by Jack Norris, RD
- Summary
- Research review
- Table 1. Observational research of LCN3 intakes and colorectal cancer
- Table 2. RCTs of LCN3 intakes and colorectal cancer
- References
Author’s note: For this article, I used AI tools in the literature search, analysis, data extraction, and copyediting. The conclusions reflect my assessment after verifying study details. (Last reviewed: March 2026.)
Colorectal cancer refers to cases of colon cancer and/or rectum cancer; i.e., it’s not defined as a cancer that involves both the colon and the rectum.
In their 2026 meta-analysis of cohort studies looking at associations between diet and cancer, Dunneram et al. (2026) found an increased risk of colorectal cancer (CRC) among vegans, primarily driven by rectal cancer cases. In the study, pescatarians tended to have a lower risk of colon cancer, raising the question of whether the long-chain omega-3s found in fish might be protective (see Site-specific cancer incidence: Dunneram 2026 meta-analysis). Here, we review the randomized controlled trials and observational cohort studies on long-chain omega-3s and colorectal cancer.
Summary
Given the preponderance of null findings in well-powered individual cohorts, the inconsistency across studies, and effect sizes close to 1.00, a causal protective effect of fish intake on colorectal cancer risk remains uncertain, if not unlikely. This conclusion is reinforced by the evidence from randomized controlled trials (RCTs). Of the two largest trials, together enrolling over 40,000 adults, neither VITAL nor ASCEND found a significant effect of EPA+DHA supplementation on cancer incidence or mortality. Cancer was a prespecified primary endpoint in VITAL. Even under the optimistic assumption that long-chain omega-3s prevent colorectal cancer, the number needed to treat to prevent one case of colorectal cancer among vegans appears impractically high.
Research review
The bulk of observational cohort evidence suggests that fish intake has little to no effect on colorectal cancer risk (more details are listed in Table 1).
Four large cohorts, NHS/HPFS (Song et al., 2014, ~2.8 million person-years), JPHC (Kobayashi et al., 2004; 88,658 adults), CPS-II NC (Daniel et al., 2009; 99,080 adults), and VITAL (Kantor et al., 2014, ~6.7 years, 68,109 adults), found null associations overall for fish intake and marine n-3 fatty acids. In VITAL, dietary EPA+DHA, total EPA+DHA, and dark fish intake were not significantly associated with colorectal cancer risk. High use of fish oil supplements showed a borderline inverse association (HR 0.51, 95% CI 0.26–1.00, P-trend=0.06) that did not meet conventional significance thresholds and was not corrected for multiple comparisons. The Singapore Chinese Health Study (Butler et al., 2009, ~10 years, 61,321 adults), conducted in a population with traditionally high background fish consumption as part of a largely plant- and fish-based diet, found a higher risk of colorectal cancer with EPA+DHA intake; it was significant only in the first 5 years of follow-up, suggesting reverse causation.
Some studies report modest inverse associations. In the Physicians’ Health Study (PHS), Hall et al. (2008) found a 37% lower risk of colorectal cancer in men who ate fish ≥5 times/week compared with those who ate it less than once weekly (RR 0.63, 95% CI 0.42–0.95). In the UK Biobank (Zhang et al., 2024, ~12.9 years, 253,138 adults), higher plasma omega-3 concentration was associated with lower colon cancer risk (per-SD HR 0.95, 95% CI 0.92–0.99, P-trend=0.006 in the additionally adjusted model), but the association was not observed for rectal cancer (HR 0.97, P-trend=0.198), and the colon result did not survive FDR correction in the main model (adjusted P-trend=0.076). Neither study’s findings withstand scrutiny for multiple comparisons: the PHS didn’t apply any correction, and the borderline P-values (0.02–0.05) across multiple exposures and outcomes would likely not survive even a modest false discovery rate correction; in the UK Biobank, the colon result didn’t survive FDR correction in the main model.
The most robust evidence in favor of omega-3 intakes was for EPIC (Aglago et al., 2020, ~15 years, ~476,000 adults), which found lower colorectal cancer risk with higher fish (Q5 vs Q1: HR 0.88, 95% CI 0.80–0.96) and EPA+DPA+DHA intake (Q5 vs Q1: HR 0.86, 95% CI 0.78–0.95). Unlike most studies in this article, the primary findings would likely survive multiplicity correction.
A 2022 meta-analysis by Caini et al., pooling 25 cohort studies (~2.2 million participants), found a statistically significant but small reduction in colorectal cancer risk with higher fish intake (SRR 0.94, 95% CI 0.89–0.99), with a consistent dose–response of ~4% lower risk per 50 g/day. Although heterogeneity was low, the effect size was close to the null value, and site-specific estimates did not individually reach statistical significance.
Four randomized controlled trials have examined the effect of EPA+DHA supplementation on cancer outcomes, and none found a significant benefit (more details are listed in Table 2).
The strongest evidence comes from VITAL (Manson et al., 2019), the only trial in which cancer was a prespecified primary endpoint. Among 25,871 U.S. adults supplemented with 460 mg EPA + 380 mg DHA per day for a median of 5.3 years, total invasive cancer incidence and cancer mortality were both null. Because cancer was central to VITAL’s design and the trial enrolled a diverse general population, its findings carry the greatest evidentiary weight.
ASCEND (2018) used the same dose over a longer follow-up (mean 7.4 years) in 15,480 adults with diabetes, but cancer was an exploratory rather than prespecified endpoint, and cancer outcomes weren’t corrected for multiple comparisons. Both overall cancer incidence and site-specific outcomes were null. ASCEND’s findings are consistent with VITAL but should be interpreted as supportive rather than primary evidence, given the exploratory framework and the population’s elevated competing risks of cardiovascular disease.
Two smaller trials in secondary cardiovascular prevention populations provide additional evidence, though limited. SU.FOL.OM3 (Andreeva et al., 2012, 400 mg EPA + 200 mg DHA/d, 5 years) found null results overall for cancer incidence and mortality, but reported a signal of higher cancer incidence and mortality in women (HR 3.02, 95% CI 1.33–6.89 and HR 5.49, 95% CI 1.18–25.97 respectively) that was not corrected for multiple testing. The Risk and Prevention Study (Roncaglioni et al., 2013, ~1 g EPA+DHA/d, 5 years) found no significant effect on cancer incidence or mortality.
Across all four trials, no consistent signal of benefit or harm emerged. The two largest and most relevant trials, VITAL and ASCEND, converge on a null finding despite different populations, follow-up durations, and primary endpoints. The large HR for women in SU.FOL.OM3 is likely a multiple testing artifact given the small sample size and lack of correction.
For my conclusions, see the Summary.
| Table 1. Observational research of LCN3 intakes and colorectal cancer | |||||
|---|---|---|---|---|---|
| Study | Design | Population | Results | Notes | |
| Fish and CRC risk: Meta-Analysis Caini, et al., 2022 Multi-country |
Systematic review and meta-analysis of prospective cohort studies | 25 studies; ~2,228,377 participants; 25,777 CRC cases | Primary: CRC incidence Highest vs lowest fish intake: CRC SRR 0.94 95% CI 0.89–0.99, P=0.023; Colon SRR 0.94 95% CI 0.88–1.01, P=0.089; Rectal SRR 0.94 95% CI 0.87–1.03, P=0.173; Dose–response per 50 g/day: CRC SRR 0.96 95% CI 0.92–0.99, P=0.021; Colon SRR 0.96 95% CI 0.92–1.01, P=0.140; Rectal SRR 0.95 95% CI 0.89–1.02, P=0.174; Heterogeneity: low (I2≈0–13% across models); | Fish definitions varied across studies (fish alone vs fish+seafood; intake as g/day or frequency). | |
| EPIC Aglago, et al., 2020 Multiple European countries |
European Prospective Investigation into Cancer and Nutrition (EPIC); ~15-year prospective cohort; absolute intake model | 476,160 adults for intake analysis, case-control for plasma n3 analysis; 6,291 CRC cases | Primary: CRC incidence Total fish continuous per 100 g/d (3.5 oz/d): HR 0.90 95% CI 0.82–0.98; Total fish Q5 (>51 g/d) vs Q1 (<9 g/d): HR 0.88 95% CI 0.80–0.96; Fatty fish Q5 (>18 g/d) vs Q1 (<1 g/d): HR 0.90 95% CI 0.82–0.98; EPA+DPA+DHA Q5 (>470 mg/d) vs Q1 (<77 mg/d): HR 0.86 95% CI 0.78-0.95, P-trend=0.010 Not corrected for multiple tests, but likely remain significant. No significant findings for rectal cancer; |
Adjusted for major CRC risk factors including BMI, height, alcohol, red and processed meat, fiber, dairy, physical activity, smoking, education; no strong sex heterogeneity | |
| NHS HPFS Song, et al., 2014 United States |
Prospective cohorts; Nurses’ Health Study (NHS), women only, 1984–2010, ~23.9 yrs avg follow-up; Health Professional Follow-up Study (HPFS), men only, 1986–2010, ~20.6 yrs avg follow-up | NHS: 1,822,490 person-years, 1,469 CRC cases; HPFS: 968,868 person-years, 987 CRC cases |
Primary: CRC incidence NHS and HPFS: overall null for total fish and marine n‑3 intake. Some subsites (proximal, distal, rectal) bordered on statistical significance (both positive and negative) but wouldn’t withstand correction for multiple testing. | Standard lifestyle/diet covariates | |
| SCHS Butler, et al., 2009 Singapore |
Singapore Chinese Health Study (SCHS), ~10-year prospective cohort, relative intake model | 61,321 adults; 961 CRC cases | Primary: CRC incidence EPA+DHA Q4 (~404 mg/d) vs Q1 (~131 mg/d): HR 1.22 95% CI 1.02–1.45, P-trend=0.03, not corrected for multiple testing; Sensitivity analysis: finding for overall CRC significant only for the first 5 years of follow-up suggesting possible reverse causation, HR 1.35 95% CI 1.01–1.80, P-trend=.04. Secondaries: Advanced CRC EPA+DHA Q4 (~404 mg/d) vs Q1 (~131 mg/d): HR 1.33 95% CI 1.05–1.70, P-trend=0.01; LCN3:n-6 ratio Q4 vs Q1: HR 1.45 95% CI 1.12–1.87, P-trend=0.01 |
Fats expressed per 1,000 kcal or % energy; adjusted for sex, age, year, dialect, education, smoking, alcohol, BMI, family history, diabetes, physical activity; additional adjustment for meat/preserved fish/fiber/folate didn’t materially change results. | |
| CPS-II NC Daniel, et al., 2009 United States |
Cancer Preventtion Study II Nutrition Cohort, ~6-year prospective cohort | 99,080 adults; 869 CRC cases | Primary: CRC incidence: overall null EPA+DHA Q4 (≥~245 mg/day) vs Q1 (<100 mg/day): women RR 0.94 95% CI 0.72–1.24, P-trend=0.83; men RR 1.00 95% CI 0.75–1.33, P-trend=0.90 |
Adjusted for age, education, smoking, alcohol, BMI, physical activity, family history, diabetes; additional adjustment for meat, preserved fish, fiber, folate didn’t materially change results | |
| PHS Hall, et al., 2008 United States |
Physicians’ Health Study (PHS), ~22-year prospective cohort of male physicians | 21,406 men; 500 CRC cases | Primary: CRC incidence Fish intake ≥5/week vs <1/week: RR 0.63 95% CI 0.42–0.95, P-trend=0.02; EPA+DHA (from fish) Q4 vs Q1: RR 0.76 95% CI 0.59–0.98, P-trend=0.02; Primary outcomes not corrected for multiple testing but likely wouldn’t withstand. Sensitivity analysis removing first 5 years of follow-up: P-trend=0.03 for inverse relationship between CRC and greater fish intake; suggests early cases didn’t drive results. Secondaries Colon cancer: fish ≥5/week vs <1/week RR 0.62 95% CI 0.38–1.00, P-trend=0.04; Rectal cancer: P-trend=0.29 |
No mg/day cutpoints reported for LCN3 intakes; adjusted for age, randomized aspirin assignment, smoking, BMI, multivitamin use, diabetes, vigorous exercise | |
| JPHC Kobayashi, et al., 2004 Japan |
Japan Public Health Center (JPHC); prospective study; follow-up ~7.8 years | 88,658 adults (42,525 men, 46,133 women); 705 pathologically confirmed CRC cases | Primary: CRC incidence – overall null Fish intake Q4 (84–112 g/d) vs Q1 (12–21 g/d): colon: RR 1.07 95% CI 0.77–1.48, P-trend=NS; rectal: RR 0.95 95% CI 0.63–1.43, P-trend=NS; EPA, DHA, long-chain n-3, and n-3:n-6 ratio: null for colon and rectal, P-trend=NS |
High background fish intake; weight constituting a serving size varied significantly across foods; adjusted for demographic and lifestyle covariates | |
| UK Biobank Zhang, et al., 2024 United Kingdom |
UK Biobank; ~12.9-year prospective cohort; plasma fatty acid biomarker model | 253,138 adults; 3,634 colon cases; 2,868 rectal cases | Primary: CRC incidence Plasma omega-3% per SD, additionally adjusted:
Colon HR 0.95 95% CI 0.92–0.99, P-trend=0.006; FDR correction not applied to additionally adjusted model; main model colon result does not survive FDR correction (adjusted P-trend=0.076); Rectal HR 0.97 95% CI 0.93–1.00, P-trend=0.198 (null); |
Plasma fatty acids reflect recent dietary intake and endogenous metabolism, not long-term intake, expressed as % of total fatty acids; main model adjusted for age, sex, ethnicity, Townsend deprivation index, BMI, smoking, alcohol, physical activity, assessment center; additonally adjusted model adds diabetes, aspirin, processed meat, waist-hip ratio, family history | |
| VITAL Kantor, et al., 2014 United States |
Vitamins and Lifestyle (VITAL) study; ~6.7-year prospective cohort | 68,109 adults aged 50–76; 488 incident invasive CRC cases | Primary (CRC incidence): Fish oil supplement use (10-yr use), high use (≥4 d/wk &; ≥3 yrs) vs no use: HR 0.51 95% CI 0.26–1.00, P-trend=0.06; Dietary EPA+DHA Q4 vs Q1: HR 0.92 95% CI 0.68–1.24, P-trend=0.61; Total EPA+DHA (diet + supplements) Q4 vs Q1: HR 0.88 95% CI 0.65–1.20, P-trend=0.56; Dark fish Q4 vs Q1: HR 0.77 95% CI 0.55–1.07, P-trend=0.40; Sensitivity analysis excluding first 2 years of follow-up: fish oil high use HR 0.38 95% CI 0.16–0.93, P-trend=0.05; dark fish Q4 HR 0.66 95% CI 0.45–0.99, P-trend=0.15; Subgroup analyses: Secondaries: |
Adjusted for age, sex, race/ethnicity, education, BMI, energy intake, physical activity, alcohol, smoking, multivitamin use, calcium, fiber, fruit/vegetable intake, red/processed meat, aspirin, NSAID use, family history of CRC, colonoscopy/sigmoidoscopy history, polyp history, HRT, cardiovascular disease, memory loss, cholesterol medications, omega-6 intake. | |
Abbreviations
|
|||||
| Table 2. RCTs of LCN3 intakes and colorectal cancer | ||||||
|---|---|---|---|---|---|---|
| Trial | Description | Population | Dose | Baseline | Result | Notes |
| VITAL Manson, et al., 2019 United States |
Randomized, double-blind, placebo-controlled; factorial (multiple interventions w/different groups testing each) with vitamin D; median follow-up 5.3 y | 25,871 participants; Men ≥50 y; women ≥55 y; U.S. general population | 460 mg EPA + 380 mg DHA/d | ALA/d NR; EPA+DHA/d NR; ALA status NR; LCN3 status NR | Primary outcome: total invasive cancer (null); secondaries: cancer deaths (null) | Stratified by: intake (< median 1.5 vs ≥ median 1.5 servings/week; no interaction); n3 status (none) |
|
ASCEND ASCEND Study Collaborative Group, 2018 United Kingdom |
Mean 7.4 y; randomized, double-blind, placebo-controlled; factorial with aspirin | 15,480 participants; adults with diabetes without atherosclerotic cardiovascular disease; ≥40 y (mean 63.3 ± 9.2) | 460 mg EPA + 380 mg DHA/d; placebo olive oil | ALA/d NR; EPA+DHA/d NR; ALA status NR; LCN3 status NR | Primary outcome: none; Exploratory: total cancer (null); Secondaries: site-specific cancers (null) |
No subgrouping by intake; no subgrouping by n3 status; product supplied by industry |
|
SU.FOL.OM3 Andreeva, et al., 2012 France |
5 y; randomized, double-blind, placebo-controlled; 2×2 factorial with B vitamins (4 groups: omega-3 only, B vitamins only, both omega-3 + B vitamins, placebo) | 2,501 participants; adults 45–80 y with recent myocardial infarction, unstable angina, or ischemic stroke (≤12 mo); mean 61.3 ± 9.0 | 400 mg EPA + 200 mg DHA/d | ALA/d NR; EPA+DHA/d NR; ALA status NR; plasma EPA median: 1.1%; plasma DHA median: 2.5% | Cancer outcomes (ancillary/exploratory): total cancer (overall null); cancer mortality (overall null); women: higher cancer incidence (HR 3.02, 95% CI 1.33–6.89) and mortality (HR 5.49, 95% CI 1.18–25.97), P values NR, not corrected for multiple testing | Stratified by: intake (none), n3 status (none); no cancer effect from B vitamins |
|
Risk and Prevention Study (RPS) Roncaglioni, et al., 2013 Italy |
5-year (median), double-blind, placebo-controlled; primary end point was the cumulative rate of death, nonfatal myocardial infarction, and nonfatal stroke | 12,513 adults with multiple cardiovascular risk factors; no prior myocardial infarction | 850–1,000 mg EPA+DHA/day (1 g/day n-3 ethyl esters; EPA:DHA 0.9:1–1.5:1); placebo: olive oil | Baseline fish intake similar across arms (P=0.76): never/seldom ~24%, 1×/wk ~43%, 2×/wk ~27%, ≥3×/wk ~6%; no omega-3 biomarkers reported | Cancer incidence: 490/6,239 vs 453/6,266 P=0.19; cancer mortality: HR 1.09 (95% CI 0.85–1.39), P=0.51 (unadjusted). | No cancer-specific subgrouping. |
Abbreviations
|
||||||
References
Butler LM, Wang R, Koh WP, Stern MC, Yuan JM, Yu MC. Marine n-3 and saturated fatty acids in relation to risk of colorectal cancer in Singapore Chinese: a prospective study. Int J Cancer. 2009 Feb 1;124(3):678-86. No supplementary material available.
Daniel CR, McCullough ML, Patel RC, Jacobs EJ, Flanders WD, Thun MJ, Calle EE. Dietary intake of omega-6 and omega-3 fatty acids and risk of colorectal cancer in a prospective cohort of U.S. men and women. Cancer Epidemiol Biomarkers Prev. 2009 Feb;18(2):516-25. No supplementary material available.
Hall MN, Chavarro JE, Lee IM, Willett WC, Ma J. A 22-year prospective study of fish, n-3 fatty acid intake, and colorectal cancer risk in men. Cancer Epidemiol Biomarkers Prev. 2008 May;17(5):1136-43. Erratum in: Cancer Epidemiol Biomarkers Prev. 2008 Oct;17(10):2901. No supplementary material available.
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(1):32-40. No supplementary material available.
Song M, Chan AT, Fuchs CS, Ogino S, Hu FB, Mozaffarian D, Ma J, Willett WC, Giovannucci EL, 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 Nov 15;135(10):2413-23. Supplementary material available.