Reviews/Alcohol and Health/Alcohol raises colorectal cancer risk
Cancer · ALC-CA-03 · v2
Borderline: within 4 points of the Established/Likely boundary, so a re-grade could land either side.
The evidence here consistently points the same way: drinking alcohol is linked to a higher chance of colorectal cancer. A very large national cohort [18], a pooled analysis that also used a genetics-based method to probe cause [22], and two independent cohorts [25], [27] all find risk rising with intake. The genetic method matters because it is harder to explain away by other habits such as smoking or diet. The main wrinkle is dose: at light or moderate drinking the picture is mixed, and one study even found a lower risk at a few drinks a week [25], so the flat wording 'alcohol consumption increases risk' is cleanest at heavier intake. The direction of the effect is not what divided the panel; its size at the low end is.
A prospective cohort of 23,323,730 Korean adults reported a colorectal cancer hazard ratio of 1.12 (95% CI 1.11-1.14) for drinkers versus non-drinkers [18]. A pooled observational and Mendelian-randomization study found a 10% higher risk per 14 g/d (OR 1.10, 95% CI 1.07-1.13) and a genetically instrumented OR of 1.25 (95% CI 1.01-1.57), supporting causation beyond confounding [22]. Independent cohorts reproduce a positive dose-response: heavy versus light lifetime drinking HR 1.91 (95% CI 1.17-3.12) [25] and advanced lesions rising per 10 g/day (OR 1.09, 95% CI 1.00-1.19) [27]. Case-control and cross-sectional data concur that alcohol intake associates with colorectal cancer or advanced neoplasia [4], [9], and the association holds across molecular subtypes [19]. The contested point: The split is over dose, not direction: [22] states the association is 'primarily driven by heavy alcohol consumption (>28 g/d)' and is J-shaped once non-drinkers are included, and [25] found 7 to <14 drinks/wk carried a lower risk (HR 0.79, 95% CI 0.64-0.97). An unqualified 'any alcohol increases risk' therefore sits uneasily with the low-intake evidence even as the overall direction holds.
All 27 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Alcohol and Health, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The claim is unqualified but the effect is concentrated at heavy intake: [22] states the association is 'primarily driven by heavy alcohol consumption (>28 g/d)' and is J-shaped once non-drinkers are included, so 'any alcohol increases risk' overreaches.
Low/moderate intake shows the opposite sign in parts of the corpus: [25] found 7 to <14 drinks/wk associated with LOWER CRC risk (HR 0.79), and animal studies [6] and [7] found low-to-moderate ethanol neutral or beneficial for colonic markers.
Where it remains uncertain. Low-to-moderate intake is mixed and sometimes protective, with [25] reporting HR 0.79 at 7 to <14 drinks/wk and [22] describing a J-shaped curve driven by intake above 28 g/d.
Aleatory
At light-to-moderate intake the estimates straddle no effect and in [25] cross to HR 0.79, so individual low-dose risk is intrinsically noisy.
Epistemic
Whether the curve is monotonic or J-shaped depends on inclusion of non-drinkers; [22] reports non-linearity (P=0.002) that the corpus does not settle.
Inferential
[9] and [27] score advanced neoplasia/lesions rather than incident colorectal cancer, so inference to cancer itself is indirect.
The load-bearing papers from a corpus of 27, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Largest cohort; CRC HR 1.12 (1.11-1.14), positive and precise.
“Light drinking as well as moderate to heavy alcohol consumption significantly increased the risks of the three gastrointestinal cancers (HR 1.51; 95% CI, 1.43-1.60; HR 1.08; 95% CI, 1.06-1.09; HR 1.12; 95% CI, 1.11-1.14) compared with non-drinkers”
Panel auditInternal validity: high · scope of inference: Korean adults undergoing national health screening.Key limitations: self-reported intake · possible residual confounding · observational
Dose-response estimate in pooled observational data.
“MR analyses supported that higher genetically predicted alcohol consumption was associated with CRC risk (ORIVW-per 1SD = 1.25; 95% CI: 1.01, 1.57)”
Panel auditInternal validity: high · scope of inference: Alcohol and CRC risk among drinkers, with causal MR support.Key limitations: driven by heavy intake · J-shaped with non-drinkers · lower subtype power
Independent cohort confirming higher intake raises risk.
“Current drinkers with an average lifetime alcohol intake of 14 or more drinks per week, compared with one drink or less per week, had a higher risk of CRC (HR, 1.25; 95% CI, 1.01-1.53)”
Panel auditInternal validity: high · scope of inference: U.S. screening-trial adults; lifetime alcohol and CRC risk.Key limitations: self-reported intake · non-monotonic (moderate lower) · site-specific variation
Case-control supporting alcohol among strongest CRC influences.
“the multivariable analysis showed that obesity, smoking history, alcoholism and diabetes mellitus have the strongest influence on the formation of colorectal cancer”
Panel auditInternal validity: low · scope of inference: Hospital patients aged 50+ in one Chinese centre.Key limitations: single-centre case-control · recall bias · no alcohol-specific OR
4 of 27 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Causal (MR) plus large reproducible cohort evidence places the directional claim in the Established band, held just above threshold because its unqualified wording glosses genuine dose-dependence at low intake.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Low-dose dose-response. The claim's validity at light-to-moderate intake is where the corpus divides [22], [25]. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Pooled individual-participant dose-response analysis modelling non-drinkers separately with spline fits for incident CRC.
Open question: Is there a threshold below which alcohol does not raise colorectal cancer risk, or is any intake harmful? Why it is still open: [18] reports risk even at light drinking while [25] finds lower risk at 7 to <14 drinks/wk and [22] describes a J-shaped curve.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19