnish Mobile Clinic Health Survey335 men, and

163 women
Men: RR = 1.04

Women: RR = 0.96
Men: RR = 1.07

Women: RR = 0.91
Men: RR = 1.09

Women: RR = 0.68
Men: RR = 0.99

Women: RR = 0.59
8) Fraser GE (1992)The Adventist Health Study134 nonfatal MI, and

463 CHD death
Nonfatal MI: RR = 1.11

CHD death: RR = 1.10
Nonfatal MI: RR = 1.04

CHD death: RR = 1.09
Nonfatal MI: RR = 1.04

CHD death: RR = 1.09
Nonfatal MI: RR = 1.04

CHD death: RR = 1.09
2) Daviglus ML (1997)The Western Electric Study430RR = 0.88RR = 0.88RR = 0.84RR = 0.62
1) Oomen CM (2000)The Seven Countries Study463RR = 0.93RR = 0.93RR = 0.95RR = 1.08
Total number of cases: 13.451Average RR = 0.95Average RR = 0.95Average RR = 0.89Average RR = 0.90

* = Lowest level of consumption was not the reference goup. RR for the lowest level of consumption was set to 1. Other RR's were adjusted for this difference in setting.

Table 2.
Prospective studies of total fish and coronary heart disease mortality. Statified by intake in servings fish/week:
AuthorCohort nameCases< 1 serving/week1 serving/week2 servings/week≥ 3 servings/week
46) Manger MS (2010)The WENBIT Study76HR = 1HR = 1HR = 1HR = 0.91
45) Tomasallo C (2010)No cohort name defined34 captains, and

26 referents
Captains: HR = 1.48

Referents: HR = 0.74
Captains: HR = 1.81

Referents: HR = 0.31
Captains: HR = 1.81

Referents: HR = 0.31
Captains: HR = 1.81

Referents: HR = 0.31
38) Kaushik S (2008)The Blue Mountains Eye Study184HR = 1HR = 0.88HR = 0.91HR = 0.91
36) Yamagishi K (2008)The JACC Study419 IHD,

107 cardiac arrest, and

307 heart failure
IHD: HR = 1

Cardiac arrest: HR = 1

Heart failure: HR = 1
IHD: HR = 1

Cardiac arrest: HR = 1

Heart failure: HR = 1
IHD: HR = 1.00

Cardiac arrest: HR = 1.22

Heart failure: HR = 0.92
IHD: HR = 0.98

Cardiac arrest: HR = 1.04

Heart failure: HR = 0.70
34) de Goede J (2010)The Dutch part of the EPIC Study82HR = 0.81HR = 0.52HR = 0.52HR = 0.52
33) Iso H (2006)The JPHC Study62HR = 1HR = 1HR = 0.70HR = 0.98
32) Nakamura Y (2005)The NIPPON DATA80124RR = 1*RR = 0.69RR = 0.69RR = 0.66
31) Ness AR (2005)The Boyd Orr Cohort298RR = 1RR = 0.99RR = 1.18RR = 1.18
30) Chang-Claude J (2005)The German Vegetarian Study60RR = 1.61RR = 2.11RR = 2.11RR = 2.11
29) Folsom AR (2004)The Iowa Women's Health Study922RR = 1.06RR = 0.86RR = 0.75RR = 1.04
28) Erkkilä AT (2003)The EUROASPIRE Study16RR = 1.59RR = 1.59RR = 1.59RR = 1.32
26) Osler M (2003)No cohort name defined247HR = 0.95*HR = 0.92HR = 0.90HR = 0.90
24) Mozaffarian D (2003)The Cardiovascular Health Study247HR = 0.89HR = 0.77HR = 0.53HR = 0.47
21) Whiteman D (1999)The OXCHECK Study93RR = 1RR = 1.38RR = 1.38RR = 1.37
20) Mann JI (1997)The Oxford Vegetarian Study64DRR = 121DRR = 123DRR = 123DRR = 123
19) Yuan JM (2001)The Shanghai Cohort Study113 MI, and

74 other IHD
MI: RR = 1

Other IHD: RR = 1
MI: RR = 1

Other IHD: RR = 1
MI: RR = 0.54

Other IHD: RR = 1.08
MI: RR = 0.56

Other IHD: RR = 0.84
15) Albert CM (1998)The Physician's Health Study308RR = 1.09RR = 0.82RR = 0.91RR = 0.86
14) Hu FB (2002)The Nurses' Health Study484RR = 0.91RR = 0.66RR = 0.73RR = 0.64
14) Ascherio A (1995)The Health Professionals Follow-up Study264RR = 0.87RR = 0.86RR = 0.71RR = 0.67
13) Järvinen R (2006)The Finnish Mobile Clinic Health Survey335 men, and

163 women
Men: RR = 1.04

Women: RR = 0.96
Men: RR = 1.07

Women: RR = 0.91
Men: RR = 1.09

Women: RR = 0.68
Men: RR = 0.99

Women: RR = 0.59
8) Fraser GE (1992)The Adventist Health Study463RR = 1.1RR = 1.09RR = 1.09RR = 1.09
2) Daviglus ML (1997)The Western Electric Study430RR = 0.88RR = 0.88RR = 0.84RR = 0.62
1) Oomen CM (2000)The Seven Countries Study463RR = 0.93RR = 0.93RR = 0.95RR = 1.08
Total number of cases: 6,465Average RR = 1.00Average RR = 0.93Average RR = 0.90Average RR = 0.90

* = Lowest level of consumption was not the reference goup. RR for the lowest level of consumption was set to 1. Other RR's were adjusted for this difference in setting.

Table 3.
Prospective studies of total fish and nonfatal MI. Statified by intake in servings fish/week:
AuthorCohort nameCases< 1 serving/week1 serving/week2 servings/week≥ 3 servings/week
34) de Goede J (2010)The Dutch part of The EPIC Study252HR = 1.01HR = 1.01HR = 1.01HR = 1.01
33) Iso H (2006)The JPHC Study196HR = 1HR = 1HR = 0.89HR = 0.67
24) Mozaffarian D (2003)The Cardiovascular Health Study363HR = 0.91HR = 0.71HR = 0.75HR = 0.67
15) Morris MC (1995)The Physician's Health Study259RR = 1RR = 1.4RR = 1.2RR = 1
14) Hu FB (2002)The Nurses' Health Study1,029RR = 0.89RR = 0.74RR = 0.68RR = 0.71
14) Ascherio A (1995)The Health Professionals Follow-up Study554RR = 0.62RR = 0.8RR = 0.67RR = 0.77
8) Fraser GE (1992)The Adventist Health Study134RR = 1.11RR = 1.04RR = 1.04RR = 1.04
Total number of cases: 2,787Average RR = 0.88Average RR = 0.87Average RR = 0.80Average RR = 0.78

Table 4.
Prospective studies of total fish and coronary heart disease risk.
Stratified by gender:
AuthorCohort nameCasesRelative Risk among menRelative Risk among women
44) Holmberg S (2009)No cohort name defined138OR = 1.00 (0.49-2.06)-
36) Iso H (2007)The JACC Study617 men, and

406 women
HR = 0.99 (0.80-1.22)HR = 0.79 (0.61-1.02; P = < 0.10)
34) Bjerregaard LJ (2010)The Danish part of the EPIC Study854 men, and

268 women
HR = 0.87 (0.69-1.10)HR = 0.85 (0.55-1.32)
34) Buckland G (2009)The Spanish part of the EPIC Study480 men, and

126 women
HR = 0.78 (0.62-0.98; P = 0.04)HR = 0.98 (0.63-1.52; P = 0.93)
33) Iso H (2006)The JPHC Studynot definedHR = 0.53 (0.30-0.94)-
29) Folsom AR (2004)The Iowa Women's Health Study922-RR = 1.04 (0.80-1.34; P = 0.31)
26) Osler M (2003)No cohort name defined349 men, and

142 women
HR = 1.05 (0.73-1.51; P = 0.94) for the highest vs third quartile of consumption
(RR reference group = 1.03)
HR = 0.64 (0.32-1.24; P = 0.21) for the highest vs third quartile of consumption
(RR reference group = 1.02)
23) Nagata C (2002)The Takayama Study63 men, and

52 women
HR = 1.05 (0.56-1.97; P = 0.91)HR = 0.73 (0.37-145; P = 0.37)
19) Yuan JM (2001)The Shanghai Cohort Study113 MI, and

74 other IHD
MI: RR = 0.35 (0.17-0.72; P = 0.02).

Other IHD: RR = 0.92 (0.41-2.06; P = 0.34).
-
18) Pietinen P (1997)The ATBC Study635RR = 1.12 (0.87-1.45; P = 0.09)-
17) Gillum RF (2000)The NHANES I StudyNot defined
(2,007 total)
White men: RR = 0.86 (0.65-1.13).

Black men: RR = 1.05 (0.50-2.19).
White women: RR = 0.97 (0.74-1.28).

Black women: RR = 0.90 (0.51-1.60).
16) Soinio M (2003)No cohort name defined74 men, and

43 women
No significant associationNo significant association
15) Albert CM (1998)The Physician's Health Study737RR = 1.00 (0.62-1.60; P = 0.67)-
14) Bernstein AM (2010)The Nurses' Health Study3,162-RR = 0.81 (0.72-0.90; P = < 0.001)
14) Ascherio A (1995)The Health Professionals Follow-up Study811RR = 0.90 (0.63-1.28; P = 0.70)-
13) Järvinen R (2006)The Finnish Mobile Clinic Health Survey335 men, and

163 women
RR = 1.00 (0.70-1.43; P = 0.83)RR = 0.59 (0.36-0.99; P = 0.02)
12) Kromhout D (1995)No cohort name definednot definedRR = 0.41 (0.20-0.86)RR = 0.64 (0.25-1.63)
11) Dolecek TA (1992)The Multiple Risk Factor Intervention Trial175RR = 0.61 (P = < 0.05)-
10) Hirayama T (1990)No cohort name definedIHD: 2,170 men, and 1,378 women.

Hypertensive heart disease: 559 men, and 613 women.

Other heart disease: not defined
IHD: RR = 1.20 (0.76-1.90).

Hypertensive heart disease: RR = 2.14 (1.07-4.27).

Other heart disease: RR = 1.02 (0.60-1.75).

All RR's are for low vs high consumption.
IHD: RR = 1.49 (1.02-2.19).

Hypertensive heart disease: RR = 0.52 (0.20-1.31).

Other heart disease: RR = 1.36 (0.97-1.92).

All RR's are for low vs high consumption.
8) Fraser GE (1997)The Aventist Health Study110 men, and

254 women
HR = 0.68 (0.25-1.83)HR = 0.82 (0.45-1.52)
6) Lapidus L (1986)No cohort name defined23-No significant association
4) Rodriguez BL (1996)The Honolulu Heart Programnot definedNo significant association-
3) Vollset SE (1985)No cohort name defined967No significant association (P = 0.93)-
2) Daviglus ML (1997)The Western Electric Study430RR = 0.62 (0.40-0.94; P = 0.04)-
1) Streppel MT (2008)The Dutch part of the Seven Countries Study336HR = 0.73 (0.47-1.13; P = 0.16)-
1) Oomen CM (2000)The Finnish & Italian part of the Seven Countries Study242 Finland, and

116 Italy
Finland: RR = 1.25 (0.89-1.76; P = 0.20).

Italy: RR = 0.67 (0.33-1.39; P = 0.33).
-
Total number of cases: 10,385 men, and
7,552 women
Average RR = 0.86Average RR = 0.90
Excluding data from Hirayama T [10].Total number of cases: 6,615 men, and
5,495 women
Average RR = 0.90Average RR = 0.84
Excluding data from Hirayama T [10].
And including coronary-bypass artery grafting as CHD (Ascherio A [14])
Total number of cases: 7,347 menAverage RR = 0.95-


Table 5.
Prospective studies of total fish and coronary heart disease risk among US cohorts:
AuthorCohort nameCasesEnd pointRelative Risk (RR)
45) Tomasallo C (2010)No cohort name defined34 captains, and

26 referents
MortalityCaptains: HR = 1.81 (0.53-6.16).

Referents: HR = 0.31 (0.10-0.96; P = < 0.05)
29) Folsom AR (2004)The Iowa Women's Health Study922MortalityRR = 1.04 (0.80-1.34; P = 0.31)
24) Mozaffarian D (2003)The Cardiovascular Health Study247 IHD death, and

363 nonfatal MI
RiskIHD death: HR = 0.47 (0.27-0.82; P = 0.002).

Nonfatal MI: HR = 0.67 (0.42-1.07; P = 0.10).
17) Gillum RF (2000)The NHANES I StudyNot defined
(2,007 total)
RiskWhite men: RR = 0.86 (0.65-1.13).

Black men: RR = 1.05 (0.50-2.19).

White women: RR = 0.97 (0.74-1.28).

Black women: RR = 0.90 (0.51-1.60).
15) Albert CM (1998)The Physician's Health Study737RiskRR = 1.00 (0.62-1.60; P = 0.67)
14) Bernstein AM (2010)The Nurses' Health Study3,162RiskRR = 0.81 (0.72-0.90; P = < 0.001)
14) Ascherio A (1995)The Health Professionals Follow-up Study811RiskRR = 0.90 (0.63-1.28; P = 0.70)
11) Dolecek TA (1992)The Multiple Risk Factor Intervention Trial175MortalityRR = 0.61 (P = < 0.05)
8) Fraser GE (1992)The Adventist Health Study134 nonfatal MI, and

463 CHD death
RiskNonfatal MI: RR = 1.04 (0.55-1.96).

CHD death: RR = 1.09 (0.73-1.61).
4) Rodriguez BL (1996)The Honolulu Heart Programnot definedRiskNo significant association
2) Daviglus ML (1997)The Western Electric Study430MortalityRR = 0.62 (0.40-0.94; P = 0.04)
Total number of cases: 9,511Average RR = 0.86


Table 6.
Prospective studies of total fish and coronary heart disease risk among European cohorts:
AuthorCohort nameCasesEnd pointRelative Risk (RR)
46) Manger MS (2010)The WENBIT Study210RiskHR = 0.93 (0.63-1.40; P = 0.72)
44) Holmberg S (2009)No cohort name defined138RiskOR = 1.00 (0.49-2.06)
34) de Goede J (2010)The Dutch part of the EPIC Study82 CHD mortality, and

252 nonfatal MI
RiskCHD mortality: HR = 0.52 (0.28-0.95; P = 0.02).

Nonfatal MI: HR = 1.01 (0.71-145; P= 0.14).
34) Bjerregaard LJ (2010)The Danish part of the EPIC Study854 men, and

268 women
RiskMen: HR = 0.87 (0.69-1.10).

Women: HR = 0.85 (0.55-1.32).
34) Buckland G (2009)The Spanish part of the EPIC Study606RiskHR = 0.83 (0.68-1.02; P = 0.82)
31) Ness AR (2005)The Boyd Orr Cohort298MortalityRR = 1.18 (0.80-1.76; P = 0.6)
30) Chang-Claude J (2005)The German Vegetarian Study60MortalityRR = 2.11 (1.13-3.96; P = 0.03)
28) Erkkilä AT (2003)The EUROASPIRE Study34RiskRR = 0.49 (0.17-1.41; P = 0.21)
26) Osler M (2003)No cohort name defined491RiskHR = 0.93 (0.68-1.27; P = 0.55)
22) Wennberg M (2011)The NSHDS263RiskOR = 1.21 (0.43-3.33; P = 0.52)
21) Whiteman D (1999)The OXCHECK Study93MortalityRR = 1.36 (0.57-3.25)
20) Mann JI (1997)The Oxford Vegetarian Study64MortalityDRR = 123 (70-217; P = NS)
18) Pietinen P (1997)The ATBC Study635MortalityRR = 1.12 (0.87-1.45; P = 0.09)
16) Soinio M (2003)No cohort name defined117RiskNo significant association
13) Järvinen R (2006)The Finnish Mobile Clinic Health Survey335 men, and

163 women
MortalityMen: RR = 1.00 (0.70-1.43; P = 0.83).

Women: RR = 0.59 (0.36-0.99; P = 0.02).
12) Kromhout D (1995)No cohort name defined58MortalityRR = 0.51 (0.29-0.89)
5) Norell SE (1986)The Cohort of Swedish Twins800MortalityRR = 0.85 (0.69-1.06)
3) Vollset SE (1985)No cohort name defined967MortalityNo significant association (P = 0.93)
1) Streppel MT (2008)The Dutch part of the Seven Countries Study336MortalityHR = 0.73 (0.47-1.13; P = 0.16)
1) Oomen CM (2000)The Finnish & Italian part of the Seven Countries Study242 Finland, and

116 Italy
MortalityFinland: RR = 1.25 (0.89-1.76; P = 0.20).

Italy: RR = 0.67 (0.33-1.39; P = 0.33).
Total number of cases: 7,482Average RR = 0.95


Table 7.
Prospective studies of total fish and coronary heart disease risk among Asian cohorts:
AuthorCohort nameCasesEnd pointRelative Risk (RR)
36) Yamagishi K (2008)The JACC Study419 IHD,

107 cardiac arrest, and

307 heart failure
MortalityIHD: HR = 0.86 (0.62-1.19; P = 0.41).

Cardiac arrest: HR = 0.73 (0.36-1.46; P = 0.16).

Heart failure: HR = 0.76 (0.53-1.07; P = 0.10).
33) Iso H (2006)The JPHC Study258RiskHR = 0.63 (0.38-1.04; P = 0.25)
32) Nakamura Y (2005)The NIPPON DATA80124MortalityRR = 0.86 (0.33-2.23; P = 0.51) for the highest vs second quintile of consumption
23) Nagata C (2002)The Takayama Study63 men, and

52 women
MortalityMen: HR = 1.05 (0.56-1.97; P = 0.91).

Women: HR = 0.73 (0.37-145; P = 0.37).
19) Yuan JM (2001)The Shanghai Cohort Study113 MI, and

74 other IHD
MortalityMI: RR = 0.35 (0.17-0.72; P = 0.02).

Other IHD: RR = 0.92 (0.41-2.06; P = 0.34).
10) Hirayama T (1990)No cohort name definedIHD: 2,170 men, and 1,378 women.

Hypertensive heart disease: 559 men, and 613 women.

Rheumatic heart disease: 364.

Other heart disease: not defined
MortalityIHD: RR's for low vs high consumption are 1.20 (0.76-1.90) for men, and 1.49 (1.02-2.19) for women.

Hypertensive heart disease: RR's for low vs high consumption are 2.14 (1.07-4.27) for men, and 0.52 (0.20-1.31) for women.

Rheumatic heart disease: RR = 0.95 (0.79-1.14).

Other heart disease: RR's for low vs high consumption are 1.02 (0.60-1.75) for men, and 1.36 (0.97-1.92) for women.
7) Tanaka H (1987)The Shibata Studynot definedRiskRR = 1.23
6) Lapidus L (1986)No cohort name defined23RiskNo significant association
Total number of cases: 6,624Average RR = 0.85