Running

Exercise and Cancer Risk: What the Data Actually Says

July 12, 2026 · by Brendon Kennedy

I track my weekly runs on this site, and I want to be honest about why — because the honest reason is smaller and duller than the headline everyone reaches for. “Exercise cuts your cancer risk” is a real finding, but it’s a population finding, and the gap between “populations that move more get fewer of certain cancers” and “my Tuesday three-miler bought me back a tumor” is the whole point of this post. This is the running log for a research catalog, not a health guarantee. So here’s what I’m actually reading, and how it shapes what the tracker on this page does and doesn’t claim.

The big numbers

The single most-cited dataset is Moore et al. 2016 in JAMA Internal Medicine, a pooled analysis of 12 cohorts — 1.44 million adults, 186,932 cancers, a median of 11 years of follow-up. Comparing people at the 90th percentile of leisure-time activity to those at the 10th, the researchers reported lower incidence of 13 cancers. A few of the hazard ratios: esophageal adenocarcinoma HR 0.58 (95% CI 0.37–0.89), lung 0.74 (0.71–0.77), kidney 0.77 (0.70–0.85), endometrial 0.79 (0.68–0.92), colon 0.84 (0.77–0.91), and breast 0.90 (0.87–0.93).

Those line up with what the National Cancer Institute’s fact sheet pulls from the broader meta-analytic literature: roughly 19% lower colon-cancer risk in the most active (2016 meta-analysis of 126 studies), 12–21% lower breast-cancer risk (2016 meta-analysis of 38 cohort studies), and about 20% lower endometrial-cancer risk (meta-analysis of 33 studies) for the most-active versus least-active groups.

When you ask which of these the evidence graders are most confident about, the World Cancer Research Fund / AICR rates the evidence “convincing” that physical activity decreases risk for exactly three: colon, postmenopausal breast, and endometrial cancer, with “probable” evidence for premenopausal breast at vigorous intensity. Those are the three I anchor on, because “convincing” is the top rung of their grading and everything below it carries more doubt.

Why I refuse to put a personal number on it

Here’s the part I care most about getting right. Every figure above is a relative risk across a population — it describes how the rate of a cancer differs between a more-active group and a less-active group. It does not, and cannot, tell any individual how many of their own future cancers they’ve “prevented.” Taking a population hazard ratio and spending it on yourself is a statistical category error, and I’m not going to build a widget that does it.

Three reasons this matters, all straight from the sources:

• These are observational studies. The NCI states it plainly: “observational studies cannot prove a causal relationship,” and the case for causation rests on consistency across populations plus a plausible mechanism — not on a controlled experiment where researchers made people run. Active people also tend to smoke less, eat differently, and see doctors more; those confounders are hard to fully scrub out.
• The same dataset shows activity going the wrong way for some cancers. In Moore 2016, more activity was associated with higher risk of malignant melanoma (HR 1.27) and prostate cancer (HR 1.05). Nobody thinks running causes melanoma — it’s a marker for more time outdoors in the sun, and the prostate signal is commonly read as more screening among health-conscious men. That’s a clean, built-in reminder that these arrows are associations tangled with behavior and detection, not tidy dose-response causes.
• The mechanisms are biologically plausible but not individually measurable. The proposed pathways — lower circulating estrogen and insulin, less chronic inflammation, faster bowel transit, better weight regulation — are why researchers lean toward a causal read for colon, breast, and endometrial cancer. Real, but there’s no dial in my body I can watch to confirm any of it is happening to me.

So the honest individual claim is: “regular activity is associated with lower risk of specific cancers at the population level, and I’d rather be in the moving group than the still one.” That’s it. That’s the whole earned statement.

Why the tracker reports guideline-coverage, not risk

This is the design decision behind the running widget on this site. Instead of faking a personal risk figure, it reports one thing I can actually measure: how much of the published activity guideline I covered this week.

The WHO and the American Cancer Society both put the adult target at 150–300 minutes of moderate or 75–150 minutes of vigorous activity per week, plus limiting sedentary time. Those are the exposure levels the epidemiology is built on, so “what fraction of that floor did I hit?” is a defensible, honest metric. It’s a measure of the input the studies studied — not an output the studies can’t give me. Green means I met the guideline; anything less is just an honest percentage. No mortality math, no tumors-avoided counter, no dashboard pretending to price my longevity.

What I’m actually doing

Right now I run 2–4 miles a week, which is short of the guideline floor and I know it. Running is vigorous, so I measure my weeks against the 75-minute vigorous target rather than the 150-minute moderate one, and most weeks I land somewhere under it. I don’t pretend the exact minutes are fixed — they drift week to week — so the bar on this page just totals whatever I actually ran, instead of a tidy average I’d be tempted to round up. The near-term goal is boring and specific: get consistent weeks above the vigorous floor before I add distance. “Consistent” here means repeatable, not heroic — clearing that floor on back-to-back weeks rather than one strong week and then nothing. I’m logging it in public mostly to keep myself honest, and because a coverage bar is a lot harder to lie to than a vibe.

Not medical advice. This is a personal catalog of research I’m reading and habits I’m testing on myself. Nothing here diagnoses, treats, or prevents any disease, and it isn’t a substitute for a qualified clinician. Talk to your doctor before changing diet, fasting, exercise, or medication — especially with ADHD medication, alcohol, or a personal or family cancer history.

← All writing