- Email: [email protected]
Fish consumption and risk of myocardial infarction: a systematic review and dose-response meta-analysis suggests a regional difference
Accepted Manuscript Fish consumption and risk of myocardial infarction: a systematic review and dose-response meta-analysis suggests a regional difference
Ahmad Jayedi, Mahdieh Sadat Zargar, Sakineh Shab-Bidar PII: DOI: Reference:
S0271-5317(18)30550-5 https://doi.org/10.1016/j.nutres.2018.10.009 NTR 7955
To appear in:
Nutrition Research
Received date: Revised date: Accepted date:
9 May 2018 8 September 2018 26 October 2018
Please cite this article as: Ahmad Jayedi, Mahdieh Sadat Zargar, Sakineh Shab-Bidar , Fish consumption and risk of myocardial infarction: a systematic review and doseresponse meta-analysis suggests a regional difference. Ntr (2018), https://doi.org/10.1016/ j.nutres.2018.10.009
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Fish consumption and risk of myocardial infarction: a systematic review and dose-response meta-analysis suggests a regional difference a
b
Ahmad Jayedi , Mahdieh Sadat Zargar , Sakineh Shab-Bidar
c
T
Nursing Care Research Center, Semnan University of Medical Sciences, Semnan, Iran
IP
b
Food (Salt) Safety Research Center, Semnan University of Medical Sciences, Semnan, Iran
Department of Community, Nutrition, School of Nutritional Science and Dietetics, Tehran University of
CR
a
c*
US
Medical Sciences, Tehran, Iran
AN
*Corresponding author: Sakineh Shab-Bidar, Associate Professor, Department of Community
M
Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences, P. O. Box
AC
CE
PT
ED
14155/6117, Tehran, Iran , Telefax: +98(21)88955979, Email: [email protected]
1
ACCEPTED MANUSCRIPT Abbreviations: CVD; cardiovascular diseases CHD; coronary heart disease MI; myocardial infarction
AC
CE
PT
ED
M
AN
US
CR
IP
T
PUFA; polyunsaturated fatty acid
2
ACCEPTED MANUSCRIPT Article outline
1. Introduction .............................................................................................................................................. 5 2. Methods and materials ............................................................................................................................. 6 2.1 Search strategy.................................................................................................................................... 6
T
2.2 Eligibility and study selection .............................................................................................................. 7
IP
2.3 Data extraction and quality assessment ............................................................................................. 7 2.4 Statistical analyses .............................................................................................................................. 8
CR
3. Results ....................................................................................................................................................... 9 3.1 Fish consumption and risk of myocardial infarction ......................................................................... 10
US
3.2 Sensitivity analysis, subgroup analysis, and publication bias ........................................................... 10 3.3 Nonlinear dose-response analysis .................................................................................................... 11
AN
3.4 Quality of meta-evidence.................................................................................................................. 11 4. Discussion................................................................................................................................................ 11 Acknowledgements:.................................................................................................................................... 17
AC
CE
PT
ED
M
References .................................................................................................................................................. 18
3
ACCEPTED MANUSCRIPT Abstract Limited evidence suggests that the association between fish consumption and risk of cardiovascular disease may be confounded by some regional-related factors. We aimed to quantify the association of fish consumption with risk of myocardial infarction (MI), and to
T
clarify the shape of the dose-response relation in Western and Asian countries. A systematic
IP
literature review was performed in PubMed and Scopus from inception to January 2018.
CR
Prospective observational studies reporting risk estimates of MI for three or more quantitave categories of fish intake were included. A random-effects dose-response meta-analysis was
US
conducted. Eleven prospective cohort studies, comprising a total of 398,221 participants and
AN
8468 cases of MI were analyzed. A significant inverse association was found for the highest compared with the lowest category of fish intake (RR: 0.73, 95%CI: 0.59, 0.87; I2 = 72%), and
M
for a 15 g/d (105 g/week, approximately equal to a one serving/week) increment in fish
ED
consumption (RR: 0.96, 95%CI: 0.94, 0.99; I2 = 65%). A subgroup analysis showed a significant inverse association only in the subgroup of Asian studies as compared to Western studies. A
PT
nonlinear dose-response analysis suggested a linear decrement in the risk with the increase in
CE
fish consumption in the analysis of Asian studies. A modest U-shaped association was observed in the analysis of Western studies. In conclusion, higher fish consumption was associated with a
AC
lower risk of MI. However, considering the observed regional difference in this association, further observational studies are needed to provide more detailed explanations about this difference. Keywords: fish; cardiovascular disease; meta-analysis; myocardial infarction; regional difference.
4
ACCEPTED MANUSCRIPT 1. Introduction Cardiovascular diseases (CVD) are the leading causes of death worldwide [1]. It has been estimated that about 18 million people died from CVD in 2015 (31% of all worldwide deaths); of these, about 7.5 million deaths were attributable to coronary heart disease (CHD) [2]. Myocardial
T
infarction (MI) is the most common manifestation of coronary artery disease [3], and about one-
IP
third of deaths in developed countries are due to MI [4]. Despite the high prevalence and many
CR
socio-economic burden of CVD, a substantial proportion of these diseases can be prevented via life-style related changes [5]. Hypertension, type 2 diabetes, adiposity, dyslipidemia, sedentary
AN
across all age subgroups, and all over the world [6].
US
life-style, and smoking have been proposed to be the main risk factors for MI in both sexes,
Besides these modifiable risk factors, a potentially contributing role for dietary factors
M
has been proposed. In particular, it has been demonstrated that adherence to healthier dietary
ED
patterns and lifestyle choices may contribute to lower risk of non-communicable diseases [7]. Fish, mainly due to the high content of omega-3 polyunsaturated fatty acids (PUFAs), are
PT
considered as one of the most important cardioprotective dietary components. Moreover, epidemiological evidence demonstrated the protective effect of fish and omega-3 PUFAs toward
CE
cerebrovascular disease [8], neurodegenerative diseases including dementia and Alzheimer’s
AC
disease [9], and depression [10]. Several meta-analyses of observational studies have shown that a higher fish consumption was associated with lower risk of stroke [11], CHD [12, 13], and cerebrovascular disease [14]. However, a recent meta-analysis of prospective cohort studies suggested an evident regional difference in the association between fish consumption and risk of CVD mortality [15]; in a way that, a significant inverse association was found only in the subgroup of Asian studies as compared to Western studies. Although the underlying mechanisms behind this difference 5
ACCEPTED MANUSCRIPT have not been clearly specified, some possible explanations such as method of preparing fish, different types of fish consumed in different regions, and potential local contaminations have been proposed. A previous meta-analysis of 18 prospective cohort and case-control studies demonstrated
T
that each 100 g/d increment in fish consumption was associated with a 5% lower risk of acute
IP
coronary syndrome [16], but potential regional difference in this association was not examined.
CR
Considering the fact that MI is one of the most important underlying causes of morbidity and mortality around the world, determining the possible regional difference in the association
US
between fish consumption and risk of MI may have important public health implications, and can
AN
provide a relatively new field to investigate the possible confounding effects of regional-related factors on the association between fish consumption and risk of CVD. Therefore, the objective of
M
this study was to test the association of fish consumption with risk of MI in the general
ED
population, and to investigate the possible regional difference in this association.
2.1 Search strategy
PT
2. Approach
CE
This study followed the requirements of the Meta-analysis of Observational Studies in Epidemiology (MOOSE) [17]. . checklist to write this systematic review and report the results
AC
[17].A systematic literature search was performed with the use of PubMed and Scopus, from inception to January 2018. The following combination of keywords was used to identify potential relevant articles: [“fish” OR “fishes” OR “seafood” OR “fish protein” OR “fish products” OR “marine” OR “animal protein”] AND [“death” OR “survival” OR “mortality” OR “fatal” OR “event” OR “events” OR “prognosis” OR “prognostic” OR “outcome” OR “Stroke” OR “Cerebrovascular disease” OR “intracranial hemorrhage” OR “Cerebral infarction” OR “CVD” OR “Cardiovascular disease” OR “Cardiovascular” OR “Myocardial infarction” OR 6
ACCEPTED MANUSCRIPT “Ischemic heart disease” OR “Coronary heart disease” OR “Cancer” OR “Carcinoma” OR “Neoplasm” OR “CHD” OR “IHD”] AND [“prospective” OR “prospectively” OR “Cohort” OR “Cohorts” OR “Longitudinal” OR “observational” OR “Observation” OR “Follow-up” OR “Nested”]. The reference lists of retrieved articles and relevant reviews were also manually
IP
T
searched. The search was restricted to articles published in the English language. 2.2 Eligibility and study selection
CR
Two independent authors (AJ, MSZ) reviewed the titles and abstracts of all studies identified, and selected prospective observational studies which (1) were conducted among adults aged 18
US
years or older; (2) reported fish intake as exposure and in at least 3 quantitative categories; (3)
AN
reported the outcome of interest as total, fatal, or nonfatal MI; (4) reported risk estimates (relative risk (RR) or hazard ratio) and their corresponding 95% confidence interval of MI for category
of
fish
consumption;
and
M
each
(5)
reported
numbers
of
cases
and
ED
participants/personyears or non-cases in each category of fish consumption, or reported sufficient
PT
information to estimate those numbers.
2.3 Data extraction and quality assessment
CE
Two independent investigators (AJ, MSZ) extracted the following information from eligible studies: first author’s name, publication year, study name, country, age range and/or mean age
AC
(year), number of participants/cases, dietary assessment method, exposure levels, reported risk estimates, and the 95%CI of MI across categories of fish intake, and covariates adjusted in the multivariate analysis. We selected effect estimates based on models with the most comprehensive covariate adjustments from each study. The Newcastle-Ottawa scale was used to assess the quality of included studies [18]. Furthermore, to provide more reliable measures for judgment about the quality of meta-evidence, we applied the NutriGrade scoring system (a
7
ACCEPTED MANUSCRIPT maximum of 10 points) [19]. Any discrepancies were resolved through discussion under supervision of a third author (SS-B). 2.4 Statistical analyses
The relative risk (RR) and 95%CI was considered as the effect size of all studies. The reported
T
hazard ratios were considered as equal to RR. For the highest versus lowest category meta-
IP
analysis, the reported risk estimates for the highest versus lowest category of fish intake were
CR
combined using the DerSimonian and Laird random-effects model [20]. If studies reported results across sex or other subgroups separately (for example, for fatal and nonfatal MI), we
US
combined subgroup-specific risk estimates using a fixed-effects model and used the combined
AN
effect size for meta-analysis. The linear dose-response relation was estimated using a generalized least squares trend estimation, according to the methods developed by Greenland and
M
Longnecker [21, 22]. This method needs distribution of cases and participants/personyears or
ED
non-cases and adjusted RR and its 95%CI across categories of fish intake. Study-specific results were combined using a random-effects model. The RR and its 95%CI were calculated for a 15
PT
g/d increment in fish consumption (105 g/week, approximately equal to a one serving/week). If
CE
the numbers of participants/cases or personyears have not been reported in the primary studies, we estimated them by dividing the total number of participants/cases or personyears by the
AC
number of categories, if the exposure was defined as quantiles [23] (this method was used only in one study). Between-studies heterogeneity was explored using Cochrane’s Q test of heterogeneity and I2 statistic (P < 0.05) [24]. Publication bias was assessed using funnel plots asymmetry and tested by Egger’s asymmetry test [25] and Begg’s test (P< 0.10)[26]. To test the potential effect of each study on pooled effect size, influence analysis was done with the stepwise exclusion of each study at a time. Subgroup analyses were done according to some of
8
ACCEPTED MANUSCRIPT the study and participant’s characteristics. A potential non-linear association was examined by modeling fish intake levels using restricted cubic splines with three knots at fixed percentiles (10, 50, and 90%) of the distribution [27].A P-value for non-linearity of the meta-analysis was calculated by testing the null hypothesis that the coefficient of the second spline was equal to
CR
IP
Texas, USA). A P-value <0.05 was considered statistically significant.
T
zero. All analyses were conducted with Stata software, version 13 (Stata Corp, College Station,
US
3. Results The initial systematic search identified 6836 articles, of which 652 were duplicates, and 6100 were considered non-relevant and excluded at the initial screening of the title and abstract. Of the
AN
remaining 84 articles, another 73 articles were eliminated by full text assessing, which respective reasons for study exclusion are presented in Fig. 1. Ultimately, 11 prospective cohort studies
M
with a total of 398,221 participants and 8468 cases of MI were included in the final analysis [28-
ED
38]. Five studies were from the US [28-30, 33, 36], three studies were from Europe [31, 32, 35], and three studies were from Asia [34, 37, 38]. Four studies included only men [28-30, 38], one
PT
study included only women [33], and the remainders included both sexes. Five studies reported
CE
total (fatal and nonfatal) MI as the outcome of interest [28, 29, 32, 34, 35], three studies reported risk estimate for fatal MI [30, 37, 38], two studies reported nonfatal MI [33, 36], and one study
AC
reported RR for fatal and nonfatal MI, separately [31]. All of the studies controlled for body mass index and smoking in their multivariate analyses, most studies controlled for alcohol consumption (n=10), and some of the studies controlled for energy intake (n=7) and intake of fruit and vegetables (n=5). All of the studies were at high quality. The general characteristics of the studies are presented in Table 1 and the numbers of cases and participants/personyears and
9
ACCEPTED MANUSCRIPT reported risk estimates of MI across categories of fish intake in each study are provided in Supplemental Table S1. 3.1 Fish consumption and risk of myocardial infarction
Eleven studies reported sufficient information for the association of fish consumption with risk
T
of MI. Highest compared with the lowest category of fish intake was associated with a 27%
IP
lower risk of MI (RR: 0.73, 95%CI: 0.59, 0.87), with high heterogeneity (I2 = 72%, 95%CI:
CR
48%, 85%; Pheterogeneity< 0.001) (Supplemental Figure S1). The linear dose-response analysis
US
indicated that a 15 g/d (105 g/week, approximately equal to a one serving/week) increment in fish consumption was associated with a 4% lower risk (RR: 0.96, 95%CI: 0.94, 0.99), with
AN
moderate-to-high heterogeneity (I2 = 65%, 95%CI: 32%, 88%; Pheterogeneity = 0.002) (Fig. 2). The
M
pooled RRs of fatal and nonfatal MI for a 15 g/d increment in fish intake were 0.90 (95%CI:
ED
0.80, 1.00; I2 = 83%, Pheterogeneity< 0.001; n = 6 studies), and 0.99 (95%CI: 0.96, 1.01; I2 = 0%,
PT
Pheterogeneity = 0.42; n = 5 studies), respectively.
3.2 Sensitivity analysis, subgroup analysis, and publication bias
CE
The pooled RR did not change materially with the stepwise exclusion of each study at a time (RR ranged between 0.95 and 0.97). None of the excluded studies explained the large degree of
AC
the heterogeneity in the data. The subgroup analyses yielded a significant inverse association only in the subgroup of Asian studies as compared to Western studies, studies with lower number of cases (<500 vs >500), studies without adjustment for physical activity, and among studies that controlled for energy intake (Table 2). The subgroup analyses suggested that geographical location, number of cases, and adjustment for energy intake were potential sources of the heterogeneity (Table 2). Egger’s regression asymmetry test (P = 0.001), and Begg’s rank
10
ACCEPTED MANUSCRIPT correlation test (P = 0.006) showed high evidence of publication bias (Supplemental Figure S2). 3.3 Nonlinear dose-response analysis
All of the studies were eligible for inclusion in the nonlinear dose-response analysis. A nonlinear
T
dose-response analysis demonstrated a linear inverse association in the main analysis (P for
IP
nonlinearity = 0.64, Fig. 3). We tested the dose-response relation across geographical location,
CR
which suggested a linear and U-shaped association in the analyses of Asian and Western studies, respectively (Fig. 4). In addition, the risk of fatal MI linearly decreased with the increase in fish
US
consumption (P for nonlinearity = 0.44; n = 6 studies), whereas a modest U-shaped association
AN
was observed in the analysis of nonfatal MI (P for nonlinearity = 0.003; n = 5 studies) (Fig. 4). 3.4 Quality of meta-evidence
M
All of the studies were at high quality (≥ 7 points). Additionally, the NutriGrade meta-evidence
ED
rating was “moderate” in this meta-analysis (NutriGrade score: 7.5); which suggests that there is
PT
a moderate confidence in effect estimate and further research may change the effect estimate.
CE
4. Discussion The present meta-analysis of prospective cohort studies provides supportive evidence about the cardioprotective properties of fish, and shows that higher fish consumption is associated with a
AC
27% lower risk of MI. Also, a 15 g/d increment of fish consumption was associated with a 4% lower risk of MI. A subgroup analysis based on geographical location suggested a significant inverse association in the subgroup of Asian studies, but not in Western studies. These findings are in agreement with those of previous meta-analyses of observational studies. A previous meta-analysis of 13 cohort studies showed that a 20 g/d increment of fish intake was associated with a 7% lower risk of CHD mortality [12]. Another meta-analysis of 18
11
ACCEPTED MANUSCRIPT cohort and case-control studies exhibited a 5% reduction in the risk of acute coronary syndrome for each 100 g/week increment in fish consumption [16]. However, possible difference across geographical location has not been well investigated. In the current review, a nonlinear doseresponse analysis demonstrated that the risk of MI linearly decreased in the analysis of Asian
T
studies, whereas a modest U-shaped association was observed in the analysis of Western studies.
IP
Also, we found a U-shaped association in the analysis of nonfatal MI, and this may be due to the
CR
fact that all of the studies included in the analysis of nonfatal MI were from Western countries. However, some important limitations must be acknowledged when interpreting the present
US
results. First, of the 11 studies included in the present review, only six studies controlled for
AN
physical activity (very important lifestyle factor in CVD); seven studies controlled for energy intake (important for body mass index and CVD); and five studies controlled for fruit and
M
vegetable intake (many CVD benefits). Also, subgroup analyses showed non-significant
ED
associations in the subgroups of the studies that controlled for physical activity and intake of fruit and vegetables. In general, higher fish consumption is associated with healthier dietary
PT
habits and better compliance with dietary guidelines. Thus, the possible confounding effects of
CE
these variables should not be ignored. Second, a large degree of the heterogeneity was observed in the high vs. low and in the linear dose-response analyses. The subgroup analyses suggested
AC
that geographical location, number of cases, and adjustment for energy intake were potential sources of the between-study heterogeneity. Additionally, of the 11 studies included in the present analysis, only one study reported effect size greater than one; hence the observed heterogeneity can be mainly attributable to differences in effect sizes of the studies examined, rather than inconsistencies in the direction of the association. Third, both Egger’s and Begg’s tests showed high evidence of publication bias, which may make our results biased toward a
12
ACCEPTED MANUSCRIPT greater effect size. Fourth, the number of Asian studies was low (n=3). Thus, the interpretation of the analysis of Asian subgroup should be made with caution. However, Asian studies were largescale, population-based, prospective cohort studies with high quality scores and long-term follow-up durations that controlled for much of the confounding variables. Fifth, the upper limit
T
of fish consumption in the analysis of Western studies (~ 125 g/d) was much lower than the
IP
corresponding intake in the analysis of Asian studies (~ 175 g/d). Thus, we have no conclusive
CR
evidence regarding fish consumption of > 125 g/d in Western countries. Finally, the number of fish items ascertained in dietary questionnaires varied substantially across studies, and almost all
US
of the studies did not take method of cooking fish into account in their dietary assessments.
AN
Additionally, only one study examined the association of different types of fish with risk of MI, and only one study evaluated the association across method of cooking fish. Thus, we were
M
unable to test the proposed hypothesis about different cardiovascular outcomes of different types
ED
of fish and different methods of cooking fish.
Cardioprotective properties of fish have been already well investigated. Fish, especially
PT
fatty fish, are rich in omega-3 fatty acids [39]. Long-chain omega-3 PUFAs including
CE
eicosapentaenoic acid and docosahexaenoic acid have anti-arrhythmic, anti-inflammatory, antihypertensive, and lipid-lowering properties [40-43]. They can improve vascular function, can
AC
mitigate oxidative stress [44-46], and may reduce platelet aggregation [47]. Also, some other cardiovascular-related features including reducing blood viscosity, reducing vascular intimal hyperplasia, and increasing arterial compliance have been attributed to dietary omega-3 PUFAs [48]. Additionally, fish are rich in other healthy dietary components such as essential amino acids, trace elements, vitamins, and other types of fats [39, 49].
13
ACCEPTED MANUSCRIPT A possible regional difference in the association between fish consumption and risk of all-cause and CVD mortality has been previously reported in previous meta-analyses [15, 50, 51]. Another meta-analysis of cohort studies suggested a positive association between fish intake and risk of type 2 diabetes among US populations, no association in European populations, and
T
an inverse association in Asian/Australian populations [52],
IP
The possible genetic, environmental, and lifestyle-related factors behind this difference
CR
have not been well investigated. However, some possible explanations have been proposed. People in different regions consume different types of fish. A cross-sectional analysis within the
US
EPIC study in Europe indicated that type of dietary fish differ substantially across countries [53].
AN
In general, fatty fish have higher content of omega-3 PUFAs as compared to lean fish and hence are expected to have better cardioprotective properties [12, 36]. To our knowledge, few studies
M
have assessed the distinct association of fatty and lean fish with risk of CVD. The Danish Diet,
ED
Cancer and Health study showed a non-significant inverse association between fatty fish consumption and risk of MI, and by contrast, suggested a non-significant trend toward a higher
PT
risk along with the increase in lean fish consumption [32]. Another prospective analysis on about
CE
20,000 participants in Italy exhibited that consumption of ≥ 1 serving/week of fatty fish, compared with none, was associated with a lower risk of CHD; whereas consumption of ≥ 1
AC
serving/week of lean fish was associated with a non-significant higher risk [54]. Also, two largescale, population-based, prospective cohort studies in China suggested that a higher intake of fatty fish was more strongly associated with a lower risk of all-cause and CVD mortality as compared to other types of fish [55]. However, fatty fish have a higher content of polychlorinated biphenyls which has been shown to be associated with a higher risk of CVD events [56, 57]. By contrast to the Chinese
14
ACCEPTED MANUSCRIPT cohorts [55], a recent prospective evaluation among 70,000 Swedish men and women showed a modest U-shaped association between fatty fish consumption with risk of all-cause mortality, especially among women [58]. When we compared the baseline characteristic of participants of these two studies, a higher intake of fruit and vegetables and lower intakes of red and processed
T
meat and total energy were observed in Chinese populations as compared with Swedish,
IP
especially in upper categories of fish intake; suggesting that other dietary habits may have a
CR
mediatory effect in the association between fish consumption and risk of mortality. Also, different method of cooking fish should be taken into consideration. Deep-frying is
US
a common method of preparing fish in Western countries which may diminish their healthy
AN
properties [36, 59]. The cooking method would be a factor in overall omega-3 PUFAs delivery, but sub optimal cooking methods may be counteracted by consuming more fish. By contrast,
M
people in Asian countries use healthier methods such as steaming and stir-frying [55]. These
ED
differences, in part, may contribute to the observed regional difference in the association between fish consumption and health outcomes [36, 58, 60]. Two large, prospective cohort
PT
studies in the US and Europe indicated that cooking methods for red meat and fish may make a
CE
distinction in their association with risk of type 2 diabetes [61, 62]; although, there is inconsistent evidence [63]. However, evidence regarding health outcomes of different types of
AC
fish and different methods of cooking fish is scarce, and further well-designed, prospective cohort studies may be needed to evaluate the association of different types of fish and different methods of cooking fish with risk of CVD. Our result about the U-shaped association between fish consumption and risk of MI in Western countries is in accordance with that of a recent meta-analysis of prospective cohort studies which has suggested a similar U-shaped association between fish consumption and risk
15
ACCEPTED MANUSCRIPT of all-cause and CVD mortality in Western countries [15]. These findings suggest that some possible factors such as unhealthy cooking methods, potential local contaminations, and adherence to an unhealthy dietary pattern may diminish healthy properties of fish. Fish are one of the main dietary sources of very long-chain omega-3 PUFAs and therefore are considered as one
T
the most important components of a healthy diet. Thus, promoting healthier methods for cooking
IP
fish and increasing the consumption of plant-based omega-3 PUFAs may be good suggestions to
CR
aid in the primary prevention of CVD.
The present study has several strengths. We showed a regional difference in the
US
association between fish intake and risk of MI. Considering the fact that coronary artery diseases
AN
are the leading cause of morbidity and mortality around the world, determining the possible optimal fish consumption for preventing these diseases may have important public health
M
implications. The present study provided a relatively new field to investigate the possible effects
ED
of regional-related factors on the association between fish consumption and risk of CVD. In addition, we included large-scale, prospective cohort studies with high rates of follow-up, high
PT
quality scores, and long-term follow-up durations.
CE
In summary, the current review presents supportive evidence about the cardioprotective properties of fish, and suggests that higher fish consumption is associated with a lower risk of
AC
MI. However, a possible difference was observed between Asian and Western countries. Further well-performed observational studies are warranted to provide more exact explanations about the observed difference in this association. Promoting healthier methods of cooking fish may be a good suggestion to aid in the primary prevention of CHD in Western countries.
16
ACCEPTED MANUSCRIPT Acknowledgment: This research did not receive any specific grant from funding agencies in the public, commercial,
AC
CE
PT
ED
M
AN
US
CR
IP
T
or not-for-profit sectors. The authors have no conflict of interest.
17
ACCEPTED MANUSCRIPT References: [1] Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 2016;388:1459-544.
T
[2] Global action plan for the prevention and control of non-communicable diseases 2013–2020.
IP
Geneva, Switzerland: World Health Organization; 2013.
CR
[3] Reed GW, Rossi JE, Cannon CP. Acute myocardial infarction. Lancet 2017;389:197-210. [4] Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS. Population trends in the incidence and
US
outcomes of acute myocardial infarction. N Engl J Med 2010;362:2155-65.
AN
[5] Franco M, Cooper RS, Bilal U, Fuster V. Challenges and opportunities for cardiovascular disease prevention. Am J Med 2011;124:95-102.
M
[6] Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, et al. Effect of potentially modifiable risk
study. Lancet 2004;364:937-52.
ED
factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control
PT
[7] Grosso G, Bella F, Godos J, Sciacca S, Del Rio D, Ray S, et al. Possible role of diet in cancer: systematic
2017;75:405-19.
CE
review and multiple meta-analyses of dietary patterns, lifestyle factors, and cancer risk. Nutr Rev
AC
[8] Chowdhury R, Stevens S, Gorman D, Pan A, Warnakula S, Chowdhury S, et al. Association between fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review and meta-analysis. Bmj 2012;345:e6698. [9] Wu S, Ding Y, Wu F, Li R, Hou J, Mao P. Omega-3 fatty acids intake and risks of dementia and Alzheimer's disease: a meta-analysis. Neurosci Biobehav Rev 2015;48:1-9.
18
ACCEPTED MANUSCRIPT [10] Grosso G, Pajak A, Marventano S, Castellano S, Galvano F, Bucolo C, et al. Role of omega-3 fatty acids in the treatment of depressive disorders: a comprehensive meta-analysis of randomized clinical trials. PLoS One 2014;9:e96905. [11] Larsson SC, Orsini N. Fish consumption and the risk of stroke: a dose-response meta-analysis. Stroke
T
2011;42:3621-3.
IP
[12] He K, Song Y, Daviglus ML, Liu K, Van Horn L, Dyer AR, et al. Accumulated evidence on fish
CR
consumption and coronary heart disease mortality: a meta-analysis of cohort studies. Circulation 2004;109:2705-11.
US
[13] Zheng J, Huang T, Yu Y, Hu X, Yang B, Li D. Fish consumption and CHD mortality: an updated meta-
AN
analysis of seventeen cohort studies. Public Health Nutr 2012;15:725-37. [14] Chowdhury R, Stevens S, Gorman D, Pan A, Warnakula S, Chowdhury S, et al. Association between
ED
and meta-analysis. Bmj 2012;345:e6698.
M
fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review
[15] Jayedi A, Shab-Bidar S, Eimeri S, Djafarian K. Fish consumption and risk of all-cause and
CE
Health Nutr 2018:1-10.
PT
cardiovascular mortality: a dose-response meta-analysis of prospective observational studies. Public
[16] Leung Yinko SS, Stark KD, Thanassoulis G, Pilote L. Fish consumption and acute coronary syndrome:
AC
a meta-analysis. Am J Med 2014;127:848-57.e2. [17] Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. Jama 2000;283:2008-12. [18] Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603-5.
19
ACCEPTED MANUSCRIPT [19] Schwingshackl L, Knuppel S, Schwedhelm C, Hoffmann G, Missbach B, Stelmach-Mardas M, et al. Perspective: NutriGrade: A Scoring System to Assess and Judge the Meta-Evidence of Randomized Controlled Trials and Cohort Studies in Nutrition Research. Adv Nutr 2016;7:994-1004. [20] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-88.
T
[21] Berlin JA, Longnecker MP, Greenland S. Meta-analysis of epidemiologic dose-response data.
IP
Epidemiology 1993;4:218-28.
CR
[22] Orsini N, Bellocco R, Greenland S. Generalized least squares for trend estimation of summarized dose-response data. Stata Journal 2006;6:40.
US
[23] Schwingshackl L, Schwedhelm C, Hoffmann G, Lampousi AM, Knuppel S, Iqbal K, et al. Food groups
AN
and risk of all-cause mortality: a systematic review and meta-analysis of prospective studies. Am J Clin Nutr 2017;105:1462-73.
ED
British Medical Journal 2003;327:557.
M
[24] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ:
[25] Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical
PT
test. Bmj 1997;315:629-34.
CE
[26] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994:1088-101.
AC
[27] Orsini N, Li R, Wolk A, Khudyakov P, Spiegelman D. Meta-analysis for linear and nonlinear doseresponse relations: examples, an evaluation of approximations, and software. American journal of epidemiology 2011;175:66-73. [28] Albert CM, Hennekens CH, O'Donnell CJ, Ajani UA, Carey VJ, Willett WC, et al. Fish consumption and risk of sudden cardiac death. Jama 1998;279:23-8. [29] Ascherio A, Rimm EB, Stampfer MJ, Giovannucci EL, Willett WC. Dietary intake of marine n-3 fatty acids, fish intake, and the risk of coronary disease among men. N Engl J Med 1995;332:977-82.
20
ACCEPTED MANUSCRIPT [30] Daviglus ML, Stamler J, Orencia AJ, Dyer AR, Liu K, Greenland P, et al. Fish consumption and the 30year risk of fatal myocardial infarction. N Engl J Med 1997;336:1046-53. [31] de Goede J, Geleijnse JM, Boer JM, Kromhout D, Verschuren WM. Marine (n-3) fatty acids, fish consumption, and the 10-year risk of fatal and nonfatal coronary heart disease in a large population of
T
Dutch adults with low fish intake. J Nutr 2010;140:1023-8.
IP
[32] Gammelmark A, Nielsen MS, Bork CS, Lundbye-Christensen S, Tjonneland A, Overvad K, et al.
prospective Danish cohort study. Br J Nutr 2016;116:167-77.
CR
Association of fish consumption and dietary intake of marine n-3 PUFA with myocardial infarction in a
US
[33] Hu FB, Bronner L, Willett WC, Stampfer MJ, Rexrode KM, Albert CM, et al. Fish and omega-3 fatty
AN
acid intake and risk of coronary heart disease in women. Jama 2002;287:1815-21. [34] Iso H, Kobayashi M, Ishihara J, Sasaki S, Okada K, Kita Y, et al. Intake of fish and n3 fatty acids and
ED
Cohort I. Circulation 2006;113:195-202.
M
risk of coronary heart disease among Japanese: the Japan Public Health Center-Based (JPHC) Study
[35] Kuhn T, Teucher B, Kaaks R, Boeing H, Weikert C, Buijsse B. Fish consumption and the risk of
PT
myocardial infarction and stroke in the German arm of the European Prospective Investigation into
CE
Cancer and Nutrition (EPIC-Germany). Br J Nutr 2013;110:1118-25. [36] Mozaffarian D, Lemaitre RN, Kuller LH, Burke GL, Tracy RP, Siscovick DS. Cardiac benefits of fish
AC
consumption may depend on the type of fish meal consumed: the Cardiovascular Health Study. Circulation 2003;107:1372-7. [37] Yamagishi K, Iso H, Date C, Fukui M, Wakai K, Kikuchi S, et al. Fish, omega-3 polyunsaturated fatty acids, and mortality from cardiovascular diseases in a nationwide community-based cohort of Japanese men and women the JACC (Japan Collaborative Cohort Study for Evaluation of Cancer Risk) Study. J Am Coll Cardiol 2008;52:988-96.
21
ACCEPTED MANUSCRIPT [38] Yuan JM, Ross RK, Gao YT, Yu MC. Fish and shellfish consumption in relation to death from myocardial infarction among men in Shanghai, China. Am J Epidemiol 2001;154:809-16. [39] He K. Fish, long-chain omega-3 polyunsaturated fatty acids and prevention of cardiovascular disease--eat fish or take fish oil supplement? Prog Cardiovasc Dis 2009;52:95-114.
T
[40] Breslow JL. n-3 fatty acids and cardiovascular disease. Am J Clin Nutr 2006;83:1477s-82s.
IP
[41] Brouwer IA, Geelen A, Katan MB. n-3 Fatty acids, cardiac arrhythmia and fatal coronary heart
CR
disease. Prog Lipid Res 2006;45:357-67.
[42] Cottin SC, Sanders TA, Hall WL. The differential effects of EPA and DHA on cardiovascular risk
US
factors. Proc Nutr Soc 2011;70:215-31.
AN
[43] Mozaffarian D, Wu JH. Omega-3 fatty acids and cardiovascular disease: effects on risk factors, molecular pathways, and clinical events. J Am Coll Cardiol 2011;58:2047-67.
M
[44] Colussi G, Catena C, Novello M, Bertin N, Sechi LA. Impact of omega-3 polyunsaturated fatty acids
ED
on vascular function and blood pressure: Relevance for cardiovascular outcomes. Nutr Metab Cardiovasc Dis 2017;27:191-200.
PT
[45] Li Q, Yu Q, Na R, Liu B. Omega-3 polyunsaturated fatty acids prevent murine dilated
CE
cardiomyopathy by reducing oxidative stress and cardiomyocyte apoptosis. Exp Ther Med 2017;14:61528.
AC
[46] Zanetti M, Gortan Cappellari G, Barbetta D, Semolic A, Barazzoni R. Omega 3 Polyunsaturated Fatty Acids Improve Endothelial Dysfunction in Chronic Renal Failure: Role of eNOS Activation and of Oxidative Stress. Nutrients 2017;9. [47] von Schacky C. n-3 fatty acids and the prevention of coronary atherosclerosis. Am J Clin Nutr 2000;71:224s-7s. [48] Simopoulos AP. Omega-3 fatty acids and cardiovascular disease: The epidemiological evidence. Environ Health Prev Med 2002;6:203-9.
22
ACCEPTED MANUSCRIPT [49] Lu Z, Chen TC, Zhang A, Persons KS, Kohn N, Berkowitz R, et al. An evaluation of the vitamin D3 content in fish: Is the vitamin D content adequate to satisfy the dietary requirement for vitamin D? J Steroid Biochem Mol Biol 2007;103:642-4. [50] Wan Y, Zheng J, Wang F, Li D. Fish, long chain omega-3 polyunsaturated fatty acids consumption,
T
and risk of all-cause mortality: a systematic review and dose-response meta-analysis from 23
IP
independent prospective cohort studies. Asia Pac J Clin Nutr 2017;26:939-56.
CR
[51] Zhao LG, Sun JW, Yang Y, Ma X, Wang YY, Xiang YB. Fish consumption and all-cause mortality: a meta-analysis of cohort studies. Eur J Clin Nutr 2016;70:155-61.
US
[52] Wallin A, Di Giuseppe D, Orsini N, Patel PS, Forouhi NG, Wolk A. Fish consumption, dietary long-
AN
chain n-3 fatty acids, and risk of type 2 diabetes: systematic review and meta-analysis of prospective studies. Diabetes Care 2012;35:918-29.
M
[53] Welch AA, Lund E, Amiano P, Dorronsoro M, Brustad M, Kumle M, et al. Variability of fish
ED
consumption within the 10 European countries participating in the European Investigation into Cancer and Nutrition (EPIC) study. Public Health Nutr 2002;5:1273-85.
PT
[54] Bonaccio M, Ruggiero E, Di Castelnuovo A, Costanzo S, Persichillo M, De Curtis A, et al. Fish intake is
CE
associated with lower cardiovascular risk in a Mediterranean population: Prospective results from the Moli-sani study. Nutr Metab Cardiovasc Dis 2017;27:865-73.
AC
[55] Takata Y, Zhang X, Li H, Gao YT, Yang G, Gao J, et al. Fish intake and risks of total and cause-specific mortality in 2 population-based cohort studies of 134,296 men and women. Am J Epidemiol 2013;178:46-57.
[56] Bergkvist C, Berglund M, Glynn A, Wolk A, Akesson A. Dietary exposure to polychlorinated biphenyls and risk of myocardial infarction - a population-based prospective cohort study. Int J Cardiol 2015;183:242-8.
23
ACCEPTED MANUSCRIPT [57] Bergkvist C, Kippler M, Larsson SC, Berglund M, Glynn A, Wolk A, et al. Dietary exposure to polychlorinated biphenyls is associated with increased risk of stroke in women. J Intern Med 2014;276:248-59. [58] Bellavia A, Larsson SC, Wolk A. Fish consumption and all-cause mortality in a cohort of Swedish men
T
and women. J Intern Med 2017;281:86-95.
IP
[59] Sauvaget C, Nagano J, Allen N, Grant EJ, Beral V. Intake of animal products and stroke mortality in
CR
the Hiroshima/Nagasaki Life Span Study. Int J Epidemiol 2003;32:536-43.
[60] Owen AJ, Magliano DJ, O'Dea K, Barr EL, Shaw JE. Polyunsaturated fatty acid intake and risk of
US
cardiovascular mortality in a low fish-consuming population: a prospective cohort analysis. Eur J Nutr
AN
2016;55:1605-13.
[61] Liu G, Zong G, Hu FB, Willett WC, Eisenberg DM, Sun Q. Cooking Methods for Red Meats and Risk of
M
Type 2 Diabetes: A Prospective Study of U.S. Women. Diabetes Care 2017;40:1041-9.
ED
[62] Wallin A, Di Giuseppe D, Orsini N, Akesson A, Forouhi NG, Wolk A. Fish consumption and frying of fish in relation to type 2 diabetes incidence: a prospective cohort study of Swedish men. Eur J Nutr
PT
2017;56:843-52.
CE
[63] Patel PS, Forouhi NG, Kuijsten A, Schulze MB, van Woudenbergh GJ, Ardanaz E, et al. The prospective association between total and type of fish intake and type 2 diabetes in 8 European
AC
countries: EPIC-InterAct Study. Am J Clin Nutr 2012;95:1445-53.
24
ACCEPTED MANUSCRIPT 6. Figure legends Figure 1. Literature search and study selection process for inclusion in meta-analysis of fish consumption and risk of myocardial infarction. Figure 2 Relative risk of myocardial infarction for a 15 g/d increment in fish consumption.
T
Figure 3 Dose-response associations between fish consumption and risk of myocardial infarction
IP
(All studies).
CR
Figure 4 Dose-response associations between fish consumption and risk of myocardial infarction
AC
CE
PT
ED
M
AN
US
(Subgroup analyses). MI, myocardial infarction.
25
ACCEPTED MANUSCRIPT Table 1. General characteristics of the studies included in meta-analysis of fish consumption and risk of myocardial infarction. Author name, year
Study name, country
Age range/ mean age (y)
Follow-up duration (y)
Participants
Gender
Number of cases
Albert [28], 1998
Physicians’ Health Study, US
40 – 84/ 53
11
20,551
M
Incident MI: 737
Exposure assessment
Quality score (max. 9 points)
Adjustments
FFQ
9
Age, aspirin and beta-carotene treatment assignment, evidence of CVDs, BMI, smoking status, alcohol consumption, history of DM, HTN, and hypercholesterolemia, vigorous exercise, and vitamin E, vitamin C, and multivitamin use.
FFQ
8
Age, BMI, smoking habits, alcohol consumption, history of HTN, history of DM; history of hypercholesterolemia, family history of myocardial infarction before 60 years of age, profession, and quintile group for intake of n-3 fatty acids.
FFQ
8
Age, education, religion, SBP, serum cholesterol, number of cigarettes smoked per day, BMI, presence or absence of DM, presence or absence of electrocardiographic abnormalities, and daily intake of energy, cholesterol, saturated,
T P
I R
C S
Ascherio [29], 1995
40 – 75
Health Professionals Follow-up Study, US
6
44,895
D E
A
M
M
T P
E C
Daviglus [30], 1997
C A
Chicago Western Electric Study, US
40 –55/ 47
30
U N
Total MI: 811
Nonfatal MI: 554
1,822
M
Fatal MI: 293
26
ACCEPTED MANUSCRIPT
de Goede [31], 2010
20 – 65/ 42
Monitoring Project on Risk Factors for Chronic Diseases (MORGEN) study, Netherlands
11.5
21,342
M/W
FFQ
7
Age, gender, BMI, total energy intake, ethanol intake, cigarette smoking, social economic status, vitamin or mineral supplement use, use of drugs for HTN or hypercholesterolemia, family history of CVD, SFA, fruit, and vegetables.
C S
U N
A
Gammelmark [32], 2016
T P
I R
Nonfatal MI: 252
monounsaturated, and polyunsaturated fatty acids, total protein, carbohydrate, alcohol, iron, thiamine, riboflavin, niacin, vitamin C, beta carotene, and retinol.
Danish Diet, Cancer and Health study, Denmark
50-64
D E
17
54,904
M
Fatal MI: 64
M/W
Incident MI: 3028
FFQ
8
Age, smoking, BMI, waist circumference, physical activity, alcohol intake, educational level, menopausal status (women), history of DM, HTN, hypercholesterolemia, total energy intake, intake of fruits and vegetables and intake of nuts.
W
Nonfatal MI: 1029
FFQ
8
Age, time periods, smoking status, BMI, alcohol intake, menopausal status and postmenopausal hormone use, vigorous to moderate activity, number of times aspirin was used per week, multivitamin use, vitamin E supplement use and
T P
E C
Hu [33], 2002
C A
Nurses’ Health Study, US
34-59
16
84,688
27
ACCEPTED MANUSCRIPT history of HTN, hypercholesterolemia, DM.
Iso [34], 2006
Japan Public Health Center-Based (JPHC) Study Cohort I, Japan
4059/50
11-12
41,578
M/W
Total MI: 221
FFQ
8
Age, sex, cigarette smoking, alcohol intake, BMI, histories of HTN, and DM, medication use for hypercholesterolemia, education level, sports at leisure time, quintiles of dietary intake of fruits, vegetables, SFA, monounsaturated fat, n-6 polyunsaturated fat, cholesterol, and total energy, and PHC.
FFQ
7
Stratified by age at baseline and study centers, adjusted for sex, energy intake, alcohol intake, BMI, waist circumference, physical activity, educational attainment, smoking and prevalent DM.
FFQ
8
Age, gender, education, DM, current smoking, pack-years of smoking, tuna/other fish and fried fish/fish sandwich consumption, BMI, systolic blood pressure, LDL cholesterol, HDL cholesterol, triglycerides, C-reactive protein, and intake of
T P
I R
C S
Ku¨hn [35], 2013
European Prospective Investigation into Cancer and Nutrition (EPIC-Germany), Germany
3565/51
8.1
48,315
D E
M/W
A
M
Nonfatal MI: 488
T P
E C
Mozaffarian [36], 2003
C A
Cardiovascular Health Study, US
≥ 65
9.3
U N
Total MI: 605
Fatal MI: 117
3910
W/M
Nonfatal MI: 363
28
ACCEPTED MANUSCRIPT saturated fat, alcohol, beef/pork, fruits, and vegetables.
Yamagishi [37], 2008
Yuan [38], 2001
Japan Collaborative Cohort Study for Evaluation of Cancer Risk (JACC) Study, Japan
40-79 /56
12.7
57,972
M/W
Fatal MI: 329
FFQ
7
Age, gender, HTN and DM, smoking status, alcohol consumption, BMI, mental stress, walking, sports, education levels, total energy, and dietary intakes of cholesterol, saturated and n-6 polyunsaturated fatty acids, vegetables, and fruit.
8
Age, total energy intake, level of education, BMI, current smoker at recruitment, average no. of cigarettes smoked per day, no. of alcoholic drinks consumed per week, history of DM, and history of HTN.
T P
I R
45 – 64/ 55
Prospective epidemiologic study of diet and cancer, China
12
18,244
M
U N
A
D E
M
C S
Fatal MI: 113
Interview based on questionnaire
Abbreviations: BMI, body mass index; CVD, cardiovascular disease; DM, diabetes mellitus; FFQ, food frequency questionnaire; HDL, high-density lipoprotein; HTN, hypertension; LDL, low-density lipoprotein; MI, myocardial infarction; M, men; PHC, public health center; SBP, systolic blood pressure; SFA, saturated fatty acid; W, women.
T P
E C
C A
29
ACCEPTED MANUSCRIPT
n
RR (95%CI)
I2(%), Pheterogeneity
11
0.96 (0.94-0.99)
65, 0.002
Men
5
0.98 (0.94-1.02)
74, 0.004
Women
2
0.98 (0.93-1.01)
0, 1.00
Both
6
0.95 (0.90-1.00)
67, 0.009
US + Europe
8
0.97 (0.94-1.00)
65, 0.005
Asia
3
0.94 (0.91-0.97)
IP
Table 2. Relative risk of myocardial infarction for a 15 g/d increment in fish consumption.
6
0.96 (0.92-1.01)
5
0.96 (0.92-1.00)
< 500
6
0.92 (0.88-0.97)
57, 0.05
> 500
5
1.00 (0.98-1.01)
0, 0.56
0.97 (0.94-0.99)
61, 0.006
All studies
1
2
Pbetween -
Sex
≥ 12 years
AN
Number of cases
10
No
1
Yes
6
Fruit and vegetable intake
0.98 (0.96-1.00)
45, 0.12
5
0.92 (0.84-0.99)
81, < 0.001
5
0.95 (0.90-1.00)
73, 0.006
6
0.97 (0.93-1.00)
63, 0.02
Yes
7
0.93 (0.89-0.98)
72, 0.001
No
4
1.00 (0.97-1.02)
0, 0.72
Yes
PT
No
2
< 0.001
0.02
0.41
0.71
0.02
P-heterogeneity within subgroups with the use of a random-effects model. P-heterogeneity between subgroups with the use of a fixed-effects model.
AC
1
0.60
CE
Energy intake
64, 0.03
-
No
0.007
71, 0.005
0.84 (0.72-0.96)
ED
Physical activity
Yes
M
Adjustment for confounders Alcohol consumption
CR
< 12 years
0, 0.73
US
Follow-up duration
T
Geographical region
0.72
30
Figure 1
Figure 2
Figure 3
Figure 4
Ahmad Jayedi, Mahdieh Sadat Zargar, Sakineh Shab-Bidar PII: DOI: Reference:
S0271-5317(18)30550-5 https://doi.org/10.1016/j.nutres.2018.10.009 NTR 7955
To appear in:
Nutrition Research
Received date: Revised date: Accepted date:
9 May 2018 8 September 2018 26 October 2018
Please cite this article as: Ahmad Jayedi, Mahdieh Sadat Zargar, Sakineh Shab-Bidar , Fish consumption and risk of myocardial infarction: a systematic review and doseresponse meta-analysis suggests a regional difference. Ntr (2018), https://doi.org/10.1016/ j.nutres.2018.10.009
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Fish consumption and risk of myocardial infarction: a systematic review and dose-response meta-analysis suggests a regional difference a
b
Ahmad Jayedi , Mahdieh Sadat Zargar , Sakineh Shab-Bidar
c
T
Nursing Care Research Center, Semnan University of Medical Sciences, Semnan, Iran
IP
b
Food (Salt) Safety Research Center, Semnan University of Medical Sciences, Semnan, Iran
Department of Community, Nutrition, School of Nutritional Science and Dietetics, Tehran University of
CR
a
c*
US
Medical Sciences, Tehran, Iran
AN
*Corresponding author: Sakineh Shab-Bidar, Associate Professor, Department of Community
M
Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences, P. O. Box
AC
CE
PT
ED
14155/6117, Tehran, Iran , Telefax: +98(21)88955979, Email: [email protected]
1
ACCEPTED MANUSCRIPT Abbreviations: CVD; cardiovascular diseases CHD; coronary heart disease MI; myocardial infarction
AC
CE
PT
ED
M
AN
US
CR
IP
T
PUFA; polyunsaturated fatty acid
2
ACCEPTED MANUSCRIPT Article outline
1. Introduction .............................................................................................................................................. 5 2. Methods and materials ............................................................................................................................. 6 2.1 Search strategy.................................................................................................................................... 6
T
2.2 Eligibility and study selection .............................................................................................................. 7
IP
2.3 Data extraction and quality assessment ............................................................................................. 7 2.4 Statistical analyses .............................................................................................................................. 8
CR
3. Results ....................................................................................................................................................... 9 3.1 Fish consumption and risk of myocardial infarction ......................................................................... 10
US
3.2 Sensitivity analysis, subgroup analysis, and publication bias ........................................................... 10 3.3 Nonlinear dose-response analysis .................................................................................................... 11
AN
3.4 Quality of meta-evidence.................................................................................................................. 11 4. Discussion................................................................................................................................................ 11 Acknowledgements:.................................................................................................................................... 17
AC
CE
PT
ED
M
References .................................................................................................................................................. 18
3
ACCEPTED MANUSCRIPT Abstract Limited evidence suggests that the association between fish consumption and risk of cardiovascular disease may be confounded by some regional-related factors. We aimed to quantify the association of fish consumption with risk of myocardial infarction (MI), and to
T
clarify the shape of the dose-response relation in Western and Asian countries. A systematic
IP
literature review was performed in PubMed and Scopus from inception to January 2018.
CR
Prospective observational studies reporting risk estimates of MI for three or more quantitave categories of fish intake were included. A random-effects dose-response meta-analysis was
US
conducted. Eleven prospective cohort studies, comprising a total of 398,221 participants and
AN
8468 cases of MI were analyzed. A significant inverse association was found for the highest compared with the lowest category of fish intake (RR: 0.73, 95%CI: 0.59, 0.87; I2 = 72%), and
M
for a 15 g/d (105 g/week, approximately equal to a one serving/week) increment in fish
ED
consumption (RR: 0.96, 95%CI: 0.94, 0.99; I2 = 65%). A subgroup analysis showed a significant inverse association only in the subgroup of Asian studies as compared to Western studies. A
PT
nonlinear dose-response analysis suggested a linear decrement in the risk with the increase in
CE
fish consumption in the analysis of Asian studies. A modest U-shaped association was observed in the analysis of Western studies. In conclusion, higher fish consumption was associated with a
AC
lower risk of MI. However, considering the observed regional difference in this association, further observational studies are needed to provide more detailed explanations about this difference. Keywords: fish; cardiovascular disease; meta-analysis; myocardial infarction; regional difference.
4
ACCEPTED MANUSCRIPT 1. Introduction Cardiovascular diseases (CVD) are the leading causes of death worldwide [1]. It has been estimated that about 18 million people died from CVD in 2015 (31% of all worldwide deaths); of these, about 7.5 million deaths were attributable to coronary heart disease (CHD) [2]. Myocardial
T
infarction (MI) is the most common manifestation of coronary artery disease [3], and about one-
IP
third of deaths in developed countries are due to MI [4]. Despite the high prevalence and many
CR
socio-economic burden of CVD, a substantial proportion of these diseases can be prevented via life-style related changes [5]. Hypertension, type 2 diabetes, adiposity, dyslipidemia, sedentary
AN
across all age subgroups, and all over the world [6].
US
life-style, and smoking have been proposed to be the main risk factors for MI in both sexes,
Besides these modifiable risk factors, a potentially contributing role for dietary factors
M
has been proposed. In particular, it has been demonstrated that adherence to healthier dietary
ED
patterns and lifestyle choices may contribute to lower risk of non-communicable diseases [7]. Fish, mainly due to the high content of omega-3 polyunsaturated fatty acids (PUFAs), are
PT
considered as one of the most important cardioprotective dietary components. Moreover, epidemiological evidence demonstrated the protective effect of fish and omega-3 PUFAs toward
CE
cerebrovascular disease [8], neurodegenerative diseases including dementia and Alzheimer’s
AC
disease [9], and depression [10]. Several meta-analyses of observational studies have shown that a higher fish consumption was associated with lower risk of stroke [11], CHD [12, 13], and cerebrovascular disease [14]. However, a recent meta-analysis of prospective cohort studies suggested an evident regional difference in the association between fish consumption and risk of CVD mortality [15]; in a way that, a significant inverse association was found only in the subgroup of Asian studies as compared to Western studies. Although the underlying mechanisms behind this difference 5
ACCEPTED MANUSCRIPT have not been clearly specified, some possible explanations such as method of preparing fish, different types of fish consumed in different regions, and potential local contaminations have been proposed. A previous meta-analysis of 18 prospective cohort and case-control studies demonstrated
T
that each 100 g/d increment in fish consumption was associated with a 5% lower risk of acute
IP
coronary syndrome [16], but potential regional difference in this association was not examined.
CR
Considering the fact that MI is one of the most important underlying causes of morbidity and mortality around the world, determining the possible regional difference in the association
US
between fish consumption and risk of MI may have important public health implications, and can
AN
provide a relatively new field to investigate the possible confounding effects of regional-related factors on the association between fish consumption and risk of CVD. Therefore, the objective of
M
this study was to test the association of fish consumption with risk of MI in the general
ED
population, and to investigate the possible regional difference in this association.
2.1 Search strategy
PT
2. Approach
CE
This study followed the requirements of the Meta-analysis of Observational Studies in Epidemiology (MOOSE) [17]. . checklist to write this systematic review and report the results
AC
[17].A systematic literature search was performed with the use of PubMed and Scopus, from inception to January 2018. The following combination of keywords was used to identify potential relevant articles: [“fish” OR “fishes” OR “seafood” OR “fish protein” OR “fish products” OR “marine” OR “animal protein”] AND [“death” OR “survival” OR “mortality” OR “fatal” OR “event” OR “events” OR “prognosis” OR “prognostic” OR “outcome” OR “Stroke” OR “Cerebrovascular disease” OR “intracranial hemorrhage” OR “Cerebral infarction” OR “CVD” OR “Cardiovascular disease” OR “Cardiovascular” OR “Myocardial infarction” OR 6
ACCEPTED MANUSCRIPT “Ischemic heart disease” OR “Coronary heart disease” OR “Cancer” OR “Carcinoma” OR “Neoplasm” OR “CHD” OR “IHD”] AND [“prospective” OR “prospectively” OR “Cohort” OR “Cohorts” OR “Longitudinal” OR “observational” OR “Observation” OR “Follow-up” OR “Nested”]. The reference lists of retrieved articles and relevant reviews were also manually
IP
T
searched. The search was restricted to articles published in the English language. 2.2 Eligibility and study selection
CR
Two independent authors (AJ, MSZ) reviewed the titles and abstracts of all studies identified, and selected prospective observational studies which (1) were conducted among adults aged 18
US
years or older; (2) reported fish intake as exposure and in at least 3 quantitative categories; (3)
AN
reported the outcome of interest as total, fatal, or nonfatal MI; (4) reported risk estimates (relative risk (RR) or hazard ratio) and their corresponding 95% confidence interval of MI for category
of
fish
consumption;
and
M
each
(5)
reported
numbers
of
cases
and
ED
participants/personyears or non-cases in each category of fish consumption, or reported sufficient
PT
information to estimate those numbers.
2.3 Data extraction and quality assessment
CE
Two independent investigators (AJ, MSZ) extracted the following information from eligible studies: first author’s name, publication year, study name, country, age range and/or mean age
AC
(year), number of participants/cases, dietary assessment method, exposure levels, reported risk estimates, and the 95%CI of MI across categories of fish intake, and covariates adjusted in the multivariate analysis. We selected effect estimates based on models with the most comprehensive covariate adjustments from each study. The Newcastle-Ottawa scale was used to assess the quality of included studies [18]. Furthermore, to provide more reliable measures for judgment about the quality of meta-evidence, we applied the NutriGrade scoring system (a
7
ACCEPTED MANUSCRIPT maximum of 10 points) [19]. Any discrepancies were resolved through discussion under supervision of a third author (SS-B). 2.4 Statistical analyses
The relative risk (RR) and 95%CI was considered as the effect size of all studies. The reported
T
hazard ratios were considered as equal to RR. For the highest versus lowest category meta-
IP
analysis, the reported risk estimates for the highest versus lowest category of fish intake were
CR
combined using the DerSimonian and Laird random-effects model [20]. If studies reported results across sex or other subgroups separately (for example, for fatal and nonfatal MI), we
US
combined subgroup-specific risk estimates using a fixed-effects model and used the combined
AN
effect size for meta-analysis. The linear dose-response relation was estimated using a generalized least squares trend estimation, according to the methods developed by Greenland and
M
Longnecker [21, 22]. This method needs distribution of cases and participants/personyears or
ED
non-cases and adjusted RR and its 95%CI across categories of fish intake. Study-specific results were combined using a random-effects model. The RR and its 95%CI were calculated for a 15
PT
g/d increment in fish consumption (105 g/week, approximately equal to a one serving/week). If
CE
the numbers of participants/cases or personyears have not been reported in the primary studies, we estimated them by dividing the total number of participants/cases or personyears by the
AC
number of categories, if the exposure was defined as quantiles [23] (this method was used only in one study). Between-studies heterogeneity was explored using Cochrane’s Q test of heterogeneity and I2 statistic (P < 0.05) [24]. Publication bias was assessed using funnel plots asymmetry and tested by Egger’s asymmetry test [25] and Begg’s test (P< 0.10)[26]. To test the potential effect of each study on pooled effect size, influence analysis was done with the stepwise exclusion of each study at a time. Subgroup analyses were done according to some of
8
ACCEPTED MANUSCRIPT the study and participant’s characteristics. A potential non-linear association was examined by modeling fish intake levels using restricted cubic splines with three knots at fixed percentiles (10, 50, and 90%) of the distribution [27].A P-value for non-linearity of the meta-analysis was calculated by testing the null hypothesis that the coefficient of the second spline was equal to
CR
IP
Texas, USA). A P-value <0.05 was considered statistically significant.
T
zero. All analyses were conducted with Stata software, version 13 (Stata Corp, College Station,
US
3. Results The initial systematic search identified 6836 articles, of which 652 were duplicates, and 6100 were considered non-relevant and excluded at the initial screening of the title and abstract. Of the
AN
remaining 84 articles, another 73 articles were eliminated by full text assessing, which respective reasons for study exclusion are presented in Fig. 1. Ultimately, 11 prospective cohort studies
M
with a total of 398,221 participants and 8468 cases of MI were included in the final analysis [28-
ED
38]. Five studies were from the US [28-30, 33, 36], three studies were from Europe [31, 32, 35], and three studies were from Asia [34, 37, 38]. Four studies included only men [28-30, 38], one
PT
study included only women [33], and the remainders included both sexes. Five studies reported
CE
total (fatal and nonfatal) MI as the outcome of interest [28, 29, 32, 34, 35], three studies reported risk estimate for fatal MI [30, 37, 38], two studies reported nonfatal MI [33, 36], and one study
AC
reported RR for fatal and nonfatal MI, separately [31]. All of the studies controlled for body mass index and smoking in their multivariate analyses, most studies controlled for alcohol consumption (n=10), and some of the studies controlled for energy intake (n=7) and intake of fruit and vegetables (n=5). All of the studies were at high quality. The general characteristics of the studies are presented in Table 1 and the numbers of cases and participants/personyears and
9
ACCEPTED MANUSCRIPT reported risk estimates of MI across categories of fish intake in each study are provided in Supplemental Table S1. 3.1 Fish consumption and risk of myocardial infarction
Eleven studies reported sufficient information for the association of fish consumption with risk
T
of MI. Highest compared with the lowest category of fish intake was associated with a 27%
IP
lower risk of MI (RR: 0.73, 95%CI: 0.59, 0.87), with high heterogeneity (I2 = 72%, 95%CI:
CR
48%, 85%; Pheterogeneity< 0.001) (Supplemental Figure S1). The linear dose-response analysis
US
indicated that a 15 g/d (105 g/week, approximately equal to a one serving/week) increment in fish consumption was associated with a 4% lower risk (RR: 0.96, 95%CI: 0.94, 0.99), with
AN
moderate-to-high heterogeneity (I2 = 65%, 95%CI: 32%, 88%; Pheterogeneity = 0.002) (Fig. 2). The
M
pooled RRs of fatal and nonfatal MI for a 15 g/d increment in fish intake were 0.90 (95%CI:
ED
0.80, 1.00; I2 = 83%, Pheterogeneity< 0.001; n = 6 studies), and 0.99 (95%CI: 0.96, 1.01; I2 = 0%,
PT
Pheterogeneity = 0.42; n = 5 studies), respectively.
3.2 Sensitivity analysis, subgroup analysis, and publication bias
CE
The pooled RR did not change materially with the stepwise exclusion of each study at a time (RR ranged between 0.95 and 0.97). None of the excluded studies explained the large degree of
AC
the heterogeneity in the data. The subgroup analyses yielded a significant inverse association only in the subgroup of Asian studies as compared to Western studies, studies with lower number of cases (<500 vs >500), studies without adjustment for physical activity, and among studies that controlled for energy intake (Table 2). The subgroup analyses suggested that geographical location, number of cases, and adjustment for energy intake were potential sources of the heterogeneity (Table 2). Egger’s regression asymmetry test (P = 0.001), and Begg’s rank
10
ACCEPTED MANUSCRIPT correlation test (P = 0.006) showed high evidence of publication bias (Supplemental Figure S2). 3.3 Nonlinear dose-response analysis
All of the studies were eligible for inclusion in the nonlinear dose-response analysis. A nonlinear
T
dose-response analysis demonstrated a linear inverse association in the main analysis (P for
IP
nonlinearity = 0.64, Fig. 3). We tested the dose-response relation across geographical location,
CR
which suggested a linear and U-shaped association in the analyses of Asian and Western studies, respectively (Fig. 4). In addition, the risk of fatal MI linearly decreased with the increase in fish
US
consumption (P for nonlinearity = 0.44; n = 6 studies), whereas a modest U-shaped association
AN
was observed in the analysis of nonfatal MI (P for nonlinearity = 0.003; n = 5 studies) (Fig. 4). 3.4 Quality of meta-evidence
M
All of the studies were at high quality (≥ 7 points). Additionally, the NutriGrade meta-evidence
ED
rating was “moderate” in this meta-analysis (NutriGrade score: 7.5); which suggests that there is
PT
a moderate confidence in effect estimate and further research may change the effect estimate.
CE
4. Discussion The present meta-analysis of prospective cohort studies provides supportive evidence about the cardioprotective properties of fish, and shows that higher fish consumption is associated with a
AC
27% lower risk of MI. Also, a 15 g/d increment of fish consumption was associated with a 4% lower risk of MI. A subgroup analysis based on geographical location suggested a significant inverse association in the subgroup of Asian studies, but not in Western studies. These findings are in agreement with those of previous meta-analyses of observational studies. A previous meta-analysis of 13 cohort studies showed that a 20 g/d increment of fish intake was associated with a 7% lower risk of CHD mortality [12]. Another meta-analysis of 18
11
ACCEPTED MANUSCRIPT cohort and case-control studies exhibited a 5% reduction in the risk of acute coronary syndrome for each 100 g/week increment in fish consumption [16]. However, possible difference across geographical location has not been well investigated. In the current review, a nonlinear doseresponse analysis demonstrated that the risk of MI linearly decreased in the analysis of Asian
T
studies, whereas a modest U-shaped association was observed in the analysis of Western studies.
IP
Also, we found a U-shaped association in the analysis of nonfatal MI, and this may be due to the
CR
fact that all of the studies included in the analysis of nonfatal MI were from Western countries. However, some important limitations must be acknowledged when interpreting the present
US
results. First, of the 11 studies included in the present review, only six studies controlled for
AN
physical activity (very important lifestyle factor in CVD); seven studies controlled for energy intake (important for body mass index and CVD); and five studies controlled for fruit and
M
vegetable intake (many CVD benefits). Also, subgroup analyses showed non-significant
ED
associations in the subgroups of the studies that controlled for physical activity and intake of fruit and vegetables. In general, higher fish consumption is associated with healthier dietary
PT
habits and better compliance with dietary guidelines. Thus, the possible confounding effects of
CE
these variables should not be ignored. Second, a large degree of the heterogeneity was observed in the high vs. low and in the linear dose-response analyses. The subgroup analyses suggested
AC
that geographical location, number of cases, and adjustment for energy intake were potential sources of the between-study heterogeneity. Additionally, of the 11 studies included in the present analysis, only one study reported effect size greater than one; hence the observed heterogeneity can be mainly attributable to differences in effect sizes of the studies examined, rather than inconsistencies in the direction of the association. Third, both Egger’s and Begg’s tests showed high evidence of publication bias, which may make our results biased toward a
12
ACCEPTED MANUSCRIPT greater effect size. Fourth, the number of Asian studies was low (n=3). Thus, the interpretation of the analysis of Asian subgroup should be made with caution. However, Asian studies were largescale, population-based, prospective cohort studies with high quality scores and long-term follow-up durations that controlled for much of the confounding variables. Fifth, the upper limit
T
of fish consumption in the analysis of Western studies (~ 125 g/d) was much lower than the
IP
corresponding intake in the analysis of Asian studies (~ 175 g/d). Thus, we have no conclusive
CR
evidence regarding fish consumption of > 125 g/d in Western countries. Finally, the number of fish items ascertained in dietary questionnaires varied substantially across studies, and almost all
US
of the studies did not take method of cooking fish into account in their dietary assessments.
AN
Additionally, only one study examined the association of different types of fish with risk of MI, and only one study evaluated the association across method of cooking fish. Thus, we were
M
unable to test the proposed hypothesis about different cardiovascular outcomes of different types
ED
of fish and different methods of cooking fish.
Cardioprotective properties of fish have been already well investigated. Fish, especially
PT
fatty fish, are rich in omega-3 fatty acids [39]. Long-chain omega-3 PUFAs including
CE
eicosapentaenoic acid and docosahexaenoic acid have anti-arrhythmic, anti-inflammatory, antihypertensive, and lipid-lowering properties [40-43]. They can improve vascular function, can
AC
mitigate oxidative stress [44-46], and may reduce platelet aggregation [47]. Also, some other cardiovascular-related features including reducing blood viscosity, reducing vascular intimal hyperplasia, and increasing arterial compliance have been attributed to dietary omega-3 PUFAs [48]. Additionally, fish are rich in other healthy dietary components such as essential amino acids, trace elements, vitamins, and other types of fats [39, 49].
13
ACCEPTED MANUSCRIPT A possible regional difference in the association between fish consumption and risk of all-cause and CVD mortality has been previously reported in previous meta-analyses [15, 50, 51]. Another meta-analysis of cohort studies suggested a positive association between fish intake and risk of type 2 diabetes among US populations, no association in European populations, and
T
an inverse association in Asian/Australian populations [52],
IP
The possible genetic, environmental, and lifestyle-related factors behind this difference
CR
have not been well investigated. However, some possible explanations have been proposed. People in different regions consume different types of fish. A cross-sectional analysis within the
US
EPIC study in Europe indicated that type of dietary fish differ substantially across countries [53].
AN
In general, fatty fish have higher content of omega-3 PUFAs as compared to lean fish and hence are expected to have better cardioprotective properties [12, 36]. To our knowledge, few studies
M
have assessed the distinct association of fatty and lean fish with risk of CVD. The Danish Diet,
ED
Cancer and Health study showed a non-significant inverse association between fatty fish consumption and risk of MI, and by contrast, suggested a non-significant trend toward a higher
PT
risk along with the increase in lean fish consumption [32]. Another prospective analysis on about
CE
20,000 participants in Italy exhibited that consumption of ≥ 1 serving/week of fatty fish, compared with none, was associated with a lower risk of CHD; whereas consumption of ≥ 1
AC
serving/week of lean fish was associated with a non-significant higher risk [54]. Also, two largescale, population-based, prospective cohort studies in China suggested that a higher intake of fatty fish was more strongly associated with a lower risk of all-cause and CVD mortality as compared to other types of fish [55]. However, fatty fish have a higher content of polychlorinated biphenyls which has been shown to be associated with a higher risk of CVD events [56, 57]. By contrast to the Chinese
14
ACCEPTED MANUSCRIPT cohorts [55], a recent prospective evaluation among 70,000 Swedish men and women showed a modest U-shaped association between fatty fish consumption with risk of all-cause mortality, especially among women [58]. When we compared the baseline characteristic of participants of these two studies, a higher intake of fruit and vegetables and lower intakes of red and processed
T
meat and total energy were observed in Chinese populations as compared with Swedish,
IP
especially in upper categories of fish intake; suggesting that other dietary habits may have a
CR
mediatory effect in the association between fish consumption and risk of mortality. Also, different method of cooking fish should be taken into consideration. Deep-frying is
US
a common method of preparing fish in Western countries which may diminish their healthy
AN
properties [36, 59]. The cooking method would be a factor in overall omega-3 PUFAs delivery, but sub optimal cooking methods may be counteracted by consuming more fish. By contrast,
M
people in Asian countries use healthier methods such as steaming and stir-frying [55]. These
ED
differences, in part, may contribute to the observed regional difference in the association between fish consumption and health outcomes [36, 58, 60]. Two large, prospective cohort
PT
studies in the US and Europe indicated that cooking methods for red meat and fish may make a
CE
distinction in their association with risk of type 2 diabetes [61, 62]; although, there is inconsistent evidence [63]. However, evidence regarding health outcomes of different types of
AC
fish and different methods of cooking fish is scarce, and further well-designed, prospective cohort studies may be needed to evaluate the association of different types of fish and different methods of cooking fish with risk of CVD. Our result about the U-shaped association between fish consumption and risk of MI in Western countries is in accordance with that of a recent meta-analysis of prospective cohort studies which has suggested a similar U-shaped association between fish consumption and risk
15
ACCEPTED MANUSCRIPT of all-cause and CVD mortality in Western countries [15]. These findings suggest that some possible factors such as unhealthy cooking methods, potential local contaminations, and adherence to an unhealthy dietary pattern may diminish healthy properties of fish. Fish are one of the main dietary sources of very long-chain omega-3 PUFAs and therefore are considered as one
T
the most important components of a healthy diet. Thus, promoting healthier methods for cooking
IP
fish and increasing the consumption of plant-based omega-3 PUFAs may be good suggestions to
CR
aid in the primary prevention of CVD.
The present study has several strengths. We showed a regional difference in the
US
association between fish intake and risk of MI. Considering the fact that coronary artery diseases
AN
are the leading cause of morbidity and mortality around the world, determining the possible optimal fish consumption for preventing these diseases may have important public health
M
implications. The present study provided a relatively new field to investigate the possible effects
ED
of regional-related factors on the association between fish consumption and risk of CVD. In addition, we included large-scale, prospective cohort studies with high rates of follow-up, high
PT
quality scores, and long-term follow-up durations.
CE
In summary, the current review presents supportive evidence about the cardioprotective properties of fish, and suggests that higher fish consumption is associated with a lower risk of
AC
MI. However, a possible difference was observed between Asian and Western countries. Further well-performed observational studies are warranted to provide more exact explanations about the observed difference in this association. Promoting healthier methods of cooking fish may be a good suggestion to aid in the primary prevention of CHD in Western countries.
16
ACCEPTED MANUSCRIPT Acknowledgment: This research did not receive any specific grant from funding agencies in the public, commercial,
AC
CE
PT
ED
M
AN
US
CR
IP
T
or not-for-profit sectors. The authors have no conflict of interest.
17
ACCEPTED MANUSCRIPT References: [1] Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 2016;388:1459-544.
T
[2] Global action plan for the prevention and control of non-communicable diseases 2013–2020.
IP
Geneva, Switzerland: World Health Organization; 2013.
CR
[3] Reed GW, Rossi JE, Cannon CP. Acute myocardial infarction. Lancet 2017;389:197-210. [4] Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS. Population trends in the incidence and
US
outcomes of acute myocardial infarction. N Engl J Med 2010;362:2155-65.
AN
[5] Franco M, Cooper RS, Bilal U, Fuster V. Challenges and opportunities for cardiovascular disease prevention. Am J Med 2011;124:95-102.
M
[6] Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, et al. Effect of potentially modifiable risk
study. Lancet 2004;364:937-52.
ED
factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control
PT
[7] Grosso G, Bella F, Godos J, Sciacca S, Del Rio D, Ray S, et al. Possible role of diet in cancer: systematic
2017;75:405-19.
CE
review and multiple meta-analyses of dietary patterns, lifestyle factors, and cancer risk. Nutr Rev
AC
[8] Chowdhury R, Stevens S, Gorman D, Pan A, Warnakula S, Chowdhury S, et al. Association between fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review and meta-analysis. Bmj 2012;345:e6698. [9] Wu S, Ding Y, Wu F, Li R, Hou J, Mao P. Omega-3 fatty acids intake and risks of dementia and Alzheimer's disease: a meta-analysis. Neurosci Biobehav Rev 2015;48:1-9.
18
ACCEPTED MANUSCRIPT [10] Grosso G, Pajak A, Marventano S, Castellano S, Galvano F, Bucolo C, et al. Role of omega-3 fatty acids in the treatment of depressive disorders: a comprehensive meta-analysis of randomized clinical trials. PLoS One 2014;9:e96905. [11] Larsson SC, Orsini N. Fish consumption and the risk of stroke: a dose-response meta-analysis. Stroke
T
2011;42:3621-3.
IP
[12] He K, Song Y, Daviglus ML, Liu K, Van Horn L, Dyer AR, et al. Accumulated evidence on fish
CR
consumption and coronary heart disease mortality: a meta-analysis of cohort studies. Circulation 2004;109:2705-11.
US
[13] Zheng J, Huang T, Yu Y, Hu X, Yang B, Li D. Fish consumption and CHD mortality: an updated meta-
AN
analysis of seventeen cohort studies. Public Health Nutr 2012;15:725-37. [14] Chowdhury R, Stevens S, Gorman D, Pan A, Warnakula S, Chowdhury S, et al. Association between
ED
and meta-analysis. Bmj 2012;345:e6698.
M
fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review
[15] Jayedi A, Shab-Bidar S, Eimeri S, Djafarian K. Fish consumption and risk of all-cause and
CE
Health Nutr 2018:1-10.
PT
cardiovascular mortality: a dose-response meta-analysis of prospective observational studies. Public
[16] Leung Yinko SS, Stark KD, Thanassoulis G, Pilote L. Fish consumption and acute coronary syndrome:
AC
a meta-analysis. Am J Med 2014;127:848-57.e2. [17] Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. Jama 2000;283:2008-12. [18] Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603-5.
19
ACCEPTED MANUSCRIPT [19] Schwingshackl L, Knuppel S, Schwedhelm C, Hoffmann G, Missbach B, Stelmach-Mardas M, et al. Perspective: NutriGrade: A Scoring System to Assess and Judge the Meta-Evidence of Randomized Controlled Trials and Cohort Studies in Nutrition Research. Adv Nutr 2016;7:994-1004. [20] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-88.
T
[21] Berlin JA, Longnecker MP, Greenland S. Meta-analysis of epidemiologic dose-response data.
IP
Epidemiology 1993;4:218-28.
CR
[22] Orsini N, Bellocco R, Greenland S. Generalized least squares for trend estimation of summarized dose-response data. Stata Journal 2006;6:40.
US
[23] Schwingshackl L, Schwedhelm C, Hoffmann G, Lampousi AM, Knuppel S, Iqbal K, et al. Food groups
AN
and risk of all-cause mortality: a systematic review and meta-analysis of prospective studies. Am J Clin Nutr 2017;105:1462-73.
ED
British Medical Journal 2003;327:557.
M
[24] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ:
[25] Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical
PT
test. Bmj 1997;315:629-34.
CE
[26] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994:1088-101.
AC
[27] Orsini N, Li R, Wolk A, Khudyakov P, Spiegelman D. Meta-analysis for linear and nonlinear doseresponse relations: examples, an evaluation of approximations, and software. American journal of epidemiology 2011;175:66-73. [28] Albert CM, Hennekens CH, O'Donnell CJ, Ajani UA, Carey VJ, Willett WC, et al. Fish consumption and risk of sudden cardiac death. Jama 1998;279:23-8. [29] Ascherio A, Rimm EB, Stampfer MJ, Giovannucci EL, Willett WC. Dietary intake of marine n-3 fatty acids, fish intake, and the risk of coronary disease among men. N Engl J Med 1995;332:977-82.
20
ACCEPTED MANUSCRIPT [30] Daviglus ML, Stamler J, Orencia AJ, Dyer AR, Liu K, Greenland P, et al. Fish consumption and the 30year risk of fatal myocardial infarction. N Engl J Med 1997;336:1046-53. [31] de Goede J, Geleijnse JM, Boer JM, Kromhout D, Verschuren WM. Marine (n-3) fatty acids, fish consumption, and the 10-year risk of fatal and nonfatal coronary heart disease in a large population of
T
Dutch adults with low fish intake. J Nutr 2010;140:1023-8.
IP
[32] Gammelmark A, Nielsen MS, Bork CS, Lundbye-Christensen S, Tjonneland A, Overvad K, et al.
prospective Danish cohort study. Br J Nutr 2016;116:167-77.
CR
Association of fish consumption and dietary intake of marine n-3 PUFA with myocardial infarction in a
US
[33] Hu FB, Bronner L, Willett WC, Stampfer MJ, Rexrode KM, Albert CM, et al. Fish and omega-3 fatty
AN
acid intake and risk of coronary heart disease in women. Jama 2002;287:1815-21. [34] Iso H, Kobayashi M, Ishihara J, Sasaki S, Okada K, Kita Y, et al. Intake of fish and n3 fatty acids and
ED
Cohort I. Circulation 2006;113:195-202.
M
risk of coronary heart disease among Japanese: the Japan Public Health Center-Based (JPHC) Study
[35] Kuhn T, Teucher B, Kaaks R, Boeing H, Weikert C, Buijsse B. Fish consumption and the risk of
PT
myocardial infarction and stroke in the German arm of the European Prospective Investigation into
CE
Cancer and Nutrition (EPIC-Germany). Br J Nutr 2013;110:1118-25. [36] Mozaffarian D, Lemaitre RN, Kuller LH, Burke GL, Tracy RP, Siscovick DS. Cardiac benefits of fish
AC
consumption may depend on the type of fish meal consumed: the Cardiovascular Health Study. Circulation 2003;107:1372-7. [37] Yamagishi K, Iso H, Date C, Fukui M, Wakai K, Kikuchi S, et al. Fish, omega-3 polyunsaturated fatty acids, and mortality from cardiovascular diseases in a nationwide community-based cohort of Japanese men and women the JACC (Japan Collaborative Cohort Study for Evaluation of Cancer Risk) Study. J Am Coll Cardiol 2008;52:988-96.
21
ACCEPTED MANUSCRIPT [38] Yuan JM, Ross RK, Gao YT, Yu MC. Fish and shellfish consumption in relation to death from myocardial infarction among men in Shanghai, China. Am J Epidemiol 2001;154:809-16. [39] He K. Fish, long-chain omega-3 polyunsaturated fatty acids and prevention of cardiovascular disease--eat fish or take fish oil supplement? Prog Cardiovasc Dis 2009;52:95-114.
T
[40] Breslow JL. n-3 fatty acids and cardiovascular disease. Am J Clin Nutr 2006;83:1477s-82s.
IP
[41] Brouwer IA, Geelen A, Katan MB. n-3 Fatty acids, cardiac arrhythmia and fatal coronary heart
CR
disease. Prog Lipid Res 2006;45:357-67.
[42] Cottin SC, Sanders TA, Hall WL. The differential effects of EPA and DHA on cardiovascular risk
US
factors. Proc Nutr Soc 2011;70:215-31.
AN
[43] Mozaffarian D, Wu JH. Omega-3 fatty acids and cardiovascular disease: effects on risk factors, molecular pathways, and clinical events. J Am Coll Cardiol 2011;58:2047-67.
M
[44] Colussi G, Catena C, Novello M, Bertin N, Sechi LA. Impact of omega-3 polyunsaturated fatty acids
ED
on vascular function and blood pressure: Relevance for cardiovascular outcomes. Nutr Metab Cardiovasc Dis 2017;27:191-200.
PT
[45] Li Q, Yu Q, Na R, Liu B. Omega-3 polyunsaturated fatty acids prevent murine dilated
CE
cardiomyopathy by reducing oxidative stress and cardiomyocyte apoptosis. Exp Ther Med 2017;14:61528.
AC
[46] Zanetti M, Gortan Cappellari G, Barbetta D, Semolic A, Barazzoni R. Omega 3 Polyunsaturated Fatty Acids Improve Endothelial Dysfunction in Chronic Renal Failure: Role of eNOS Activation and of Oxidative Stress. Nutrients 2017;9. [47] von Schacky C. n-3 fatty acids and the prevention of coronary atherosclerosis. Am J Clin Nutr 2000;71:224s-7s. [48] Simopoulos AP. Omega-3 fatty acids and cardiovascular disease: The epidemiological evidence. Environ Health Prev Med 2002;6:203-9.
22
ACCEPTED MANUSCRIPT [49] Lu Z, Chen TC, Zhang A, Persons KS, Kohn N, Berkowitz R, et al. An evaluation of the vitamin D3 content in fish: Is the vitamin D content adequate to satisfy the dietary requirement for vitamin D? J Steroid Biochem Mol Biol 2007;103:642-4. [50] Wan Y, Zheng J, Wang F, Li D. Fish, long chain omega-3 polyunsaturated fatty acids consumption,
T
and risk of all-cause mortality: a systematic review and dose-response meta-analysis from 23
IP
independent prospective cohort studies. Asia Pac J Clin Nutr 2017;26:939-56.
CR
[51] Zhao LG, Sun JW, Yang Y, Ma X, Wang YY, Xiang YB. Fish consumption and all-cause mortality: a meta-analysis of cohort studies. Eur J Clin Nutr 2016;70:155-61.
US
[52] Wallin A, Di Giuseppe D, Orsini N, Patel PS, Forouhi NG, Wolk A. Fish consumption, dietary long-
AN
chain n-3 fatty acids, and risk of type 2 diabetes: systematic review and meta-analysis of prospective studies. Diabetes Care 2012;35:918-29.
M
[53] Welch AA, Lund E, Amiano P, Dorronsoro M, Brustad M, Kumle M, et al. Variability of fish
ED
consumption within the 10 European countries participating in the European Investigation into Cancer and Nutrition (EPIC) study. Public Health Nutr 2002;5:1273-85.
PT
[54] Bonaccio M, Ruggiero E, Di Castelnuovo A, Costanzo S, Persichillo M, De Curtis A, et al. Fish intake is
CE
associated with lower cardiovascular risk in a Mediterranean population: Prospective results from the Moli-sani study. Nutr Metab Cardiovasc Dis 2017;27:865-73.
AC
[55] Takata Y, Zhang X, Li H, Gao YT, Yang G, Gao J, et al. Fish intake and risks of total and cause-specific mortality in 2 population-based cohort studies of 134,296 men and women. Am J Epidemiol 2013;178:46-57.
[56] Bergkvist C, Berglund M, Glynn A, Wolk A, Akesson A. Dietary exposure to polychlorinated biphenyls and risk of myocardial infarction - a population-based prospective cohort study. Int J Cardiol 2015;183:242-8.
23
ACCEPTED MANUSCRIPT [57] Bergkvist C, Kippler M, Larsson SC, Berglund M, Glynn A, Wolk A, et al. Dietary exposure to polychlorinated biphenyls is associated with increased risk of stroke in women. J Intern Med 2014;276:248-59. [58] Bellavia A, Larsson SC, Wolk A. Fish consumption and all-cause mortality in a cohort of Swedish men
T
and women. J Intern Med 2017;281:86-95.
IP
[59] Sauvaget C, Nagano J, Allen N, Grant EJ, Beral V. Intake of animal products and stroke mortality in
CR
the Hiroshima/Nagasaki Life Span Study. Int J Epidemiol 2003;32:536-43.
[60] Owen AJ, Magliano DJ, O'Dea K, Barr EL, Shaw JE. Polyunsaturated fatty acid intake and risk of
US
cardiovascular mortality in a low fish-consuming population: a prospective cohort analysis. Eur J Nutr
AN
2016;55:1605-13.
[61] Liu G, Zong G, Hu FB, Willett WC, Eisenberg DM, Sun Q. Cooking Methods for Red Meats and Risk of
M
Type 2 Diabetes: A Prospective Study of U.S. Women. Diabetes Care 2017;40:1041-9.
ED
[62] Wallin A, Di Giuseppe D, Orsini N, Akesson A, Forouhi NG, Wolk A. Fish consumption and frying of fish in relation to type 2 diabetes incidence: a prospective cohort study of Swedish men. Eur J Nutr
PT
2017;56:843-52.
CE
[63] Patel PS, Forouhi NG, Kuijsten A, Schulze MB, van Woudenbergh GJ, Ardanaz E, et al. The prospective association between total and type of fish intake and type 2 diabetes in 8 European
AC
countries: EPIC-InterAct Study. Am J Clin Nutr 2012;95:1445-53.
24
ACCEPTED MANUSCRIPT 6. Figure legends Figure 1. Literature search and study selection process for inclusion in meta-analysis of fish consumption and risk of myocardial infarction. Figure 2 Relative risk of myocardial infarction for a 15 g/d increment in fish consumption.
T
Figure 3 Dose-response associations between fish consumption and risk of myocardial infarction
IP
(All studies).
CR
Figure 4 Dose-response associations between fish consumption and risk of myocardial infarction
AC
CE
PT
ED
M
AN
US
(Subgroup analyses). MI, myocardial infarction.
25
ACCEPTED MANUSCRIPT Table 1. General characteristics of the studies included in meta-analysis of fish consumption and risk of myocardial infarction. Author name, year
Study name, country
Age range/ mean age (y)
Follow-up duration (y)
Participants
Gender
Number of cases
Albert [28], 1998
Physicians’ Health Study, US
40 – 84/ 53
11
20,551
M
Incident MI: 737
Exposure assessment
Quality score (max. 9 points)
Adjustments
FFQ
9
Age, aspirin and beta-carotene treatment assignment, evidence of CVDs, BMI, smoking status, alcohol consumption, history of DM, HTN, and hypercholesterolemia, vigorous exercise, and vitamin E, vitamin C, and multivitamin use.
FFQ
8
Age, BMI, smoking habits, alcohol consumption, history of HTN, history of DM; history of hypercholesterolemia, family history of myocardial infarction before 60 years of age, profession, and quintile group for intake of n-3 fatty acids.
FFQ
8
Age, education, religion, SBP, serum cholesterol, number of cigarettes smoked per day, BMI, presence or absence of DM, presence or absence of electrocardiographic abnormalities, and daily intake of energy, cholesterol, saturated,
T P
I R
C S
Ascherio [29], 1995
40 – 75
Health Professionals Follow-up Study, US
6
44,895
D E
A
M
M
T P
E C
Daviglus [30], 1997
C A
Chicago Western Electric Study, US
40 –55/ 47
30
U N
Total MI: 811
Nonfatal MI: 554
1,822
M
Fatal MI: 293
26
ACCEPTED MANUSCRIPT
de Goede [31], 2010
20 – 65/ 42
Monitoring Project on Risk Factors for Chronic Diseases (MORGEN) study, Netherlands
11.5
21,342
M/W
FFQ
7
Age, gender, BMI, total energy intake, ethanol intake, cigarette smoking, social economic status, vitamin or mineral supplement use, use of drugs for HTN or hypercholesterolemia, family history of CVD, SFA, fruit, and vegetables.
C S
U N
A
Gammelmark [32], 2016
T P
I R
Nonfatal MI: 252
monounsaturated, and polyunsaturated fatty acids, total protein, carbohydrate, alcohol, iron, thiamine, riboflavin, niacin, vitamin C, beta carotene, and retinol.
Danish Diet, Cancer and Health study, Denmark
50-64
D E
17
54,904
M
Fatal MI: 64
M/W
Incident MI: 3028
FFQ
8
Age, smoking, BMI, waist circumference, physical activity, alcohol intake, educational level, menopausal status (women), history of DM, HTN, hypercholesterolemia, total energy intake, intake of fruits and vegetables and intake of nuts.
W
Nonfatal MI: 1029
FFQ
8
Age, time periods, smoking status, BMI, alcohol intake, menopausal status and postmenopausal hormone use, vigorous to moderate activity, number of times aspirin was used per week, multivitamin use, vitamin E supplement use and
T P
E C
Hu [33], 2002
C A
Nurses’ Health Study, US
34-59
16
84,688
27
ACCEPTED MANUSCRIPT history of HTN, hypercholesterolemia, DM.
Iso [34], 2006
Japan Public Health Center-Based (JPHC) Study Cohort I, Japan
4059/50
11-12
41,578
M/W
Total MI: 221
FFQ
8
Age, sex, cigarette smoking, alcohol intake, BMI, histories of HTN, and DM, medication use for hypercholesterolemia, education level, sports at leisure time, quintiles of dietary intake of fruits, vegetables, SFA, monounsaturated fat, n-6 polyunsaturated fat, cholesterol, and total energy, and PHC.
FFQ
7
Stratified by age at baseline and study centers, adjusted for sex, energy intake, alcohol intake, BMI, waist circumference, physical activity, educational attainment, smoking and prevalent DM.
FFQ
8
Age, gender, education, DM, current smoking, pack-years of smoking, tuna/other fish and fried fish/fish sandwich consumption, BMI, systolic blood pressure, LDL cholesterol, HDL cholesterol, triglycerides, C-reactive protein, and intake of
T P
I R
C S
Ku¨hn [35], 2013
European Prospective Investigation into Cancer and Nutrition (EPIC-Germany), Germany
3565/51
8.1
48,315
D E
M/W
A
M
Nonfatal MI: 488
T P
E C
Mozaffarian [36], 2003
C A
Cardiovascular Health Study, US
≥ 65
9.3
U N
Total MI: 605
Fatal MI: 117
3910
W/M
Nonfatal MI: 363
28
ACCEPTED MANUSCRIPT saturated fat, alcohol, beef/pork, fruits, and vegetables.
Yamagishi [37], 2008
Yuan [38], 2001
Japan Collaborative Cohort Study for Evaluation of Cancer Risk (JACC) Study, Japan
40-79 /56
12.7
57,972
M/W
Fatal MI: 329
FFQ
7
Age, gender, HTN and DM, smoking status, alcohol consumption, BMI, mental stress, walking, sports, education levels, total energy, and dietary intakes of cholesterol, saturated and n-6 polyunsaturated fatty acids, vegetables, and fruit.
8
Age, total energy intake, level of education, BMI, current smoker at recruitment, average no. of cigarettes smoked per day, no. of alcoholic drinks consumed per week, history of DM, and history of HTN.
T P
I R
45 – 64/ 55
Prospective epidemiologic study of diet and cancer, China
12
18,244
M
U N
A
D E
M
C S
Fatal MI: 113
Interview based on questionnaire
Abbreviations: BMI, body mass index; CVD, cardiovascular disease; DM, diabetes mellitus; FFQ, food frequency questionnaire; HDL, high-density lipoprotein; HTN, hypertension; LDL, low-density lipoprotein; MI, myocardial infarction; M, men; PHC, public health center; SBP, systolic blood pressure; SFA, saturated fatty acid; W, women.
T P
E C
C A
29
ACCEPTED MANUSCRIPT
n
RR (95%CI)
I2(%), Pheterogeneity
11
0.96 (0.94-0.99)
65, 0.002
Men
5
0.98 (0.94-1.02)
74, 0.004
Women
2
0.98 (0.93-1.01)
0, 1.00
Both
6
0.95 (0.90-1.00)
67, 0.009
US + Europe
8
0.97 (0.94-1.00)
65, 0.005
Asia
3
0.94 (0.91-0.97)
IP
Table 2. Relative risk of myocardial infarction for a 15 g/d increment in fish consumption.
6
0.96 (0.92-1.01)
5
0.96 (0.92-1.00)
< 500
6
0.92 (0.88-0.97)
57, 0.05
> 500
5
1.00 (0.98-1.01)
0, 0.56
0.97 (0.94-0.99)
61, 0.006
All studies
1
2
Pbetween -
Sex
≥ 12 years
AN
Number of cases
10
No
1
Yes
6
Fruit and vegetable intake
0.98 (0.96-1.00)
45, 0.12
5
0.92 (0.84-0.99)
81, < 0.001
5
0.95 (0.90-1.00)
73, 0.006
6
0.97 (0.93-1.00)
63, 0.02
Yes
7
0.93 (0.89-0.98)
72, 0.001
No
4
1.00 (0.97-1.02)
0, 0.72
Yes
PT
No
2
< 0.001
0.02
0.41
0.71
0.02
P-heterogeneity within subgroups with the use of a random-effects model. P-heterogeneity between subgroups with the use of a fixed-effects model.
AC
1
0.60
CE
Energy intake
64, 0.03
-
No
0.007
71, 0.005
0.84 (0.72-0.96)
ED
Physical activity
Yes
M
Adjustment for confounders Alcohol consumption
CR
< 12 years
0, 0.73
US
Follow-up duration
T
Geographical region
0.72
30
Figure 1
Figure 2
Figure 3
Figure 4