What specific epidemiological studies most influenced IARC’s 2015 assessment of processed meat and colorectal cancer?
Executive summary
The IARC 2015 classification that processed meat is “carcinogenic to humans” (Group 1) and red meat “probably carcinogenic” (Group 2A) rested chiefly on a large body of epidemiological evidence — hundreds of cohort and case‑control studies pooled in systematic reviews and meta‑analyses — rather than any single landmark paper [1] [2]. The Working Group emphasized consistent positive associations for colorectal cancer across prospective cohorts and meta‑analyses (including the EPIC cohort and pooled meta‑analyses by Aune and colleagues), and summarized risk in the widely‑reported estimate that each 50 g/day of processed meat increases colorectal cancer risk by about 18% [2] [3] [4].
1. The evidence base: more than 800 studies, many prospective cohorts
IARC’s Working Group considered more than 800 studies overall — more than 700 epidemiological studies for red meat and over 400 for processed meat — drawing on prospective cohorts and case–control investigations from many countries [1] [5]. That volume of observational data — not a single trial — was the primary epidemiological substrate for the 2015 decision [1] [2].
2. The big cohorts that carried weight: EPIC and other prospective studies
Large prospective investigations such as the European Prospective Investigation into Cancer and Nutrition (EPIC) were repeatedly cited in IARC documents and related editorials as central datasets linking meat intake and colorectal cancer outcomes, and were discussed in the Working Group’s literature list [2]. Prospective cohorts reduce certain biases compared with retrospective studies, and their consistent, though modest, positive associations across populations strengthened the Working Group’s conclusion [2] [6].
3. Meta‑analyses and pooled analyses that synthesized the signal
Systematic reviews and meta‑analyses — for example those by Aune, Chan, Vieira and colleagues — distilled dozens of individual studies into summary risk estimates that IARC used to judge sufficiency of evidence; these meta‑analyses reported consistent positive associations between processed meat and colorectal cancer [2] [7]. The commonly quoted quantitative synthesis — a roughly 18% higher relative risk per 50 g/day of processed meat — derives from pooled analyses and was emphasized in WHO/IARC communications [3] [4] [5].
4. Mechanistic tie‑ins buttressed epidemiology but did not replace it
IARC also relied on mechanistic and biomarker research to interpret epidemiological associations: studies on N‑nitroso compounds, heterocyclic amines, polycyclic aromatic hydrocarbons, heme iron and DNA adduct formation were used to argue biological plausibility for the epidemiological links [6] [7]. The IARC report therefore combined observational consistency with plausible carcinogenic mechanisms to elevate processed meat to Group 1 [1] [6].
5. Who made the call and how consensus formed
The Working Group comprised 22 scientists from ten countries who reviewed the assembled epidemiology, mechanistic data and animal studies before reaching their classifications; their summary appears in Lancet Oncology and IARC’s monograph materials [2] [5]. IARC explicitly stated that the processed‑meat classification was “based on sufficient evidence from epidemiological studies that eating processed meat causes colorectal cancer” [1] [8].
6. Unresolved questions, criticisms and the scale of risk
The IARC finding prompted debate about magnitude and communication of risk: critics note that IARC’s Group 1 label indicates sufficient evidence of causation but not quantitative danger relative to other Group 1 carcinogens, and public confusion followed the 18% per‑50 g figure, which is a relative risk aggregated across studies and does not describe absolute individual risk [1] [9]. Subsequent reviews called for more research to clarify mechanisms and to refine risk estimates by meat subtype, cooking method and genetic susceptibility — gaps IARC acknowledged and that later studies have sought to address [10] [11] [12].