Mammography — screening the older woman

Mammography is the mainstay of breast cancer screening.

Research has shown mammography to be effective up to age 84 — a key figure to quote.

Breast cancer is common and does shorten life in older women: Essentials records the median age at diagnosis as about 61 and the median age at death as about 69.

However, cancer in older people may grow more slowly even though incidence and prevalence rise with age — this is the basis of overdiagnosis.

There is a lack of agreement among organisations about both the screening interval and when to stop.

ACS: annual mammography for women 55 and older, or they may transition to biennial. Women should continue if overall health is good and life expectancy is 10 years or more.

NCCN: annual mammography, with no guidelines for stopping.

USPSTF (textbook version): biennial screening mammography up to age 74; evidence insufficient for women 75 or older (I statement).

Medicare (Essentials): covers a screening mammogram every 12 months for all women aged 40 and above, plus one baseline mammogram between 35 and 39, and covers digital technologies.

There is no recommendation for or against breast self-examination, breast self-awareness, or clinician-performed clinical breast examination.

NCCN recommends both; ACS recommends breast self-awareness if the woman is interested; USPSTF recommends against both (grade D for teaching BSE, grade I for CBE).

Breast self-awareness may allow detection of lesions not seen on mammography, or interval cancers developing between mammograms — the argument in its favour.

Mammography by category: every 1–2 years up to age 80 in the Robust; consider every 1–2 years up to 75 in the Frail; consider every 1–2 years up to 70 in the Moderately demented; nothing at End of life.

Breast examination: yearly in the Robust, Frail and Moderately demented; not at End of life.

Few studies describe how women decide to stop mammography.

Some women and their providers, recognising that the expected benefit of early detection declines with remaining life expectancy, consciously decide to discontinue screening following a serious health event — a practically useful trigger point.

The "10-year rule" applies: the lag time to mortality benefit is about a decade, so continue only if life expectancy ≥10 years.

Essentials states that analysis of mammography trial data illustrates just how confusing this literature can be; one analysis suggests screening after age 69 gives a small gain in life expectancy and is modestly cost-effective (Kerlikowske, 1999).

Harms to weigh: false positives, unnecessary biopsy, pain, anxiety, and overdiagnosis with overtreatment — the last being especially serious in older women who tolerate treatment poorly.

Related USPSTF positions in Essentials: BRCA testing — grade D against routine referral if family history is not suggestive, grade B for referral if it is; chemoprevention — grade D against routine tamoxifen or raloxifene at average risk, grade B for discussion in high-risk women at low risk of adverse effects.

Practical geriatric rule: screening should stop when a woman would not accept or could not tolerate a biopsy or subsequent cancer treatment — if the result cannot change management, the test should not be done.

USPSTF revised its recommendation in April 2024. The Task Force now recommends biennial screening mammography for women aged 40 to 74 (B recommendation), concludes the evidence is insufficient for women 75 years or older (I statement), and also insufficient for supplemental ultrasound or MRI in women with dense breasts (JAMA/USPSTF, 2024). The change from the 2016 version is that biennial screening now begins at 40 rather than being individualised for 40–49; the change was estimated to save 19% more lives than starting at 50. Both digital mammography and digital breast tomosynthesis ("3D mammography") are accepted as effective mammographic modalities.

The 75-and-over question remains unresolved — but the observational data are more favourable than "insufficient evidence" suggests. No randomised trial of mammography included women over 74, so it is not known whether mammography helps them live longer; data suggest around 10 years of life expectancy is needed for a mortality benefit, while the harms — pain, anxiety, complications of biopsy after a false positive, and overdiagnosis — are immediate. On the other hand, mammography performs significantly better in older women, with higher sensitivity, specificity, cancer detection rate and positive predictive value, and lower recall and false-positive rates; modelling suggests the overdiagnosis rate is low and benefits outweigh risks until about age 90. A systematic review of 36 studies in women 75+ found mixed evidence, with some studies showing a mortality benefit and others none, and called for better comparison groups and formal estimation of overdiagnosis given the substantial competing risk of death in this age group (British Journal of Cancer, 2023).

Counterpoint on overdiagnosis in the elderly. A meta-analysis of two RCTs including 33,268 women aged 70 or older found a 28% relative reduction in breast cancer mortality (RR 0.72), but estimated the proportion of overdiagnosis in older women at 19%, higher than in younger women. A 2025 review in Nature Reviews Clinical Oncology argues the case for active monitoring of selected screen-detected breast cancers in women ≥75, drawing on the COMET trial of active monitoring for low-risk DCIS (JAMA, 2025) — a shift from "detect and treat" to "detect and observe" in the very old.

Artificial intelligence is the biggest technical change — MASAI is the landmark trial. MASAI is the first randomised controlled trial of AI in breast cancer screening and the largest trial of AI in cancer screening generally, involving 105,934 Swedish women. Interim safety results (Lancet Oncology, 2023) showed a 44% reduction in screen-reading workload; a Lancet Digital Health analysis showed a 29% increase in cancer detection without any increase in false positives; and the full results (Lancet, 2026) showed 12% fewer interval cancers, meaning fewer aggressive and advanced cancers diagnosed between screening rounds. Sensitivity was 6.7 percentage points higher (80.5% vs 73.8%) at the same specificity (98.5%), with consistent results across age and breast-density subgroups. Importantly, the authors state the findings do not support replacing radiologists — AI-supported screening still requires at least one human reader.

Breast density notification is now mandatory in the US. The FDA's updated MQSA rule (effective September 2024) requires that every mammogram report include a breast density notification to the patient. USPSTF, however, still finds the evidence insufficient to recommend supplemental ultrasound or MRI on the basis of density alone.

India — mammography is not the national strategy. ICMR recommends clinical breast examination every one to three years as the practical starting point in India, along with monthly self-examination; free breast, oral and cervical cancer screening is offered at Health and Wellness Centres for adults aged 30 to 65. Uptake remains very poor — fewer than 2% of eligible women aged 30–49 had been screened for cervical or breast cancer at the 2019–21 national survey. For Indian geriatric practice this means: CBE, not mammography, is the realistic screening tool, and no programme covers women above 65 at all.