cancers

Systematic Review
Fish Consumption and Colorectal Cancer Risk: Meta-Analysis
of Prospective Epidemiological Studies and Review of Evidence
from Animal Studies
Saverio Caini 1 , Sofia Chioccioli 2 , Elisa Pastore 1, * , Miriam Fontana 1 , Katia Tortora 1 ,
Giovanna Caderni 2,† and Giovanna Masala 1,†

1 Institute for Cancer Research, Prevention and Clinical Network (ISPRO), Via Cosimo il Vecchio 2,
50139 Florence, Italy; s.caini@ispro.toscana.it (S.C.); m.fontana@ispro.toscana.it (M.F.);
k.tortora@ispro.toscana.it (K.T.); g.masala@ispro.toscana.it (G.M.)
2 NEUROFARBA Department, Pharmacology and Toxicology Section, University of Florence, Viale Pieraccini 6,
50139 Florence, Italy; sofia.chioccioli@unifi.it (S.C.); giovanna.caderni@unifi.it (G.C.)
* Correspondence: e.pastore@ispro.toscaa.it; Tel.: +39-055416942 (ext. 704)
† These authors contributed equally to this work.

Simple Summary: We meta-analyzed prospective epidemiological studies reporting on the asso-

ciation between fish consumption and colorectal cancer (CRC) risk among humans and reviewed
preclinical studies that examined the link between fish components and colorectal carcinogenesis
in animals. By pooling results from 25 studies (encompassing over 25,000 CRC cases) published
up to November 2020, we found convincing evidence that increased fish consumption may protect
from CRC development among humans. The review of animal studies allowed identifying several



biological mechanisms able to explain the associations that have emerged in human populations.

 Dietary recommendations for cancer prevention should incorporate the evidence from this literature
Citation: Caini, S.; Chioccioli, S.; review and meta-analysis.
Pastore, E.; Fontana, M.; Tortora, K.;
Caderni, G.; Masala, G. Fish Abstract: Background: Epidemiological studies on the association between fish consumption and
Consumption and Colorectal Cancer
colorectal cancer (CRC) risk have yielded inconsistent results, despite evidence from preclinical
Risk: Meta-Analysis of Prospective
studies that long-chain ω-3 polyunsaturated fatty acids inhibit colorectal carcinogenesis. We con-
Epidemiological Studies and Review
ducted a meta-analysis of prospective epidemiological studies investigating the association between
of Evidence from Animal Studies.
fish consumption and CRC risk among humans and reviewed studies examining the link between
Cancers 2022, 14, 640. https://
doi.org/10.3390/cancers14030640
fish components and colorectal carcinogenesis in animal models. Methods: We included studies
published until November 2020. We calculated the summary risk ratio (SRR) and 95% confidence
Academic Editor: Heike Allgayer
intervals (CI) through random effects meta-analysis models in order to summarize evidence from
Received: 17 December 2021 studies among humans. Results: Twenty-five prospective epidemiological studies encompassing
Accepted: 21 January 2022 25,777 CRC cases were included. Individuals in the highest (vs. lowest) category of fish consumption
Published: 27 January 2022 had a significantly reduced risk of CRC (SRR 0.94, 95%CI 0.89–0.99). In dose–response meta-analysis,
a 50-g increment in the daily consumption of fish was associated with a statistically significant 4%
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
reduction in CRC risk (SRR 0.96, 95%CI 0.92–0.99). Preclinical studies (n = 25) identified multiple
published maps and institutional affil- mechanisms of action of fish and fish components on colorectal carcinogenesis. Conclusions: Dietary
iations. recommendations for cancer prevention should take into account the evidence from epidemiological
and preclinical studies that increasing fish consumption may be effective in preventing CRC.

Keywords: fish consumption; colorectal cancer; epidemiological studies; animal studies; review;
Copyright: © 2022 by the authors. meta-analysis
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
1. Introduction
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
Colorectal cancer (CRC) is one of the most common and deadliest malignancies
4.0/). worldwide. According to the Global Cancer Observatory (GCO) data for 2020, it ranks

Cancers 2022, 14, 640. https://doi.org/10.3390/cancers14030640 https://www.mdpi.com/journal/cancers

Cancers 2022, 14, 640 2 of 20

third globally in terms of incidence, with an estimated 1,931,590 new cases, and second
in terms of mortality, causing 935,173 deaths [1]. CRC disease burden is generally higher
in high-income industrialized countries [1], clearly suggesting the importance of lifestyle
factors in its etiology.
The role of diet in the pathogenesis of CRC is now established, yet uncertainty remains
about some specific foods and food classes. In the latest (2018) report of the World Cancer
Research Fund (WCRF) on diet, nutrition, physical activity, and CRC [2], the Expert Panel
concluded there was strong evidence that red and processed meat, alcoholic drinks, and
body fatness increase CRC risk, while whole grains, foods containing dietary fiber, and
dairy products have a protective effect. The evidence suggesting that fish consumption
reduces CRC risk was judged as still limited, albeit generally consistent.
Fish is a key component of the Mediterranean diet, and its consumption is usually
actively encouraged as it contains many high-quality proteins and essential micronutrients
(vitamins and minerals) and fatty acids [3]. In particular, fish represents the main dietary
source of long-chain ω-3 polyunsaturated fatty acids (ω-3 PUFA), among which are eicos-
apentaenoic acid (EPA) and docosahexaenoic acid (DHA), which possess anti-inflammatory
and immunomodulatory properties [4]. Studies in animal models have consistently sug-
gested that fish oil (FO) containing ω-3 PUFA inhibits colorectal carcinogenesis through
several pathways [5,6], and a recently published meta-analysis confirmed these findings
among humans by highlighting an inverse relationship between ω-3 PUFA intake, blood
levels, and CRC risk [7]. Epidemiological studies on the association between fish consump-
tion and CRC risk in human populations did not always yield consistent results [8–11], and
previous meta-analyses aggregating the available evidence from prospective investigations
failed to reach conclusive results, highlighting the need for further research [12].
Recently, findings from large population-based cohort studies have been published [13,14];
hence, conducting an updated meta-analysis seems appropriate and timely. Moreover, we
believe that coupling the evidence from epidemiological studies with that from preclinical
studies may greatly help to jointly evaluate the robustness and plausibility of the hypothesis
under study. Therefore, we conducted an up-to-date meta-analysis of prospective studies
on the association between fish consumption and CRC risk among humans and a review
of the available evidence on the link between the intake of fish and fish components and
colorectal carcinogenesis in animal models.

2. Materials and Methods

2.1. Fish Consumption and Colorectal Cancer Risk among Humans
2.1.1. Literature Search and Inclusion Criteria
The literature search and statistical analysis were conducted according to the MOOSE
guidelines for the meta-analyses of observational studies [15] and the protocol was regis-
tered in the Prospective Register of Systematic Reviews (PROSPERO CRD42022299629). On
23 November 2020, we searched PUBMED/MEDLINE and EMBASE using the following
search string: (fish OR seafood* OR shellfish) AND (colon OR rectal OR rectum OR colorec-
tal) AND (cancer OR carcinoma OR tumour OR tumor OR malignancy). After removing
duplicates, papers were initially screened based on their title and abstract; those that were
considered as potentially of interest for the present meta-analysis were then obtained and
read in full text to decide on their eligibility. Additional papers were searched by the
backward citation chaining of eligible papers and previously published literature reviews
and meta-analysis. Papers were eligible if they had a prospective study design (cohort,
nested case-control studies (NCC), and randomized clinical trials (RCT)) and reported
a measure of relative risk (RR) (e.g., hazard ratio or odds ratio), a corresponding 95%
confidence intervals (CI), or a different measure of statistical uncertainty (e.g., standard
errors, variance, or exact p-value) for the association between the consumption of fish and
the risk of colon, rectal, or colorectal cancer. Retrospective case–control studies were not
considered, as well as ecological studies, case reports, editorials/letters that did not present
original data, and studies with any design in which the outcome of interest was cancer
Cancers 2022, 14, 640 3 of 20

mortality instead of cancer risk. In the case of non-independent studies (with total overlap,
e.g., two or more reports from the same cohort, or with partial overlap, e.g., one study
reporting from a cohort consortium and a different study presenting results for a cohort that
is part of it), the older/smaller study was discarded unless it provided RR estimates that
were not available in a more recent/larger study. Two researchers independently decided
on the inclusion of each paper and any disagreement was resolved by consensus.
As illustrated in more detail below (see Results), the type of food studied in relation to
CRC risk varied across articles in that it could be limited to fish (sometimes left unspecified)
or encompass a varying combination of fish and other seafoods including crustaceans and
mollusks/shellfish. For the sake of brevity, the generic term “fish” will be used henceforth
to refer to the area of interest.

2.1.2. Data Extraction and Statistical Analysis

The following information was extracted from all eligible papers: first author and year
of publication; country in which the study was conducted; study design; source; sex and age
distribution of the whole study population and number of cancer cases (for cohort studies
and RCT), or of cancer cases and controls (for NCC); details of matching, if any, for NCC;
methods for data collection on food consumption; mean/median and maximum duration
of follow-up and total number of person–years accrued; distribution of cancer cases by site
(colon further split into proximal and distal colon, rectum, or other, and overlapping sites);
adjusted (whenever available) or unadjusted RRs and their 95% CI, alongside the number
of cases and non-cases/controls and the total number of person–years in each category of
exposure (e.g., each quantile of fish consumption); details on the statistical methods and
the variables used for adjustment.
RR estimates and corresponding 95% confidence intervals were transformed into log
relative risk (logRR) and corresponding variance as proposed by Greenland [16]. We then
conducted highest versus lowest and linear and non-linear dose–response random effects
meta-analyses to summarize the evidence on the association between fish consumption
and CRC risk. For linear dose–response meta-analysis, we followed the method described
by Greenland and Longnecker [17] to compute a trend from RRs and 95% CI calculated
across categories (e.g., quantiles) of exposure, when this was not directly provided in the
paper. The non-linear dose–response meta-analysis was conducted using 3-knots restricted
cubic splines (5th, 50th, and 95th percentiles of the distribution). Between studies, the
heterogeneity of RR estimates was quantified using the I2 statistics, which is a measure of
the total variation of effects across studies that can be attributed to actual heterogeneity
rather than chance [18]. When I2 was above 50% (which denotes large heterogeneity)
and the sample size allowed, we conducted subgroup analyses and fitted meta-regression
models (for categorical and continuous variables, respectively) to determine what study
characteristics could explain a significant share of the observed heterogeneity. Analyses
stratified by sex, cancer site (colon vs. rectum), and country were, however, planned a
priori and conducted regardless of whether or not there was substantial heterogeneity.
Publication bias was assessed by visually inspecting the funnel plot and applying the
Egger’s test [19]. Finally, the methodological quality and proneness to bias of all included
studies was rated using the Newcastle–Ottawa Scale (NOS) [20].
Statistical analyses were conducted using Stata (version 14, command metan, for
highest versus lowest meta-analysis) and R (version 4.0.0, command dosresmeta [21] for
dose–response meta-analysis) software.

2.2. Fish and Fish-Oil Intake and Colorectal Carcinogenesis in Animal Models
The literature search for preclinical studies focusing on the relationship between the
intake of fish, FO intake, and colon carcinogenesis in animal models was conducted in
PUBMED/MEDLINE using the following search string: (chemoprevention OR omega3 OR
“omega 3” OR omega-3 OR diet OR fish) AND (OR colon OR colorectal OR crc OR intestinal
OR bowel) AND (cancer* OR carcinogenesis) AND (rodent* OR rat* OR mouse OR mice).
Cancers 2022, 14, 640 4 of 20

The search was conducted on 3 May 2021, and it extends, therefore, to all the articles that
were published until April 2021. To be included in the final review, an article needed to
report original data and examine the effect of fish and/or fish oil intake in the process
of colorectal carcinogenesis and metastasis in rodent models. The article selection was
conducted according to the same procedures described above; no formal meta-analysis was
conducted because of the great diversity across studies in terms of several characteristics
(see Section 3).

3. Results
3.1. Fish Consumption and Colorectal Cancer Risk among Humans
The literature search returned 3200 non-duplicate entries, and 44 were added via
backward reference searching (Figure 1).

Articles identified through database

searching (Medline and Embase): Additional articles identified through other
sources: IDENTIFICATION
n = 3200
n = 44
(after removing duplicates)

Articles screened

n = 3244

Articles excluded after title screening:

n = 2873
SCREENING
Articles excluded after abstract screening:

n = 131

Full-text articles assessed for

elegibility:

n = 240

Full-text articles excluded, with reasons:

ELIGIBILITY
- other study design: n = 52
- review, meta-analysis: n = 51
- exposure other than fish/seafood: n = 59
- outcome other than CRC risk: n = 25
- non-independent studies: n = 15
- other: n = 13

Articles included in the systematic

review and meta-analysis:
INCLUDED
n = 25

Figure 1. Flow-chart for the selection of articles included in the systematic review and meta-analysis
on the association between fish consumption and colorectal cancer risk.

A total of 2873 and 131 papers were screened out based on their titles and abstract,
and the remaining 240 papers were read in full copy. Finally, 25 articles reporting on the
association between fish consumption and CRC risk met the inclusion criteria for the present
Cancers 2022, 14, 640 5 of 20

meta-analysis [8–11,13,14,22–40] The papers by Bamia et al. [22] and Engeset et al. [23]
were totally and partially overlapping, respectively, with the larger and more recent paper
by Aglago et al. [13], which reported from the large, multi-country European Prospective
Investigation into Cancer and Nutrition (EPIC) cohort study. Likewise, the two papers by
Murff et al. [24] and Lee et al. [11] were based on the same prospective cohort (Shanghai
Women’s Health Study). The study by Spencer et al. [25], a case–control study nested within
a consortium of UK-based dietary cohorts including the EPIC-Norfolk and the EPIC-Oxford
cohorts, had only a limited overlap with that by Aglago et al. [13], to which the two English
cohorts contributed around 15.5% of the total size. Since the overlap between the two
papers was limited, and in order not to discard valuable information, both papers were
included in the meta-analysis and treated as if they were completely independent.
The main characteristics of included papers are summarized in Table 1.
The papers were published between 1994 and 2020 and reported on studies conducted
in Europe (n = 12), the USA (n = 7), Asia (n = 5), and Australia (n = 1). In terms of design,
all were prospective cohort studies, except for Spencer et al. [25], Siezen et al. [31], and
Tiemersma et al. [36], which were nested case–control studies, and Pietinen et al., whose
study population consisted of subjects previously enrolled in a randomized controlled trial
testing the effect of alpha-tocopherol and beta-carotene supplementation on lung cancer risk
among male smokers [38]. The 25 studies included 2,228,377 individuals altogether, aged
between 15 and 99 years at cohort inception, among which a total of 25,777 CRC cases were
diagnosed during an average follow-up that varied between 4.8 and 14.9 across the studies.
Seven and two studies included only women and men, respectively; the proportion of
women ranged between 40.3% and 70.1% in the remaining sixteen papers. The breakdown
of CRC cases into subsites was fully specified in seventeen studies; the studies by Gaard
et al. and Bostick et al. included only colon cancer cases [39,40]. Information on food
consumption was collected using food-frequency questionnaires in all studies except in
Spencer et al. where 4-to-7-day diet diaries were used [25]. All included studies were of
fair to very good quality, with the scores (assigned according to the NOS tool) ranging from
the maximum allowed of 9 (for nine articles) to a minimum of 6 (for two articles) (Table S1).
The studies varied widely both in the way the exposure of interest was defined (e.g.,
(total) fish, fish and shellfish, (fish and) seafoods, etc.) and in the availability of risk
estimates for specific subtypes of fish (e.g., fatty vs. lean, oily vs. non-oily, marine vs.
fresh water, etc.) (Table S2). Dietary consumptions were reported by using as a unit of
measure either the daily intake (mostly g/day) or the frequency of consumption (e.g., times
or servings per week or month) (Table 1).
In meta-analysis, individuals in the highest (vs. lowest) category of fish consumption
had a mild yet statistically significant reduction in CRC risk (SRR 0.94, 95% CI 0.89–0.99,
p-value 0.023), based on data from 22 independent studies, with negligible heterogeneity
(I2 = 11.7%) and no evidence of publication bias (Egger’s test p-value = 0.955) (Tables 2 and S3
and Figure 2).
A comparable (in direction and strength) association emerged in stratified analyses
(expect sex-stratified SRR for colon cancer risk, which was 1.02 among both women and
men) but statistical significance was not achieved because of the lower number of available
risk estimates (Table 2). In detail, the reduction in CRC risk for those in the highest
category of fish consumption was slightly more evident among men (SRR 0.91) than
women (SRR 0.95), while not differing by cancer site (SRR 0.94 for both colon and rectum)
(Tables S4–S11) (Figures S1–S8). The association differed significantly (p-value 0.007) by
country, as the SRR was 0.91 (95% CI 0.82–1.00) for studies conducted in North America,
(n = 7), 0.90 (95% CI 0.84–0.97) for studies conducted in Europe (n = 10), and 1.12 (95%
CI 0.98–1.27) in Asian studies (n = 4).
In dose–response meta-analysis, a 50-g increment in the daily consumption of fish
was associated with a statistically significant 4% reduction in CRC risk (SRR 0.96, 95%
CI 0.92–0.99, p-value 0.021), based on data from seven independent studies, with no evi-
dence of heterogeneity between studies (I2 = 0.0%) (Table 3).
Cancers 2022, 14, 640 6 of 20

Table 1. Main characteristics of articles included in the systematic review and meta-analysis on the association between fish consumption and colorectal cancer risk.

First Author, Age Range (at Study Follow-Up Fish Consumption Anatomic Site Distribution of Colorectal Cancers
Country Study Name Study Size (a) Men (%) No. Cases
Year Cohort Inception) Length Time (Years) Assessment (b) Colon Rectal Ns or Other
European
Aglago, 2020 Prospective
Europe Investigation into 476,160 29.9% 25–70 1992–ns median 14.9 intake + frequency 6291 66.7% 33.3% 0.0%
[13] (c) Cancer and
Nutrition (EPIC)
Bradbury, 2020 UK biobank
UK 475,581 46.0% 40–69 2006–2014 mean 5.7 intake + frequency 2609 66.8% 33.2% 0.0%
[14] study
Kantor, 2014 VITamins And
USA 68,109 49.0% 50–70 2001–2008 mean 6.7 frequency 488 63.7% 26.3% 10.0%
[26] Lifestyle cohort
Nurses’ Health
76,386 0.0% 30–55 1976–2010 1469 76.9% 21.1% 2.0%
Study
Song, 2014 [27] USA mean 14.8 intake
Health
Professionals 47,143 100.0% 40–75 1986–2010 987 65.2% 21.8% 13.0%
Follow-up Study
European
Prospective
Bamia, 2013
Europe Investigation into 480,308 30.0% 25–70 1992–2010 mean 11.6 intake 4355 63.2% 36.8% 0.0%
[22] (c)
Cancer and
Nutrition
National
Institutes of
Daniel, 2011 Health
USA 492,186 59.6% 50–71 1995–2006 mean 9.1 intake 7143 71.3% 26.4% 2.3%
[28] (NIH)-AARP
Diet and Health
Study
UK Dietary
Spencer, 2010
UK Cohort 2575 45.9% ns 1985–2006 ns intake 579 ns ns 100.0%
[25]
Consortium
Shanghai
Murff, 2009
China Women’s Health 73,243 0.0% 40–70 1996–2007 ns intake 396 ns ns 100.0%
[24] (d)
Study
Shanghai
Lee, 2009 [11] China Women’s Health 74,942 0.0% 45–65 1997–2005 mean 7.4 intake 394 59.9% 40.1% 0.0%
(d)
Study
Ohsaki National
Sugawara,
Japan Health Insurance 39,498 47.7% 40–79 1995–2003 mean 7.7 intake 566 59.5% 40.5% 0.0%
2009 [29]
Cohort Study
Cancers 2022, 14, 640 7 of 20

Table 1. Cont.

First Author, Age Range (at Study Follow-Up Fish Consumption Anatomic Site Distribution of Colorectal Cancers
Country Study Name Study Size (a) Men (%) No. Cases
Year Cohort Inception) Length Time (Years) Assessment (b) Colon Rectal Ns or Other
Singapore
Butler, 2008
Singapore Chinese Health 61,321 ns 45–74 1993–ns mean 9.8 intake 961 61.5% 38.5% 0.0%
[30]
Study
Physicians’
Hall, 2008 [10] USA 21,406 100.0% ns 1982–2006 ns frequency 500 77.6% 22.4% 0.0%
Health Study

Engeset, 2007 The Norwegian

Norway Women and 64,168 0.0% 40–71 1991–2004 ns intake 254 ns ns 100.0%
[23] (c) Cancer
Monitoring
Project on
Cardiovascular
Siezen, 2006 The Disease Risk
160 ns ns 1976–2003 ns frequency 160 ns ns 100.0%
[31] Netherlands Factors +
Diagnostisch
Onderzoek Mam-
macarcinoom
Brink, 2005 The The Netherlands
2948 48.4% 55–69 1986–1993 mean 5.0 intake 608 73.7% 26.3% 0.0%
[32] Netherlands Cohort Study
The Swedish
Larsson, 2005
Sweden Mammography 61,433 0.0% 40–75 1987–2003 mean 13.9 frequency 733 53.1% 31.4% 15.5%
[33]
Cohort
The Melbourne
English, 2004
Australia Collaborative 37,112 39.5% 27–75 1990–2003 mean 9.0 frequency 451 62.5% 37.3% 0.2%
[34]
Cohort Study
The Japan Public
Health
Kobayashi,
Japan Center-based 116,194 49.3% 40–69 1990–1999 ns intake 705 ns ns 100.0%
2004 [9]
prospective
study
Women’s Health
Lin, 2004 [35] USA 37,547 0.0% 45–ns 1993–2003 mean 8.7 frequency 202 80.2% 19.8% 0.0%
Study
Monitoring
Project on
Tiemersma, The
Cardiovascular 102 54.9% 20–59 1987–1998 mean 8.5 frequency 102 61.8% ns ns
2002 [36] Netherlands
Disease Risk
Factors
Cancers 2022, 14, 640 8 of 20

Table 1. Cont.

First Author, Age Range (at Study Follow-Up Fish Consumption Anatomic Site Distribution of Colorectal Cancers
Country Study Name Study Size (a) Men (%) No. Cases
Year Cohort Inception) Length Time (Years) Assessment (b) Colon Rectal Ns or Other
Cohort
assembled within
the Mobile
Knekt, 1999
Finland Health Clinic of 9985 52.8% 15–99 1967–1990 ns intake 189 38.6% ns 61.4%
[37]
the Social
Insurance
Institution
The Alpha-
Tocopherol,
Pietinen, 1999
Finland Beta-Carotene 27,111 100.0% 50–69 1985–1995 mean 8.0 intake 185 ns ns ns
[38]
Cancer
Prevention Study
New York
University
Kato, 1997 [8] USA 14,727 0.0% 34–65 1985–1994 mean 7.1 frequency 100 84.0% 16.0% 0.0%
Women’s Health
Study
Cohort
assembled within
Gaard, 1996
Norway the Norwegian 50,535 50.1% 20–54 1977–ns mean 11.4 frequency 143 100.0% 0.0% 0.0%
[39]
National Health
Screening Service
Bostick, 1994 Iowa Women’s
USA 35,216 0.0% 55–69 1986–1990 mean 4.8 frequency 212 100.0% 0.0% 0.0%
[40] Health Study
(a)Cohort size or (for nested case–control studies) number of CRC cases. (b)
Food intake was ex-pressed in g/day except in Daniel et al. 2001 (g/1000 kcal/day). Frequency was expressed as times, servings, or meals per week
or month depending on the study. (c) Bamia et al. [22] was based on a shorter follow-up of the EPIC study than in Aglago et al. [13]. Engeset et al. [23] was based on one of the country-specific EPIC cohorts. From the studies
by Bamia et al. and Engeset et al. we con-sidered only the results not available in Aglago et al. (d) The results from Murff et al. [24] were used only when they were not available in Lee et al. [11], which was based on the
same cohort.
Cancers 2022, 14, 640 9 of 20

Cancers 2022, 14, x FOR PEER REVIEW 10 of 21

Table 2. Meta-analysis of studies for the association between fish consumption (highest vs. lowest
category of consumption) and colorectal cancer risk, overall and stratified by tumor site and sex.

(I2 = 11.7%) and no evidence of publication bias (Egger’s testUpper

Lower p-value = 0.955) (Tables 2 and 2
N Studies SRR p-Value I
S3 and Figure 2). 95% CI 95% CI
Colorectal
22 of studies for
Table 2. Meta-analysis 0.94 0.89
the association 0.99
between fish consumption 0.023 vs. lowest
(highest 11.7%
cancer
category of consumption) and colorectal cancer risk, overall and stratified by tumor site and sex.
women 13 0.95 0.87 1.05 0.310 2 13.9%
N Studies SRR Lower 95% CI Upper 95% CI p-Value
men 10 0.91 0.82 1.01 0.088 I 27.2%
Colorectal cancer 22 0.94 0.89 0.99 0.023 11.7%
women
Colon
13 0.95 0.87 1.05
15 0.94 0.88 1.010.310 0.08913.9% 12.9%
cancer
men 10 0.91 0.82 1.01 0.088 27.2%
Colon cancer women
15 7
0.94 1.02
0.88 0.88
1.01 1.190.089 0.76312.9% 12.5%
women men
7 4
1.02 1.02
0.88 0.85
1.19 1.220.763 0.86612.5% 0.0%
men 4
Rectal 1.02 0.85 1.22 0.866 0.0%
13 0.94 0.87 1.03 0.173 0.0%
Rectal cancer cancer
13 0.94 0.87 1.03 0.173 0.0%
women 5
women 0.99
5 0.79
0.99 1.26
0.79 1.260.966 0.9660.0% 0.0%
men 3
men 0.90
3 0.57
0.90 1.43
0.57 1.430.663 0.66363.2% 63.2%
SRR:
SRR: summary
summary relative
relative risk.risk.
CI: CI: confidence
confidence intervals.
intervals.

Figure
Figure 2. 2. Forestplot
Forest plotfor
for the
the association
associationbetween
betweenfishfish
consumption (highest
consumption vs. lowest
(highest vs. category of
lowest category of
consumption)
consumption) andandcolorectal
colorectalcancer
cancer risk.
risk. RR:
RR: relative
relativerisk.
risk.SRR:
SRR:summary
summary relative risk.
relative CI: CI:
risk. confi-
confidence
dence intervals. W: RR among women. M: RR among men. C: RR for colon cancer. R: RR for rectal
intervals.
cancer.
W: RR among women. M: RR among men. C: RR for colon cancer. R: RR for rectal cancer.

Results were similar

A comparable in analyses
(in direction stratified
and strength) by sex oremerged
association cancer site (Table 3).
in stratified The results of
analyses
the(expect
dose–response meta-analysis
sex-stratified SRR for colonwere
cancerheavily influenced
risk, which was 1.02 by the single
among studyand
both women of Aglago
et al.
men)[13],
butwhose percentage
statistical significancestudy
was notweight was
achieved 66.5%ofinthethe
because main
lower analysis
number and ranged
of availa-
between
ble risk59.5% and(Table
estimates 73.4%2).
inInstratified
detail, theanalysis.
reduction The p-value
in CRC riskforfor deviation
those in thefrom linearity of
highest
thecategory of fish consumption
dose–response association was was nearly
slightlysignificant
more evident among
(0.071) onlymen for (SRR 0.91)analysis.
the main than The
women (SRR 0.95), while not differing by cancer site (SRR 0.94 for both colon
visual inspection of the graph obtained by fitting a non-linear dose–response meta-analysis and rectum)
revealed that the decrease in CRC risk was observed only for the daily consumption of fish
increasing up to 50 g/day and appeared to level off (albeit confidence intervals became
increasingly wide because of limited data availability) above this value (Figure 3).
Cancers 2022, 14, 640 10 of 20

Table 3. Dose–response meta-analysis of studies for the association between fish consumption (linear
increment by 50 g/day) and colorectal cancer risk, overall and stratified by tumor site and sex.

SRR (for an Increase Lower 95% Upper 95% p-Value for

N Studies p-Value I2
by 50 g/day) CI CI Non-Linearity
Colorectal
7 0.96 0.92 0.99 0.021 0.0% 0.071
cancer
women 5 0.95 0.90 1.01 0.078 0.0% 0.880
men 6 0.97 0.92 1.02 0.188 0.0% 0.280
Colon
6 0.96 0.92 1.01 0.140 0.0% 0.991
cancer
Rectal
6 0.95 0.89 1.02 0.174 0.0% 0.248
cancer
SRR: summary relative risk. CI: confidence intervals.

3.2. Fish and Fish-Oil Intake and Colorectal Carcinogenesis in Animal Models
The literature search resulted in 281 entries altogether: upon removing duplicates and
checking inclusion criteria, a total of 25 articles were finally included in the review. The
majority of preclinical experiments with rodents dealt with the effects of FO or pure ω-3
polyunsaturated fatty acids (PUFA) on colorectal carcinogenesis, while only a few studies
were conducted by administering fish meat to the animals.
In 1986, Reddy and Maruyama, using an experimental model in which colorectal car-
cinogenesis was chemically induced (with 1,2-dimethylhydrazine (DMH) or its metabolite
azoxymethane (AOM)), first documented in rats that a diet containing FO as a source of fat
Cancers 2022, 14, x FOR PEER REVIEW 12 of 21
(Menhaden oil) reduces colon tumorigenesis when compared with a diet containing the
same amount of fat as corn oil (CO) which is rich in ω-6 PUFA [41] (Tables 4 and S11).

Non-lineardose–response
Figure 3.3. Non-linear
Figure dose–responsemeta-analysis
meta-analysis
forfor
thethe association
association between
between fishfish consumption
consumption
(g/day)
(g/day)andandcolorectal
colorectalcancer risk.
cancer SRR:
risk. summary
SRR: summaryrelative risk.
relative risk.

3.2. Fish and Fish-Oil Intake and Colorectal Carcinogenesis in Animal Models
The literature search resulted in 281 entries altogether: upon removing duplicates
and checking inclusion criteria, a total of 25 articles were finally included in the review.
The majority of preclinical experiments with rodents dealt with the effects of FO or pure
Cancers 2022, 14, 640 11 of 20

Table 4. Effects of fish or its components on different animal models of colorectal cancer.

Dietary Intervention and

Experimental Model Effects on Carcinogenesis Reference
Duration of Treatment
Diets containing Menhaden or Reduction in CRC in rats fed
AOM-induced rats Reddy, 1986 [41]
Corn oils Menhaden Oil.
Different levels of FO or CO in Reduction in CRC in rats fed
-+AOM-induced rats Reddy, 1988 [42]
the diet. high FO.
Reduction in colon
AOM-induced rats Diets containing EPA or LA carcinogenesis in the EPA Minoura, 1988 [43]
group.
Reduction in dysplastic areas
Commercial preparations of
AOM-induced mice and carcinogenesis in the FO Deschner, 1990 [44]
FO or CO.
group.
Reduction in colon
Diets containing FO and/or
AOM-induced rats carcinogenesis in rats fed high Reddy, 1991 [45]
CO.
FO.
DHA (0.7 mL by gavage)
DMH-induced rats Reduction in ACF. Takahashi, 1993 [46]
5 times a wk.
Slight non-significant
AOM-induced rats DHA (0.7 mL of by gavage) reduction of CRC. Takahashi, 1994 [47]
twice a wk.
Significant reduction in ACF.
DHA (1 mL by gavage) Significant reduction in ACF
AOM-induced rats Takahashi, 1997a [48]
5 times a wk. and in CRC.
2-amino-1-methyl-6-
DHA (1 mL by gavage)
phenylimidazo [4,5-b]pyridine Reduction in ACF. Takahashi, 1997b [49]
5 times a wk.
(PhIP)-induced rats
Diet supplemented with FO or
AOM-induced rats CO; cellulose or pectin also FO reduced CRC. Chang, 1998 [50]
tested.
Reduction in carcinogenesis
AOM-induced rats FO vs. HFML. Rao, 2001 [51]
and ACF in rats fed FO.
Reduction in SI polyps, only
Apc∆716 mice of both sexes Diet with 3% DHA for 7 wks. Oshima, 1995 [52]
in female mice fed DHA.
Diet containing a commercial Reduction in SI tumors in
Min mice mutated in Apc Paulsen, 1997 [53]
FO preparation. treated mice.
Diets containing EPA (1.5%) Reduction in SI tumors in the
Min mice mutated in Apc Petrik, 2000 [54]
or different PUFA for 8 wks. EPA group.
Reduction in polyp number in
Min mice mutated in Apc EPA-FFA fed for 12 wks. Fini, 2011 [55]
SI and colon.
EPA-FFA in the diet tested in
Mice treated with AOM/DSS
the initiation and Reduction in tumorigenesis Piazzi, 2014 [56]
to induce CAC
post-initiation phases.
Salmon compared with beef
Salmon muscle decreased
A/J Min/+ mice of both sexes or chicken muscles fed from Steppler, 2017 [57]
tumor load and size in the SI.
weaning for 10 wks.
Mice treated with AOM-DSS Tuna muscle extract rich in
Reduction in carcinogenesis. Masuda, 2018 [58]
to induce CAC Selenoeine fed for 14 wks.

The same authors found that diets containing high levels of FO and low levels of CO
were associated with fewer AOM-induced colon cancers compared with diets containing
only CO [42]. Later investigations confirmed that Menhaden oil is effective in reducing both
the initiation and post-initiation stages of carcinogenesis when compared to ω-6 PUFA-rich
Cancers 2022, 14, 640 12 of 20

CO [44,45]. Studies were conducted in parallel in which pure ω-3 PUFA was administered
instead of not-purified FO. Minoura et al. found that rats fed with eicosapentaenoic acid
(EPA) had lower colon carcinogenesis and lower prostaglandin E2 (PGE2) levels than rats
on a linoleic acid (LA) diet [43]. Takahashi et al. reported fewer DMH-induced microscopic
preneoplastic lesions (aberrant crypt foci (ACF)) and protection against AOM-induced
carcinogenesis among rats fed with pure docosahexaenoic acid (DHA) [46–49]. Subsequent
studies confirmed the role of FO in preventing AOM-induced colon carcinogenesis through
enhanced cell differentiation and apoptosis [50], particularly when compared to diets
containing a mixture of lipids simulating the Western diet (lipids derived from saturated
fats, peanut and corn oils) [51] and suggested that the effect could be mediated via reduced
cyclooxygenase-2 (COX-2) activity and prostaglandin production.
A beneficial effect of DHA was also observed in genetic models of intestinal car-
cinogenesis (Apc∆716 mice) [52] (Table 4 and Table S11). Likewise, an ω-3 PUFA ethyl
ester-enriched fish oil concentrate was shown to oppose tumor growth in the small intestine
(but not in the colon) in Apc-Min mice [53], and subsequent studies suggested that the
protective effect could be mediated via antagonism with the production of arachidonic
acid (a precursor of eicosanoids such as PGE2) [54]. More recently, highly purified EPA as
free fatty acid (EPA-FFA) was observed to reduce COX-2 expression and cell proliferation
and, eventually, prevent intestinal carcinogenesis in Apc-Min mice [55]. The potential of
EPA-FFA in preventing colon carcinogenesis was later confirmed in a preclinical model
mimicking CRC arising in the setting of inflammatory bowel disease (colitis-associated
colorectal cancer (CAC)) [56].
Regarding colon cancer metastasis, Menhaden oil, in 1989, was first shown to be able
to suppress the growth of a transplanted colon cancer cell line (CT-26 cells) and inhibit
pulmonary colonization [59] (Table 5 and Table S13).

Table 5. Effects of fish or its components on different animal models of colorectal cancer metastasis.

Dietary Intervention and Effects on Colon Cancer and

Experimental Model Reference
Duration of Treatment Metastasis
Diets containing fish or safflower
Mice inoculated with Fish oil (FO) reduced tumor growth
oils fed for 30 days before and Cannizzo, 1989 [59]
CT-26 tumor cells and pulmonary colonization.
after CT-26 transplantation.
Diets containing EPA and DHA
Mice s.c. implanted with Inhibition of tumor growth and
during and after cancer cell Iigo, 1997 [60]
Co 26 Lu tumor cells decrease in lung metastatic nodules.
transplantation.
EPA, DHA, and OA reduced
Diets containing EPA, DHA, LA,
metastasis. Tumor cells treated with
Mice s.c. implanted with or oleic acid (OA) from day 5 for a
DHA showed a very low potential Suzuki, 1997 [61]
Co 26 Lu tumor cells total of 3 wks after cell
for lung colony formation when
implantation.
injected i.v.
Diets containing EPA, LA, or PA. EPA reduced metastatic foci in liver.
Rats inoculated with
ACL-15 tumor cells inoculated at EPA and PA diet groups had Iwamoto,1998 [62]
ACL-15 tumor cells
6 wks and rats sacrificed at 9 wks. smaller liver metastatic foci.
FO increased metastasis (number
Diets containing FO or safflower
Rats injected (via portal and size) at 1 wk after implantation;
oil for 3 wks before CC531
vein) with CC531 tumor both FO and the safflower oil diets Griffini, 1998 [63]
inoculation until sacrifice after
cells increased metastasis 3 wks after
1 or 3 wks.
implantation.
Rats injected (into the Diets containing PUFA or coconut
Reduction in tumor growth in the
spleen) with CC531 tumor oil in the diet 3 days before and Gutt, 2007 [64]
PUFA group.
cells 28 days after CC 531 injection.
Mice injected (into the Diet containing EPA-FFA. Feeding
EPA-FFA administration caused a
spleen) with MC-26 tumor 2 wks before and after cell Hawcroft, 2012 [65]
reduced MC-26 liver tumor burden.
cells injection.
Cancers 2022, 14, 640 13 of 20

Similar results were seen in mice transplanted with a highly metastatic murine colon
carcinoma cell line (Co 26 Lu) and fed with EPA and DHA [60], which were also effective
in decreasing the number of lung metastasis via reduced metalloproteinase-2 and -9 activ-
ity [61]. Furthermore, EPA was also reported to inhibit the formation of liver metastatic
foci of previously injected ACL-15 rat colon cancer cells by decreasing tumor cell prolifera-
tion and adhesion to the capillary bed [62]. Despite initial findings documenting adverse
effects [63], ω-3 PUFA were reported to be able to prevent the development of colon cancer
metastases in the liver of Wag-Rij rats transplanted with an established colon cancer cell
line [64]. Subsequent experiments using EPA-FFA in the diet reported largely consistent
results [65] and identified, among the possible mechanisms of action, a shift from PGE2 to
PGE3 in tumor cells, as well as a reduced ERK signaling at the invasive edge of tumors.
As already mentioned, only a few studies documented the effect of edible parts of
fish on colon carcinogenesis (Table 4). The effect of salmon muscle was tested by Steppler
et al. in A/J Min/+ mice, Apc mutated animals showing tumorigenesis not only in the
small intestine but also in the colon [66]. Salmon-fed animals showed a slightly lower
intestinal carcinogenesis when compared to a standard diet and to diets containing meat
from terrestrial animals [57]. More recently, dietary supplementation with tuna muscle
extract rich in selenoeine (an Se-protein with antioxidant activity present in the blood and
muscle of tuna) was reported to decrease colon carcinogenesis in a colitis-associated model
of CRC in mice [58].

4. Discussion
We conducted an up-to-date meta-analysis aiming to summarize the most recent
available evidence about the association between fish consumption and CRC risk among
humans. Twenty-five studies were included that encompassed a total of over 25,000 CRC
cases arising from over 2.2 million individuals. The most important finding was the mild
yet statistically significant reduction in CRC risk for individuals in the highest (vs. lowest)
category of fish consumption, a finding that was made particularly trustworthy by the fair
consistency of risk estimates across the studies and by the lack of evidence of publication
bias. Analyses stratified by subsite showed that the effect was similar for colon and rectal
cancer, the lower number of available studies for subsite-specific analyses being, therefore,
the most likely explanation for the failure to achieve full statistical significance. Analyses
stratified by sex also yielded comparable results to the main analysis in terms of both the
direction and strength of the association, while geographical variability existed, whereby
significant results emerged only in studies conducted in Europe and North America.
The latter finding may partly be due to chance (only four of the included studies were
conducted in Asia), but geographical diversity in terms of the genetic background and
cooking methods of populations may also play a role; therefore, more research on the topic
would be desirable to disentangle the importance of the different factors potentially at play.
Finally, while based on only seven independent studies, the dose–response analysis showed
a 4% decrease in CRC risk associated with a 50-g daily increase in fish consumption.
We then carried out a review of preclinical studies that reported on the link between
the intake of fish and fish oil and colon carcinogenesis in animal models. While the
diversity across the studies in terms of experimental methods advised against applying a
formal meta-analytical approach to summarize the results, the reviewed studies were fairly
consistent in suggesting a beneficial effect of fish- and fish oil-based diets against colorectal
carcinogenesis and dissemination in rodents and identified several biological mechanisms
potentially able to underlie this effect.
The preventive effect of fish consumption on CRC risk observed in epidemiological
studies can recognize several possible explanations. Partly, the association may be at-
tributable to a replacement effect since those who eat more fish generally eat less red meat,
whose causal link with colorectal carcinogenesis is well-known. Additionally, preferring
fish instead of meat may be part of a generally healthier lifestyle encompassing other habits
effective in preventing cancer [67,68]. However, prospective epidemiological studies are
Cancers 2022, 14, 640 14 of 20

fairly consistent in detecting a reduction in CRC risk associated with fish consumption,
which is, therefore, difficult to dismiss as mostly due to bias. In this regard, preclinical stud-
ies using animal models have been instrumental in helping to understanding the process
of colorectal carcinogenesis and to give insights on how this may be affected by intrinsic
and external factors such as diet, thus, providing support to the evidence stemming from
epidemiological studies. Among animal models, both chemically-induced carcinogenesis
and genetic models (the latter, mostly based on Apc mutations in mice and rats [69]) have
been used to test the effect of dietary treatments, which are typically administered to the
animals after the chemical induction of tumorigenesis or, in genetic models, starting from
weaning. Depending on the duration of dietary treatment, either precursor lesions or
straightforward tumors were used as a primary end-point of treatment efficacy. As stated
before, the majority of animal experiments were carried out testing the effect of FO or its
pure components (e.g., EPA and DHA) on colon carcinogenesis, while fish muscle consump-
tion has not been thoroughly tested. This underlies the need to conduct studies in which
animals are fed with fish, possibly using as a control group animals fed the same number
of proteins and fat deriving from other animal muscles (i.e., beef or chicken) to resemble an
actual human diet more closely. Notably, the majority of preclinical data documented a
beneficial effect of FO and its pure compounds when compared with fats more common in
the Western diet (like ω-6 or saturated fats), which emerged when focusing on CRC, its
early phases (ACF or polyps), and even metastasis (although with some exceptions, e.g.,
Griffini et al.).
The reviewed studies highlighted some of the several biological mechanisms that may
account for the protective effect of fish and FO against colorectal carcinogenesis. A detailed
overview is beyond the scope of the present paper (interested readers may refer to existing
reviews [5,6,70,71]), but some data deserve to be briefly described here. ω-3 PUFAs affect
eicosanoids metabolism by inhibiting PGE2 production, EPA can function as a substrate for
COXs to synthesize unique 3-series prostaglandin compounds (e.g., PGE3 ) less endowed
with inflammatory action compared to 2-series prostaglandin compounds [72]. EPA and
DHA produce lipid mediators endowed with pro-resolving, immunomodulatory, and anti-
inflammatory properties (which may also explain their effect in preventing cardiovascular
diseases) [6]. The molecular basis for the health benefits of ω-3 PUFA was also ascribed to
the incorporation of these fatty acids into membrane phospholipids. Moreover, ω-3 PUFA
do not enhance the luminal concentration of secondary bile acids (unlike saturated fats
and ω-6 PUFA) and lower colon and liver activity of ornithine decarboxylase (ODC) and
tyrosine-specific protein kinase (TPK), all of which are implicated in colon carcinogene-
sis [70]. Positive epigenetic effects have also been described together with interaction of
ω-3 PUFA with nuclear receptors and transcription factors, thus, altering the proliferation,
lipid metabolism, and apoptosis of cancer cells [5]. More recently, ω-3 PUFA was reported
to be associated with higher intestinal microbial diversity, thus, improving host immune
function and eventually halting the development of CRC [71,73].
ω-3 PUFA is also the nutrient contained in fish that was most extensively investigated
in relation to CRC risk among humans, including in several prospective studies that mostly
yielded results consistent with preclinical studies. In 2020, Kim et al. published a meta-
analysis of twenty prospective studies which encompassed a total of 18,102 cases and over
1.3 million participants and found that ω-3 PUFA intake (from fish or FO supplements) was
inversely associated with CRC risk [7]. In particular, linear increments by 0.1 g/day in the
intake of EPA and DHA were associated with statistically significant 5% and 3% reductions,
respectively, in CRC risk. Similar reviews for fish components other than ω-3 PUFA are
warranted to reach a clearer understanding of the mechanisms underlying the observed
protective effect of fish consumption against CRC risk. Like for every food, however, it is
important to emphasize that the nutritional value of fish, and its association with human
disease, is better evaluated on the basis of its consumption as a whole rather than as the
effect of any single nutrient therein, both because the different nutrients may interact
between one another in complex ways once they have entered the body, and because the
Cancers 2022, 14, 640 15 of 20

food matrix affects how the food is digested and absorbed [3,74]. The finding of Kim et al.’s
meta-analysis deserves attention, however, because it is consistent with data from animal
models and confirms the role of ω-3 PUFA intake as an important (although not exclusive)
mediator of the protective effect of fish consumption against CRC risk.
The main strength of this paper lies in our having reviewed the available evidence from
both preclinical and epidemiological studies with the aim of giving as broad an overview
as possible on the role of fish in the prevention of CRC and on the possible mechanisms
underlying this effect. The meta-analysis of prospective human studies takes advantage
of a very large size and of a remarkable consistency of results across time and space: the
studies were conducted in the USA, Europe, Australia, and Asia, and participants were
followed up from as early as 1967 until as late as 2014. Results from highest-versus-lowest
and dose–response analyses were consistent, heterogeneity among studies was low-to-
negligible, and there was no evidence of publication bias, which further contributed to
the reliability of our findings. Articles included in the review of preclinical studies were
published over a period of more than three decades (1986–2018): in spite of the inevitable
differences in study methods, the data were considerably consistent and provided valuable
support to those originating from epidemiological studies, as well as many insights into
underlying causal mechanisms. Our work also has some weaknesses that are important
to acknowledge. A few papers may have been missed for having limited the search to
only two databases (PUBMED/MEDLINE and EMBASE): however, the meticulous citation
chaining of the papers included, the previously published reviews, the meta-analysis, and
the lack of evidence for publication bias in meta-analysis suggest that the number of missed
papers is unlikely to be high and their impact on the summary results minimal. Regarding
the epidemiological studies, we merged results from studies that varied in terms of how
the exposure of interest (“fish”) was defined (crustaceans and shellfish were included only
in some studies, and a clear definition was sometimes missing) and measured (g/day or
frequency of consumption). Moreover, the categories of fish consumption used to conduct
the highest-versus-lowest analysis differed across studies, which is, however, usual in
meta-analysis in the field of nutritional epidemiology. Furthermore, only a limited subset
of studies (7 out of 25) contributed to the dose–response meta-analysis and summary
results were heavily affected by a single study [13] which contributed around two-thirds of
the statistical power of that analysis. Our meta-analysis does not suggest that any major
disparity by sex exists in the association between fish consumption and CRC risk; however,
only around half of the studies provided separate risk estimates for men and women, which
may have curbed our ability to detect small differences in the strength of the association.
With regard to studies on animal models, in addition to the aforementioned very large
variability in study design, materials and methods, and to the use of fish meat (instead of
fish oil) in only a very limited subset of studies, an important limitation is the fact that some
of the potential biological mechanisms (e.g., modification of COX-2 activity, prostaglandins
production, and others) were investigated in only a few studies; therefore, confirmation is
needed to corroborate and extend current knowledge.

5. Conclusions
In conclusion, by jointly reviewing epidemiological studies among human popula-
tions and preclinical studies in animal models we found evidence that increasing fish
consumption may effectively help inhibit colorectal carcinogenesis. CRC is a disease char-
acterized by high incidence and mortality rates worldwide, so even a moderate reduction
in its risk at the individual level (such as that achievable by increasing the consumption of
fish, according to our findings) may lead to a major reduction in its disease burden at the
population level. The consistent evidence from epidemiological and preclinical studies that
fish consumption may be effective in preventing CRC should be taken into account when
making dietary recommendations for cancer prevention.
Cancers 2022, 14, 640 16 of 20

Supplementary Materials: The following supporting information can be downloaded at: https:
//www.mdpi.com/article/10.3390/cancers14030640/s1, Figure S1: Forest plot for the association
between fish consumption (highest vs. lowest category of consumption) and colorectal cancer risk
among women; Figure S2: Forest plot for the association between fish consumption (highest vs.
lowest category of consumption) and colorectal cancer risk among men; Figure S3: Forest plot
for the association between fish consumption (highest vs. lowest category of consumption) and
colon cancer risk; Figure S4: Forest plot for the association between fish consumption (highest
vs. lowest category of consumption) and colon cancer risk among women; Figure S5: Forest plot
for the association between fish consumption (highest vs. lowest category of consumption) and
colon cancer risk among men; Figure S6: Forest plot for the association between fish consumption
(highest vs. lowest category of consumption) and rectal cancer risk; Figure S7: Forest plot for the
association between fish consumption (highest vs. lowest category of consumption) and rectal cancer
risk among women; Figure S8: Forest plot for the association between fish consumption (highest vs.
lowest category of consumption) and rectal cancer risk among men; Table S1: Quality assessment
(using the Newcastle–Ottawa Scale tool) of the articles included in the meta-analysis; Table S2: Exact
definition (as reported in the paper text) of the exposures that were studied in relation to colorectal,
colon, or rectal cancer risk in each included study. Risk estimates entered in meta-analysis models
were those marked with the asterisk (*); Table S3: Studies and corresponding relative risk estimates
included in the meta-analysis of the association between fish consumption (highest vs. lowest
category of consumption) and colorectal cancer risk; Table S4: Studies and corresponding relative
risk estimates included in the meta-analysis of the association between fish consumption (highest vs.
lowest category of consumption) and colon cancer risk; Table S5: Studies and corresponding relative
risk estimates included in the meta-analysis of the association between fish consumption (highest
vs. lowest category of consumption) and rectal cancer risk; Table S6: Studies and corresponding
relative risk estimates included in the meta-analysis of the association between fish consumption
(highest vs. lowest category of consumption) and colorectal cancer risk among women; Table S7:
Studies and corresponding relative risk estimates included in the meta-analysis of the association
between fish consumption (highest vs. lowest category of consumption) and colon cancer risk among
women; Table S8: Studies and corresponding relative risk estimates included in the meta-analysis of
the association between fish consumption (highest vs. lowest category of consumption) and rectal
cancer risk among women; Table S9: Studies and corresponding relative risk estimates included
in the meta-analysis of the association between fish consumption (highest vs. lowest category of
consumption) and colorectal cancer risk among men; Table S10: Studies and corresponding relative
risk estimates included in the meta-analysis of the association between fish consumption (highest
vs. lowest category of consumption) and colon cancer risk among men; Table S11: Studies and
corresponding relative risk estimates included in the meta-analysis of the association between fish
consumption (highest vs. lowest category of consumption) and rectal cancer risk among men;
Table S12: Experiments with fish and its components on animal models of colorectal cancer; Table S13:
Experiments with fish or its components on different animal models of colorectal cancer metastasis.
Author Contributions: Conceptualization, S.C. (Saverio Caini), G.C. and G.M.; methodology, S.C.
(Saverio Caini), S.C. (Sofia Chioccioli) and G.C.; software, S.C. (Saverio Caini); validation, E.P., M.F.
and K.T.; formal analysis, S.C. (Saverio Caini); investigation, all authors; resources, G.C. and G.M.;
data curation, S.C. (Saverio Caini), S.C. (Sofia Chioccioli), E.P., M.F. and K.T.; writing—original draft
preparation, S.C. (Saverio Caini), S.C. (Sofia Chioccioli), E.P., M.F. and K.T.; writing—review and
editing, G.C. and G.M.; visualization, S.C. (Saverio Caini); supervision, S.C. (Saverio Caini), G.C. and
G.M.; project administration, G.C. and G.M.; funding acquisition, not applicable. All authors have
read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
ACF: Aberrant Crypt Foci. AOM: Azoxymethane. AOM/DSS: Azoxymethane/Dextran
Sodium Sulfate. APC: Adenomatous Polyposis Coli. CAC: Colitis-Associated Colorectal cancer.
CI: Confidence Intervals. CO: Corn Oil. COX-2: Cyclossigenase-2. CRC: Colorectal Cancer. DHA:
Docosahexaenoic Acid. DMH: 1,2-Dimethylhydrazine. EPA: Eicosapentaenoic acid. EPA-FFA: Eicos-
Cancers 2022, 14, 640 17 of 20

apentaenoic Acid as Free Fatty Acid. EPIC: European Prospective Investigation into Cancer and
Nutrition. ERK: Extracellular Signal-Regulated Kinase. FFA: Free Fatty Acid. FO: Fish Oil. GCO:
Global Cancer Observatory. HFML: High-Fat diet containing Mixed Lipids. LA: Linoleic Acid. LogRR:
Log Relative Risk. MOOSE: Guidelines for Meta-Analyses and Systematic Reviews of Observational
Studies. NCC: Nested Case–Control studies. NOS: Newcastle–Ottawa Scale. OA: Oleic Acid. ODC:
Ornithine Decarboxylase. PA: Palmitic Acid. PGE2 : Prostaglandin E2. PGE3 : Prostaglandin E3. PhIP:
2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine. PUFA (ω-3): Long chain ω-3 Polyunsturated
Fatty Acids. RCT: Randomized Clinical Trials. RR: Relative Risk. SI: Small Intestine. SRR: Summary
Risk Ratio. TPK: Tyrosine-specific Protein Kinase. WCRF: World Cancer Research Fund.

References
1. Global Cancer Observatory (GCO). Colorectal Cancer Factsheet 2020. Available online: https://gco.iarc.fr/today/data/
factsheets/cancers/10_8_9-Colorectum-fact-sheet.pdf (accessed on 31 August 2021).
2. World Cancer Research Fund/American Institute for Cancer Research. Diet, Nutrition, Physical Activity, and Cancer: A Global
Perspective. Continuous Update Project (CUP): Colorectal Cancer. Available online: https://www.wcrf.org/dietandcancer/
colorectal-cancer/ (accessed on 31 August 2021).
3. Institute of Food Research, Norwich, UK. Nutrient Analysis of Fish and Fish Products. 2013. Available online: https:
//assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/167923/Nutrient_analysis_
of_fish_and_fish_products_-_Analytical_Report.pdf (accessed on 13 October 2021).
4. Saedisomeolia, A.; Wood, L.G.; Garg, M.L.; Gibson, P.G.; Wark, P.A. Anti-inflammatory effects of long-chain n-3 PUFA in
rhinovirus-infected cultured airway epithelial cells. Br. J. Nutr. 2009, 101, 533–540. [CrossRef] [PubMed]
5. Hou, T.Y.; Davidson, L.A.; Kim, E.; Fan, Y.Y.; Fuentes, N.R.; Triff, K.; Chapkin, R.S. Nutrient-Gene Interaction in Colon Cancer,
from the Membrane to Cellular Physiology. Annu. Rev. Nutr. 2016, 36, 543–570. [CrossRef]
6. Irún, P.; Lanas, A.; Piazuelo, E. Omega-3 Polyunsaturated Fatty Acids and Their Bioactive Metabolites in Gastrointestinal
Malignancies Related to Unresolved Inflammation. A Review. Front. Pharmacol. 2019, 10, 852. [CrossRef] [PubMed]
7. Kim, Y.; Kim, J. Intake or Blood Levels of n-3 Polyunsaturated Fatty Acids and Risk of Colorectal Cancer: A Systematic Review
and Meta-analysis of Prospective Studies. Cancer Epidemiol. Biomark. Prev. 2020, 29, 288–299. [CrossRef] [PubMed]
8. Kato, I.; Akhmedkhanov, A.; Koenig, K.; Toniolo, P.G.; Shore, R.E.; Riboli, E. Prospective study of diet and female colorectal
cancer: The New York University Women’s Health Study. Nutr. Cancer 1997, 28, 276–281. [CrossRef] [PubMed]
9. Kobayashi, M.; Tsubono, Y.; Otani, T.; Hanaoka, T.; Sobue, T.; Tsugane, S.; JPHC Study Group. Fish, long-chain n-3 polyunsaturated
fatty acids, and risk of colorectal cancer in middle-aged Japanese: The JPHC study. Nutr. Cancer 2004, 49, 32–40. [CrossRef]
10. Hall, M.N.; Chavarro, J.E.; Lee, I.M.; Willett, W.C.; Ma, J. A 22-year prospective study of fish, n-3 fatty acid intake, and colorectal
cancer risk in men. Cancer Epidemiol. Biomark. Prev. 2008, 17, 1136–1143. [CrossRef]
11. Lee, S.A.; Shu, X.O.; Yang, G.; Li, H.; Gao, Y.T.; Zheng, W. Animal origin foods and colorectal cancer risk: A report from the
Shanghai Women’s Health Study. Nutr. Cancer 2009, 61, 194–205. [CrossRef]
12. Vieira, A.R.; Abar, L.; Chan, D.S.M.; Vingeliene, S.; Polemiti, E.; Stevens, C.; Greenwood, D.; Norat, T. Foods and beverages and
colorectal cancer risk: A systematic review and meta-analysis of cohort studies, an update of the evidence of the WCRF-AICR
Continuous Update Project. Ann. Oncol. 2017, 28, 1788–1802. [CrossRef]
13. Aglago, E.K.; Huybrechts, I.; Murphy, N.; Casagrande, C.; Nicolas, G.; Pischon, T.; Fedirko, V.; Severi, G.; Boutron-Ruault,
M.C.; Fournier, A.; et al. Consumption of fish and long-chain n-3 polyunsaturated fatty acids is associated with reduced risk of
colorectal cancer in a large European cohort. Clin. Gastroenterol. Hepatol. 2020, 18, 654–666.e6. [CrossRef]
14. Bradbury, K.E.; Murphy, N.; Key, T.J. Diet and colorectal cancer in UK Biobank: A prospective study. Int. J. Epidemiol. 2020, 49,
246–258. [CrossRef] [PubMed]
15. Stroup, D.F.; Berlin, J.A.; Morton, S.C.; Olkin, I.; Williamson, G.D.; Rennie, D.; Moher, D.; Becker, B.J.; Sipe, T.A.; Thacker, S.B.
Meta-analysis of observational studies in epidemiology: A proposal for reporting. Meta-analysis Of Observational Studies in
Epidemiology (MOOSE) group. JAMA 2000, 283, 2008–2012. [CrossRef] [PubMed]
16. Greenland, S. Quantitative methods in the review of epidemiologic literature. Epidemiol. Rev. 1987, 9, 1–30. [CrossRef] [PubMed]
17. Greenland, S.; Longnecker, M.P. Methods for trend estimation from summarized dose-response data, with applications to
meta-analysis. Am. J. Epidemiol. 1992, 135, 1301–1309. [CrossRef] [PubMed]
18. Higgins, J.P.; Thompson, S.G. Quantifying heterogeneity in a meta-analysis. Stat. Med. 2002, 21, 1539–1558. [CrossRef] [PubMed]
19. Egger, M.; Davey Smith, G.; Schneider, M.; Minder, C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997, 315,
629–634. [CrossRef]
20. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analysis. 2020. Available
online: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp (accessed on 2 July 2021).
21. The Comprehensive R Archive Network (CRAN). Package Dosresmeta. Available online: https://cran.r-project.org/web/
packages/dosresmeta/dosresmeta.pdf (accessed on 9 June 2021).
Cancers 2022, 14, 640 18 of 20

22. Bamia, C.; Lagiou, P.; Buckland, G.; Grioni, S.; Agnoli, C.; Taylor, A.J.; Dahm, C.C.; Overvad, K.; Olsen, A.; Tjønneland, A.; et al.
Mediterranean diet and colorectal cancer risk: Results from a European cohort. Eur. J. Epidemiol. 2013, 28, 317–328. [CrossRef]
23. Engeset, D.; Andersen, V.; Hjartåker, A.; Lund, E. Consumption of fish and risk of colon cancer in the Norwegian Women and
Cancer (NOWAC) study. Br. J. Nutr. 2007, 98, 576–582. [CrossRef]
24. Murff, H.J.; Shu, X.O.; Li, H.; Dai, Q.; Kallianpur, A.; Yang, G.; Cai, H.; Wen, W.; Gao, Y.T.; Zheng, W. A prospective study of
dietary polyunsaturated fatty acids and colorectal cancer risk in Chinese women. Cancer Epidemiol. Biomark. Prev. 2009, 18,
2283–2291. [CrossRef]
25. Spencer, E.A.; Key, T.J.; Appleby, P.N.; Dahm, C.C.; Keogh, R.H.; Fentiman, I.S.; Akbaraly, T.; Brunner, E.J.; Burley, V.; Cade, J.E.;
et al. Meat, poultry and fish and risk of colorectal cancer: Pooled analysis of data from the UK dietary cohort consortium. Cancer
Causes Control 2010, 21, 1417–1425. [CrossRef]
26. Kantor, E.D.; Lampe, J.W.; Peters, U.; Vaughan, T.L.; White, E. Long-chain omega-3 polyunsaturated fatty acid intake and risk of
colorectal cancer. Nutr. Cancer 2014, 66, 716–727. [CrossRef] [PubMed]
27. Song, M.; Chan, A.T.; Fuchs, C.S.; Ogino, S.; Hu, F.B.; Mozaffarian, D.; Ma, J.; Willett, W.C.; Giovannucci, E.L.; Wu, K. Dietary
intake of fish, ω-3 and ω-6 fatty acids and risk of colorectal cancer: A prospective study in U.S. men and women. Int. J. Cancer
2014, 135, 2413–2423. [CrossRef] [PubMed]
28. Daniel, C.R.; Cross, A.J.; Graubard, B.I.; Hollenbeck, A.R.; Park, Y.; Sinha, R. Prospective investigation of poultry and fish intake
in relation to cancer risk. Cancer Prev. Res. 2011, 4, 1903–1911. [CrossRef] [PubMed]
29. Sugawara, Y.; Kuriyama, S.; Kakizaki, M.; Nagai, M.; Ohmori-Matsuda, K.; Sone, T.; Hozawa, A.; Nishino, Y.; Tsuji, I. Fish
consumption and the risk of colorectal cancer: The Ohsaki Cohort Study. Br. J. Cancer 2009, 101, 849–854. [CrossRef] [PubMed]
30. Butler, L.M.; Wang, R.; Koh, W.P.; Yu, M.C. Prospective study of dietary patterns and colorectal cancer among Singapore Chinese.
Br. J. Cancer 2008, 99, 1511–1516. [CrossRef]
31. Siezen, C.L.; Bueno-de-Mesquita, H.B.; Peeters, P.H.; Kram, N.R.; van Doeselaar, M.; van Kranen, H.J. Polymorphisms in the
genes involved in the arachidonic acid-pathway, fish consumption and the risk of colorectal cancer. Int. J. Cancer 2006, 119,
297–303. [CrossRef]
32. Brink, M.; Weijenberg, M.P.; de Goeij, A.F.; Roemen, G.M.; Lentjes, M.H.; de Bruïne, A.P.; Goldbohm, R.A.; van den Brandt, P.A.
Meat consumption and K-ras mutations in sporadic colon and rectal cancer in The Netherlands Cohort Study. Br. J. Cancer 2005,
92, 1310–1320. [CrossRef]
33. Larsson, S.C.; Rafter, J.; Holmberg, L.; Bergkvist, L.; Wolk, A. Red meat consumption and risk of cancers of the proximal colon,
distal colon and rectum: The Swedish Mammography Cohort. Int. J. Cancer 2005, 113, 829–834. [CrossRef]
34. English, D.R.; MacInnis, R.J.; Hodge, A.M.; Hopper, J.L.; Haydon, A.M.; Giles, G.G. Red meat, chicken, and fish consumption and
risk of colorectal cancer. Cancer Epidemiol. Biomark. Prev. 2004, 13, 1509–1514.
35. Lin, J.; Zhang, S.M.; Cook, N.R.; Lee, I.M.; Buring, J.E. Dietary fat and fatty acids and risk of colorectal cancer in women. Am. J.
Epidemiol. 2004, 160, 1011–1022. [CrossRef]
36. Tiemersma, E.W.; Kampman, E.; Bueno de Mesquita, H.B.; Bunschoten, A.; van Schothorst, E.M.; Kok, F.J.; Kromhout, D. Meat
consumption, cigarette smoking, and genetic susceptibility in the etiology of colorectal cancer: Results from a Dutch prospective
study. Cancer Causes Control 2002, 13, 383–393. [CrossRef] [PubMed]
37. Knekt, P.; Järvinen, R.; Dich, J.; Hakulinen, T. Risk of colorectal and other gastro-intestinal cancers after exposure to nitrate, nitrite
and N-nitroso compounds: A follow-up study. Int. J. Cancer 1999, 80, 852–856. [CrossRef]
38. Pietinen, P.; Malila, N.; Virtanen, M.; Hartman, T.J.; Tangrea, J.A.; Albanes, D.; Virtamo, J. Diet and risk of colorectal cancer in a
cohort of Finnish men. Cancer Causes Control 1999, 10, 387–396. [CrossRef] [PubMed]
39. Gaard, M.; Tretli, S.; Løken, E.B. Dietary factors and risk of colon cancer: A prospective study of 50,535 young Norwegian men
and women. Eur. J. Cancer Prev. 1996, 5, 445–454. [PubMed]
40. Bostick, R.M.; Potter, J.D.; Kushi, L.H.; Sellers, T.A.; Steinmetz, K.A.; McKenzie, D.R.; Gapstur, S.M.; Folsom, A.R. Sugar, meat,
and fat intake, and non-dietary risk factors for colon cancer incidence in Iowa women (United States). Cancer Causes Control 1994,
5, 38–52. [CrossRef] [PubMed]
41. Reddy, B.S.; Maruyama, H. Effect of dietary fish oil on azoxymethane-induced colon carcinogenesis in male F344 rats. Cancer Res.
1986, 46, 3367–3370. [PubMed]
42. Reddy, B.S.; Sugie, S. Effect of different levels of omega-3 and omega-6 fatty acids on azoxymethane-induced colon carcinogenesis
in F344 rats. Cancer Res. 1988, 48, 6642–6647.
43. Minoura, T.; Takata, T.; Sakaguchi, M.; Takada, H.; Yamamura, M.; Hioki, K.; Yamamoto, M. Effect of dietary eicosapentaenoic
acid on azoxymethane-induced colon carcinogenesis in rats. Cancer Res. 1988, 48, 4790–4794.
44. Deschner, E.E.; Lytle, J.S.; Wong, G.; Ruperto, J.F.; Newmark, H.L. The effect of dietary omega-3 fatty acids (fish oil) on
azoxymethanol-induced focal areas of dysplasia and colon tumor incidence. Cancer 1990, 66, 2350–2356. [CrossRef]
45. Reddy, B.S.; Burill, C.; Rigotty, J. Effect of diets high in omega-3 and omega-6 fatty acids on initiation and postinitiation stages of
colon carcinogenesis. Cancer Res. 1991, 51, 487–491.
46. Takahashi, M.; Minamoto, T.; Yamashita, N.; Yazawa, K.; Sugimura, T.; Esumi, H. Reduction in formation and growth of
1,2-dimethylhydrazine-induced aberrant crypt foci in rat colon by docosahexaenoic acid. Cancer Res. 1993, 53, 2786–2789.
[PubMed]
Cancers 2022, 14, 640 19 of 20

47. Takahashi, M.; Minamoto, T.; Yamashita, N.; Kato, T.; Yazawa, K.; Esumi, H. Effect of docosahexaenoic acid on azoxymethane-
induced colon carcinogenesis in rats. Cancer Lett. 1994, 83, 177–184. [CrossRef]
48. Takahashi, M.; Fukutake, M.; Isoi, T.; Fukuda, K.; Sato, H.; Yazawa, K.; Sugimura, T.; Wakabayashi, K. Suppression of
azoxymethane-induced rat colon carcinoma development by a fish oil component, docosahexaenoic acid (DHA). Carcinogenesis
1997, 18, 1337–1342. [CrossRef] [PubMed]
49. Takahashi, M.; Totsuka, Y.; Masuda, M.; Fukuda, K.; Oguri, A.; Yazawa, K.; Sugimura, T.; Wakabayashi, K. Reduction in formation
of 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine (PhIP)-induced aberrant crypt foci in the rat colon by docosahexaenoic
acid (DHA). Carcinogenesis 1997, 18, 1937–1941. [CrossRef] [PubMed]
50. Chang, W.L.; Chapkin, R.S.; Lupton, J.R. Fish oil blocks azoxymethane-induced rat colon tumorigenesis by increasing cell
differentiation and apoptosis rather than decreasing cell proliferation. J. Nutr. 1998, 128, 491–497. [CrossRef] [PubMed]
51. Rao, C.V.; Hirose, Y.; Indranie, C.; Reddy, B.S. Modulation of experimental colon tumorigenesis by types and amounts of dietary
fatty acids. Cancer Res. 2001, 61, 1927–1933.
52. Oshima, M.; Takahashi, M.; Oshima, H.; Tsutsumi, M.; Yazawa, K.; Sugimura, T.; Nishimura, S.; Wakabayashi, K.; Taketo, M.M.
Effects of docosahexaenoic acid (DHA) on intestinal polyp development in Apc delta 716 knockout mice. Carcinogenesis 1995, 16,
2605–2607. [CrossRef]
53. Paulsen, J.E.; Elvsaas, I.K.; Steffensen, I.L.; Alexander, J. A fish oil derived concentrate enriched in eicosapentaenoic and
docosahexaenoic acid as ethyl ester suppresses the formation and growth of intestinal polyps in the Min mouse. Carcinogenesis
1997, 18, 1905–1910. [CrossRef]
54. Petrik, M.B.; McEntee, M.F.; Chiu, C.H.; Whelan, J. Antagonism of arachidonic acid is linked to the antitumorigenic effect of
dietary eicosapentaenoic acid in Apc(Min/+) mice. J. Nutr. 2000, 130, 1153–1158. [CrossRef]
55. Fini, L.; Piazzi, G.; Daoud, Y.; Selgrad, M.; Maegawa, S.; Garcia, M.; Fogliano, V.; Romano, M.; Graziani, G.; Vitaglione, P.;
et al. Chemoprevention of intestinal polyps in ApcMin/+ mice fed with western or balanced diets by drinking annurca apple
polyphenol extract. Cancer Prev. Res. 2011, 4, 907–915. [CrossRef]
56. Piazzi, G.; D’Argenio, G.; Prossomariti, A.; Lembo, V.; Mazzone, G.; Candela, M.; Biagi, E.; Brigidi, P.; Vitaglione, P.; Fogliano, V.;
et al. Eicosapentaenoic acid free fatty acid prevents and suppresses colonic neoplasia in colitis-associated colorectal cancer acting
on Notch signaling and gut microbiota. Int. J. Cancer 2014, 135, 2004–2013. [CrossRef] [PubMed]
57. Steppeler, C.; Sødring, M.; Egelandsdal, B.; Kirkhus, B.; Oostindjer, M.; Alvseike, O.; Gangsei, L.E.; Hovland, E.M.; Pierre, F.;
Paulsen, J.E. Effects of dietary beef, pork, chicken and salmon on intestinal carcinogenesis in A/J Min/+ mice. PLoS ONE 2017,
12, e0176001. [CrossRef] [PubMed]
58. Masuda, J.; Umemura, C.; Yokozawa, M.; Yamauchi, K.; Seko, T.; Yamashita, M.; Yamashita, Y. Dietary Supplementation of
Selenoneine-Containing Tuna Dark Muscle Extract Effectively Reduces Pathology of Experimental Colorectal Cancers in Mice.
Nutrients 2018, 10, 1380. [CrossRef] [PubMed]
59. Cannizzo, F., Jr.; Broitman, S.A. Postpromotional effects of dietary marine or safflower oils on large bowel or pulmonary implants
of CT-26 in mice. Cancer Res. 1989, 49, 4289–4294. [PubMed]
60. Iigo, M.; Nakagawa, T.; Ishikawa, C.; Iwahori, Y.; Asamoto, M.; Yazawa, K.; Araki, E.; Tsuda, H. Inhibitory effects of docosahex-
aenoic acid on colon carcinoma 26 metastasis to the lung. Br. J. Cancer 1997, 75, 650–655. [CrossRef] [PubMed]
61. Suzuki, I.; Iigo, M.; Ishikawa, C.; Kuhara, T.; Asamoto, M.; Kunimoto, T.; Moore, M.A.; Yazawa, K.; Araki, E.; Tsuda, H. Inhibitory
effects of oleic and docosahexaenoic acids on lung metastasis by colon-carcinoma-26 cells are associated with reduced matrix
metalloproteinase-2 and -9 activities. Int. J. Cancer 1997, 73, 607–612. [CrossRef]
62. Iwamoto, S.; Senzaki, H.; Kiyozuka, Y.; Ogura, E.; Takada, H.; Hioki, K.; Tsubura, A. Effects of fatty acids on liver metastasis of
ACL-15 rat colon cancer cells. Nutr. Cancer 1998, 31, 143–150. [CrossRef]
63. Griffini, P.; Fehres, O.; Klieverik, L.; Vogels, I.M.; Tigchelaar, W.; Smorenburg, S.M.; Van Noorden, C.J. Dietary omega-3
polyunsaturated fatty acids promote colon carcinoma metastasis in rat liver. Cancer Res. 1998, 58, 3312–3319.
64. Gutt, C.N.; Brinkmann, L.; Mehrabi, A.; Fonouni, H.; Müller-Stich, B.P.; Vetter, G.; Stein, J.M.; Schemmer, P.; Büchler, M.W. Dietary
omega-3-polyunsaturated fatty acids prevent the development of metastases of colon carcinoma in rat liver. Eur. J. Nutr. 2007, 46,
279–285. [CrossRef]
65. Hawcroft, G.; Volpato, M.; Marston, G.; Ingram, N.; Perry, S.L.; Cockbain, A.J.; Race, A.D.; Munarini, A.; Belluzzi, A.; Loadman,
P.M.; et al. The omega-3 polyunsaturated fatty acid eicosapentaenoic acid inhibits mouse MC-26 colorectal cancer cell liver
metastasis via inhibition of PGE2-dependent cell motility. Br. J. Pharmacol. 2012, 166, 1724–1737. [CrossRef]
66. Sødring, M.; Gunnes, G.; Paulsen, J.E. Spontaneous initiation, promotion and progression of colorectal cancer in the novel A/J
Min/+ mouse. Int. J. Cancer 2016, 138, 1936–1946. [CrossRef] [PubMed]
67. Wennberg, M.; Tornevi, A.; Johansson, I.; Hörnell, A.; Norberg, M.; Bergdahl, I.A. Diet and lifestyle factors associated with fish
consumption in men and women: A study of whether gender differences can result in gender-specific confounding. Nutr. J. 2012,
11, 101. [CrossRef] [PubMed]
68. Tani, S.; Kawauchi, K.; Atsumi, W.; Matsuo, R.; Ashida, T.; Imatake, K.; Suzuki, Y.; Yagi, T.; Takahashi, A.; Matsumoto, N.;
et al. Association among daily fish intake, white blood cell count, and healthy lifestyle behaviors in an apparently healthy
Japanese population: Implication for the anti-atherosclerotic effect of fish consumption. Heart Vessels 2021, 36, 924–933. [CrossRef]
[PubMed]
Cancers 2022, 14, 640 20 of 20

69. Femia, A.P.; Caderni, G. Rodent models of colon carcinogenesis for the study of chemopreventive activity of natural products.
Planta Med. 2008, 74, 1602–1607. [CrossRef]
70. Reddy, B.S. Studies with the azoxymethane-rat preclinical model for assessing colon tumor development and chemoprevention.
Environ. Mol. Mutagen. 2004, 44, 26–35. [CrossRef]
71. Song, M.; Chan, A.T. Environmental Factors, Gut Microbiota, and Colorectal Cancer Prevention. Clin. Gastroenterol. Hepatol. 2019,
17, 275–289. [CrossRef]
72. Yang, P.; Jiang, Y.; Fischer, S.M. Prostaglandin E3 metabolism and cancer. Cancer Lett. 2014, 348, 1–11. [CrossRef]
73. Piazzi, G.; Prossomariti, A.; Baldassarre, M.; Montagna, C.; Vitaglione, P.; Fogliano, V.; Biagi, E.; Candela, M.; Brigidi, P.; Balbi, T.;
et al. A Mediterranean Diet Mix Has Chemopreventive effects in a murine model of colorectal cancer modulating apoptosis and
the gut microbiota. Front. Oncol. 2019, 9, 140. [CrossRef]
74. Thorning, T.K.; Bertram, H.C.; Bonjour, J.P.; de Groot, L.; Dupont, D.; Feeney, E.; Ipsen, R.; Lecerf, J.M.; Mackie, A.; McKinley,
M.C.; et al. Whole dairy matrix or single nutrients in assessment of health effects: Current evidence and knowledge gaps. Am. J.
Clin. Nutr. 2017, 105, 1033–1045. [CrossRef]