Chris Kresser & Dr. Joel Kahn on the Joe Rogan Experience #1175, fact-checked
“That's still 50,000 to 60,000 people a year that get colon cancer if that data's right. They wouldn't get colon cancer if they just would stop eating hot dogs.”
What the evidence shows: The WHO's International Agency for Research on Cancer (IARC) classified processed meat as carcinogenic to humans in 2015, based on an analysis finding that each 50-gram daily portion (roughly one hot dog) is associated with about an 18% relative increase in colorectal cancer risk. WHO and the American Cancer Society both frame this as a relative-risk increase from a small absolute baseline risk, not a fixed count of caused cases, and WHO's own estimate attributes roughly 34,000 cancer deaths per year worldwide, not tens of thousands of U.S. colon cancer cases, to diets high in processed meat generally, not hot dogs specifically. No public-health source establishes a discrete figure of 50,000 to 60,000 U.S. colon cancer cases caused annually by hot dog consumption; deriving such a number from the 18% relative-risk statistic requires additional assumptions not supported by the primary WHO data. The claim exaggerates a modest relative-risk association into an unsupported absolute national case count and implies individual causation ("they wouldn't get colon cancer") that the underlying epidemiology does not establish.
“I scooped cholesterol out of blocked arteries. 20% of every blockage in a heart is cholesterol. It's a fact that was discovered in 1910.”
What the evidence shows: Kahn's reference to 1910 corresponds to a real finding: German chemist Adolf Windaus reported that atherosclerotic plaques from human aortas contained roughly 25-fold more cholesterol than normal aortic tissue, a discovery that helped launch the lipid hypothesis of atherosclerosis (later extended by Anitschkov's 1913 cholesterol-fed rabbit experiments). However, that 1910 finding established relative cholesterol enrichment in plaque versus healthy tissue, not a fixed "20% of every blockage" composition figure. Current research shows atherosclerotic plaques are heterogeneous in composition, mixing cholesterol (free and esterified), a necrotic lipid core, fibrous/collagen tissue, smooth muscle cells, calcification, and inflammatory cells, with proportions varying widely by lesion type, plaque stage, and vascular bed; studies of plaque cholesterol content find it differs significantly between primary and restenotic lesions and between symptomatic and asymptomatic plaques, and reviews describe lipid burden and fibrous cap thickness as variable rather than fixed at any single percentage. No allowlisted source substantiates a universal, invariant "20% cholesterol" figure for arterial blockages. The claim is best characterized as a real historical discovery cited alongside an unsupported, oversimplified statistic.
“1997, Clark, 395 randomized clinical trials, what Chris is looking for. In metabolic ward, say, you jack up the saturated fat in your diet, you jack up your cholesterol.”
What the evidence shows: Kahn's citation refers to a real, well-known study: Clarke R, Frost C, Collins R, Appleby P, Peto R, "Dietary lipids and blood cholesterol: quantitative meta-analysis of metabolic ward studies," BMJ 1997;314:112-117. The paper pooled 395 controlled dietary experiments (median duration one month) across 129 groups of healthy volunteers housed in metabolic wards, and found that replacing saturated fat with carbohydrate lowered total and LDL cholesterol, implying the reverse (raising saturated fat intake raises cholesterol) as Kahn stated. Kahn misheard or mispronounced the author's surname ("Clark" instead of "Clarke") but got the year and the "395" figure exactly right, so the underlying citation is accurate rather than fabricated. One imprecision: the metabolic ward experiments Clarke and colleagues analyzed were tightly controlled feeding studies, not uniformly described in the paper as "randomized clinical trials" in the modern sense, though many used randomized crossover or sequential-diet designs within the same volunteers. His co-guest's on-air inability to place the citation reflects unfamiliarity with the specific paper rather than any flaw in the study itself.
“No meta-analysis of randomized controlled trials and or prospective cohort studies, those are observational studies, has found any significant difference for dietary fat interventions in all-cause mo…”
What the evidence shows: Kresser claimed that no meta-analysis of randomized controlled trials or prospective cohort studies has found any significant difference in all-cause mortality from dietary fat interventions. For saturated fat specifically, this is broadly consistent with current RCT evidence: a 2020 Cochrane review of 15 trials (56,675 participants) found no significant effect on all-cause mortality (RR 0.96, 95% CI 0.90-1.03) or cardiovascular mortality (RR 0.95, 95% CI 0.80-1.12), though the same review found a statistically significant 17% reduction in combined cardiovascular events (RR 0.83, 95% CI 0.70-0.98), a different endpoint than mortality; a 2025 meta-analysis of 9 RCTs (13,532 participants) likewise found no significant mortality difference (all-cause OR 1.01, 95% CI 0.89-1.14). However, a 2015 BMJ meta-analysis of prospective cohort studies (de Souza et al.) found that while saturated fat intake showed no association with all-cause mortality (RR 0.99, 95% CI 0.91-1.09), trans fat intake was associated with a statistically significant 34% increase in all-cause mortality (RR 1.34, 95% CI 1.16-1.56) pooled across cohort studies. As an unqualified claim spanning all dietary fat types and all meta-analyses, the statement is overstated, since at least one meta-analysis of cohort data has found a significant all-cause mortality association for a dietary fat (trans fat). Narrowed specifically to saturated fat and mortality, the claim is largely consistent with the current evidence base.
“He put half of them on a low animal diet with very high plants, let half of them eat like Los Angeles people. Eight years later, 50% on the high plant, low fat diet were alive, 0% on the standard Los…”
What the evidence shows: Los Angeles cardiologist Lester Morrison did run a diet-heart study beginning in 1948, assigning about 100 post-heart-attack patients (50 per group) to a low-fat, low-cholesterol diet or to a usual-diet control group using sequential/alternate, non-randomized, unblinded allocation in his own practice. His own published data do not match Kahn's figures: an interim report (Am Heart J, 1951) found roughly 56% survival in the diet group versus 24% in controls after 8 years, while his final 1960 JAMA report found 19 of 50 diet-group patients (38%) alive versus 0 of 50 controls alive at 12 years, not 8. Kahn's "50% vs 0% at eight years" appears to conflate the diet group's approximate 8-year survival rate with the control group's 0%-alive figure, which only occurred at the later 12-year mark, and the original study's small size, lack of randomization, and lack of blinding make its numbers unsuitable as proof that diet "reverses" heart disease. Larger modern evidence, based on cohort studies of over 400,000 people, does associate plant-based diets with modestly lower cardiovascular mortality (about 8%) and incidence (about 10%), but no randomized trial has tested plant-based diets against hard cardiovascular endpoints, so the causal "reversal" framing goes beyond what current evidence establishes.