Endometrial Cancer: Activity, Adiposity, and Risk

Physical activity is associated with lower endometrial-cancer incidence, and excess body fatness has an even stronger evidence grade as a risk factor. The difficult question is how much of the activity association operates through adiposity—and whether adjusting for BMI clarifies or obscures that pathway.

Two evidence questions sit side by side

Endometrial cancer begins in the lining of the uterus. It is useful to separate two related but distinct questions: is greater physical activity associated with lower incidence, and does body fatness itself raise risk? The 2018 WCRF/AICR evidence review graded activity as probable protection against endometrial cancer. It graded greater body fatness as a convincing cause. That distinction reflects different volumes and consistency of evidence, not a claim that activity is irrelevant.

For body fatness, the WCRF/AICR endometrial report found a dose-response association: its analysis estimated a 50% higher relative risk per five-unit increase in BMI (RR 1.50; 95% CI, 1.42–1.59) across 26 studies and 18,717 cases. BMI is a population measure and does not directly measure fat distribution or metabolic health in an individual. Still, the consistent pattern across body-fatness measures underpins the stronger evidence judgment.

The activity result is more complicated. In the WCRF/AICR summary matrix, high-versus-low recreational activity was associated with RR 0.73 (95% CI, 0.58–0.93; nine studies); occupational activity was associated with RR 0.79 (95% CI, 0.71–0.88; five studies). Matthews and colleagues later pooled nine cohorts and found lower endometrial risk across recommended activity amounts before BMI adjustment, but the association did not remain after adjustment. Both results need to be reported, not selectively quoted.

The BMI adjustment puzzle

In the Matthews analysis, 7.5–15 MET-hours per week of leisure-time activity was associated with 10%–18% lower endometrial-cancer risk compared with no leisure-time activity. The association was eliminated after BMI adjustment. It is tempting to call that proof that the original result was confounded by weight. That conclusion goes beyond what the comparison can show.

There are at least two causal stories. In one, a person's body size before the measured activity pattern influences both their activity and their later cancer risk; BMI then acts as a confounder, and adjustment may reduce bias.

In another, years of activity affect adiposity, which in turn affects risk; adiposity lies on the pathway, so adjustment removes part of the total association one might want to estimate. A single BMI measure may also capture body size imperfectly, leaving residual confounding while blocking some mediated effect.

Timing is decisive. BMI measured before exposure, repeatedly over time, or at the same visit as activity does not carry the same causal meaning. Researchers need longitudinal measures and explicit causal models to distinguish a pre-existing confounder from a mediator. Adjustment can also change the question: the total activity association differs from an estimate intended to isolate pathways not operating through measured BMI.

That is why the older WCRF/AICR endometrial evidence synthesis and the later Matthews pooled analysis should not be reduced to “one is right, one is wrong.” The WCRF report noted that BMI adjustment generally did not change the direction or statistical significance across the studies then included. In the later pooled analysis, the association attenuated substantially after BMI adjustment. Differences in populations, measures, models, and data volume may contribute. The causal role of adiposity remains an open research question.

Why adiposity is biologically relevant

Endometrial tissue responds to hormonal and metabolic signals. Greater body fatness can influence estrogen availability, insulin resistance, and inflammatory pathways; these mechanisms are part of why body-fatness evidence is substantial. Physical activity can influence some of the same systems directly and may also help prevent weight gain. A pathway can be biologically plausible without proving the size of its contribution in humans.

Activity is not simply a proxy for weight. Two people with similar BMI can differ in fitness, muscle mass, visceral fat, insulin sensitivity, and activity history. Conversely, BMI adjustment does not perfectly account for all these features. Studies that rely on one snapshot of weight and activity may not capture the long-term exposures most relevant to a cancer that develops over many years.

Endometrial cancer also has risk factors outside exercise and body size. Reproductive history, age, diabetes, and some hormone exposures matter; medication choices and symptoms need clinical interpretation. This page does not estimate personal risk or suggest changing medication or weight for cancer prevention.

What future studies could clarify

Better answers require measurements that follow people over time rather than relying on one snapshot. Repeated assessment of activity, weight, waist circumference, and metabolic markers could help establish which exposures came first. A causal model should state whether its goal is to estimate the overall activity association or the portion not mediated through adiposity; the same adjustment set is not automatically right for both questions.

Long follow-up also matters because endometrial cancer may be diagnosed years after the behaviors being measured. Researchers can examine whether results persist after excluding early cases, compare activity domains, and report estimates by relevant subgroups. Such analyses can reduce reverse-causation concerns and reveal whether recreational and occupational activity behave similarly, though they still cannot remove all unmeasured confounding.

Randomized exercise studies can strengthen evidence about intermediate pathways—such as insulin sensitivity, inflammatory markers, or body composition—but they are generally too short and too small to test cancer incidence directly. A shift in a marker is useful mechanistic information, not a substitute for an outcome trial. The strongest synthesis will combine long-term cohort follow-up, well-designed intervention studies, and transparent causal reasoning.

There is also a practical communication need: studies should report absolute event rates alongside relative estimates where possible. Relative risks describe proportional differences, while absolute risk helps show how many diagnoses occur in a particular population over a defined period. The WCRF relative estimates cannot be used to calculate a personal probability without an appropriate baseline-risk model.

Until those gaps narrow, the balanced message is stable: activity is associated with lower risk, adiposity is a major independent risk factor, and the pathway connecting them is not fully resolved. Nothing in that uncertainty requires a reader to choose between movement and weight management as if one invalidates the other.

What to take from the evidence

For a reader, the most defensible summary has two parts: higher physical activity is associated with lower endometrial-cancer incidence, and the direct evidence is less certain than the evidence linking greater body fatness to higher risk. The association may include both weight-related and weight-independent pathways. The current evidence cannot reliably apportion them.

Public-health activity recommendations remain reasonable for overall health. The American Cancer Society advises adults to accumulate 150–300 minutes of moderate activity or 75–150 minutes of vigorous activity each week, or an equivalent combination. This is not a validated endometrial-cancer dose. An achievable routine may include walking, cycling, gardening, swimming, or other movement; activity can be increased gradually and adapted to health status.

Do not convert evidence about group averages into blame. Endometrial cancer can occur in people who are active or have lower body weight, and many cases arise without a single identifiable cause. Weight stigma does not improve risk communication. An unexplained change such as postmenopausal bleeding needs prompt medical assessment regardless of exercise level or body size; symptom evaluation and screening guidance belong to clinicians.

Questions, answered briefly

• 🧮 If BMI removes the association, did activity matter? The result is compatible with several explanations. It could reflect confounding, a pathway through adiposity, measurement error, or a mixture. The model alone cannot choose among them.
• 📉 Does RR 0.73 mean 27 fewer cases in every 100 women? No. It is a relative estimate for highest versus lowest recreational-activity categories, not an absolute count or personal prediction.
• 🪑 Is sitting the same as not exercising? No. Sedentary behavior is a distinct exposure. WCRF/AICR judged evidence that sedentary habits might increase endometrial risk as limited, and that finding should not be merged with activity evidence.
• 🩺 Should someone change treatment or diet based on this page? No. Clinical decisions, unexplained bleeding, and personal cancer-risk questions require a qualified healthcare professional.

Related Topics