Why might a higher BMI at 20 be linked to lower breast cancer risk later in life?

Researchers followed more than 33,000 women to test whether two measures of mammographic density could help explain why body size at different stages of life relates to breast cancer risk in strikingly different ways.

Want to read later? Download your clean, ad-free, print-friendly PDF copy by clicking here.

A large UK cohort study published in the British Journal of Cancer found that higher BMI at age 20 was associated with a lower risk of postmenopausal breast cancer, while higher BMI in later adulthood was associated with increased risk.

Background

Mammographic density (MD) reflects the proportion or amount of fibroglandular tissue in the breast, depending on the measurement method. Fibroglandular tissue contains epithelial and stromal components that can become malignant.

Higher mammographic density is associated with a higher risk of breast cancer. MD is assessed in various ways: as percentage density, density area, or absolute fibroglandular volume.

Body mass index (BMI) is an established risk factor for breast cancer, but its association varies across the life course. Previous research has shown that a higher BMI in childhood and early adulthood is associated with lower percentage MD and lower postmenopausal breast cancer risk. By comparison, a higher midlife BMI is linked to a higher risk of breast cancer.

This indicates the need to understand how excess adiposity affects breast cancer development at different stages of the life course.

Mediation studies suggested that MD might account for part of this relationship, with estimates ranging from 26% to 56%. The studies differed in the timing of adiposity exposure, density measurement, and analytical approach. Two examined childhood adiposity, while a Mendelian randomization study used genetic proxies for MD rather than directly measured phenotypes.

Currently, the UK National Health Service (NHS) uses full-field digital mammography (FFDM) as the standard. The authors of this study have previously demonstrated that, among five methods of MD measurement, the visual analog scale (VAS) method using FFDM has improved predictive performance for postmenopausal breast cancer.

It is possible that different methods may not capture the same biological pathway linking early adult BMI with postmenopausal breast cancer risk. BMI measured at the time of screening, after early adulthood, may itself affect both the MD and postmenopausal cancer risk. This makes it a potential intermediate confounder.

An earlier study, the UK Predicting Risk of Cancer at Screening (PROCAS) study, reported that early-adulthood BMI (at age 20) was associated with reduced postmenopausal breast cancer risk. In the current paper, the authors used the same cohort to examine the extent to which this inverse association was mediated by VAS percentage density and Volpara-derived fibroglandular volume (FGV).

This could help understand how adiposity over the female life course influences postmenopausal breast cancer risk.

Study characteristics

The PROCAS cohort recruited 57,898 women aged 46–73 attending routine mammographic screening in Greater Manchester between 2009 and 2015. Participants provided information on lifestyle, reproductive history, and family history of breast cancer, and mammographic density measurements.

VAS MD was assessed independently by two experienced readers and provided an area-based estimate of percentage density. Volpara-based FGV is a volume-based automated method.

Cox regression was used to estimate associations between BMI, density, and cancer risk. The mediation analysis used a counterfactual approach to decompose the total effect of BMI at age 20 into a natural direct effect and a natural indirect effect operating through MD.

Participant attributes

The main analysis included 33,821 postmenopausal women without previous breast cancer who had valid recalled weight information at age 20. BMI at age 20 was calculated using recalled weight at that age and height reported at cohort entry; denominators were smaller for analyses requiring cohort-entry BMI or mammographic density measurements. About 93% were Caucasian.

A higher BMI at age 20 was associated with a higher BMI at the time of entering the current cohort. It was also linked to a greater likelihood of prior ovary or uterus removal, earlier menarche, lower alcohol consumption, and a lower proportion of reported moderate-to-vigorous physical exercise.

A higher BMI at 20 years was associated with a lower median VAS, while FGV increased across the BMI categories. Only the VAS association remained significant after adjustment for BMI at cohort entry.

Higher BMI at cohort entry was strongly and independently associated with lower VAS, higher FGV, and greater breast fatty volume.

BMI and postmenopausal breast cancer risk

Over a median follow-up of 10.44 years, 1,261 breast cancers were included in the analysis, of which 285 were detected at the entry screen. Breast cancer incidence declined as tMI increased across categories at age 20, from 4.5% at BMI below 18.5 kg/m² to 2.6% with BMI of 30 kg/m² or higher.

Each 5 kg/m² increase in BMI at age 20 was associated with a 15% lower risk of postmenopausal breast cancer. Each 5 kg/m² increase in BMI at cohort entry was associated with a 20% higher risk.

The two density measures behaved differently, though both independently showed a positive association with postmenopausal cancer risk. Per one-standard-deviation increase, VAS percentage density was linked to a larger increase in breast cancer risk than FGV, with associated increases of 29% and 12%, respectively.

Mediation findings

For VAS percentage density, the natural indirect effect was protective. This corresponds to a 59.9% proportion of the BMI–breast cancer association being mediated through VAS density. The natural direct effect itself was less certain.

For fibroglandular volume, the pattern differed: the natural direct effect was protective, but the indirect effect through FGV was in the opposite direction (HR 1.051, 95% CI 1.016–1.083). This produced a negative proportion mediated of −37.9%, with a very wide confidence interval. FGV represented an inconsistent mediation pathway and did not provide evidence of the same mediating pathway identified using percentage density.

When BMI at cohort entry was used as the exposure rather than as an intermediate confounder, higher BMI had a direct effect associated with increased postmenopausal breast cancer risk in both MD models. FGV accounted for 33.7% of the positive association, whereas VAS percentage density produced a negative proportion-mediated effect of −45.7%.

Interpretation

According to the authors, the findings support a reduced risk of postmenopausal breast cancer with higher BMI at age 20, while higher BMI in later adulthood, and higher MD by either percentage- or volume-based measures, were associated with higher risk.

This suggests that early-adulthood adiposity and later-life adiposity may affect breast tissue through different pathways.

Higher BMI at age 20 or at cohort entry was associated with lower VAS MD scores, likely due to greater fat volume. BMI at age 20 was not associated with higher FGV in later adulthood, after adjusting for BMI. By comparison, later-adulthood adiposity was associated with a higher FGV after adjustment for early-adulthood BMI.

The results also indicate that VAS MD mediated the inverse association between BMI at age 20 and postmenopausal breast cancer risk. The wide confidence intervals mean that the mediation estimate is best read as an indication of direction and approximate magnitude rather than a precise percentage.

Indirect mediation effects involving FGV differed in direction when BMI at age 20 was compared with BMI at cohort entry. The conflicting FGV result further suggests that percentage- and volume-based measures may capture different biological pathways across the life course.

One proposed explanation is that a higher BMI at age 20 may reflect protective alterations in breast tissue composition. The authors also discuss the potential role of increased estrogen in promoting early mammary tissue differentiation, alongside lower progesterone, altered insulin-like growth factor programming, and lower prolactin later in adulthood.

Later-life adiposity may be accompanied by increased peripheral conversion of androgens to estrogen, promoting cell proliferation. Unfavorable metabolic phenotypes and insulin resistance may contribute to this process. The increased FGV associated with higher later-adulthood BMI could accompany a chronic inflammatory state in breast tissue, promoting carcinogenesis.

Limitations

Important limitations include recall of BMI at age 20, which introduces potential measurement error; participation being restricted to women attending routine screening; the predominantly Caucasian study population; possible residual confounding; causal assumptions in the mediation analysis that cannot be empirically verified; the modest proportion of mammographic-density variation explained by BMI; crude measurement of some confounders such as alcohol and exercise; lack of data on molecular breast cancer subtypes; and the smaller sample available for the volumetric FGV analyses.

Conclusion

The study suggests that the inverse association between higher BMI in early adulthood and postmenopausal breast cancer may be partly explained by its long-term association with lower percentage mammographic density. FGV, by comparison, may represent a different pathway that is more susceptible to the effects of later-life adiposity.

The authors suggest that weight trajectories across the life course may provide more information on cancer risk than a single measurement. The study does not establish the biological mechanisms underlying these associations, though the authors discuss several hormonal, developmental, metabolic, and inflammatory pathways that could contribute to them.

Higher early-adulthood BMI can adversely affect cardiometabolic health. The study does not promote such weight gain; its aim is to understand these associations and potentially improve risk prediction for women.

Journal reference:

• Jauniaux, B., Harvie, M. N., Sperrin, M., Astley, S. M., Malcomson, L., Evans, D. G., & Renehan, A. G. (2026). Early-adulthood body mass index, mammographic density and post-menopausal breast cancer risk: A mediation analysis. British Journal of Cancer, 1-11. DOI: 10.1038/s41416-026-03593-w, https://www.nature.com/articles/s41416-026-03593-w