Alzheimer's disease and Alzheimer's disease-related dementias (AD–ADRD) are progressive neurodegenerative conditions that occur more frequently in older adults. As there are no curative treatments, prevention is a public health priority. AD–ADRD onset occurs late in the life course, whereas causative exposures and pathogenic biological changes accumulate from early life.1Livingston G Huntley J Sommerlad A et al.Dementia prevention, intervention, and care: 2020 report of the Lancet Commission.Lancet. 2020; 396: 413-446Summary Full Text Full Text PDF PubMed Scopus (1661) Google Scholar This disparity creates a challenge for prevention science: the individuals who can benefit most from intervention might not yet show clinical signs of disease, making it hard to determine effectiveness without decades of follow up. Prevention trials need outcome measures that can quantify progression or reduction in AD–ADRD risk from early to midlife.2Elliott ML MRI-based biomarkers of accelerated aging and dementia risk in midlife: how close are we?.Ageing Res Rev. 2020; 61101075Crossref PubMed Scopus (14) Google Scholar To address this challenge, investigators are developing outcome measures to quantify the life course development of AD–ADRD risk. Among these outcome measures are machine-learning-derived algorithms that integrate thousands of measurements from structural MRI of the brain into a single number, referred to as brain age.3Cole JH Marioni RE Harris SE Deary IJ Brain age and other bodily ‘ages’: implications for neuropsychiatry.Mol Psychiatry. 2018; 24: 266-281Crossref PubMed Scopus (145) Google Scholar Brain age algorithms are developed by comparing brain MRI scans of young and old patients to build a predictive model. When applied to new patients, the model predicts the age at which that patients’ brain would appear typical in the sample used to train the algorithm. Ageing is the primary risk factor for AD–ADRD. Brain age algorithms are therefore interpreted to reflect increased risk of disease when a patient's brain age is greater than their chronological age, and vice versa. Although studies are beginning to link brain age with early-life exposures, trajectories of cognitive decline, and other signs of biological ageing,4Elliott ML Belsky DW Knodt AR et al.Brain-age in midlife is associated with accelerated biological aging and cognitive decline in a longitudinal birth cohort.Mol Psychiatry. 2019; 26: 3829-3838Crossref PubMed Scopus (51) Google Scholar, 5Reuben A Elliott ML Abraham WC et al.Association of childhood lead exposure with MRI measurements of structural brain integrity in midlife.JAMA. 2020; 324: 1970-1979Crossref PubMed Scopus (15) Google Scholar data that establish robust links between older brain age and life-course signs of AD–ADRD risk are needed. The study by Aaron Z Wagen and colleagues6Wagen AZ Coath W Keshavan A et al.Life course, genetic, and neuropathological associations with brain age in the 1946 British Birth Cohort: a population-based study.Lancet Healthy Longev. 2022; (published online Aug 22.)https://doi.org/10.1016/S2666-7568(22)00167-2Summary Full Text Full Text PDF PubMed Google Scholar advances the field in two important ways. First, they used independent samples to train and test their brain age algorithm, a design feature essential to establishing the validity of their measurement. The training sample included 2001 adults aged 18–90 years, a much larger sample than in many previous studies. The testing sample consisted of 456 older adults from the UK National Survey of Health and Development (NSHD), a 1946 birth cohort followed prospectively for 7 decades. Second, Wagen and colleagues conducted comprehensive testing of a wide range of established AD–ADRD risk factors, including genetics, childhood cognitive ability, education, occupation, mid-life and later-life cardiovascular and cerebrovascular disease, physical functioning, and peripheral and brain biomarkers of AD–ADRD risk. This comprehensive approach makes it possible to evaluate which aspects of AD–ADRD aetiology might be captured by brain age. Wagner and colleagues’ key findings are that brain age was unrelated to measures of genetic and early-life risk, and to contemporaneous measures of physical functioning and brain amyloid deposition. In contrast, brain age did show expected relationships with prospective measures of neurodegeneration, including whole-brain and hippocampal atrophy, white-matter hyperintensities, late-life cognitive functioning, and cerebrovascular and cardiovascular disease. These findings suggest that the current generation of brain age measures are sensitive to vascular risk factors and volumetric anatomical changes, but might not record life-course accumulations of risk. The next steps in brain age research will require the same rigorous and comprehensive approach to algorithm development and evaluation applied in this study, joined with a commitment to new strategies for the inclusion of diverse research participants. The burden of AD–ADRD arises from causes such as structural racism that are not represented in conventional clinical samples or historical birth cohorts from Europe.7Babulal GM Quiroz YT Albensi BC et al.Perspectives on ethnic and racial disparities in Alzheimer's disease and related dementias: update and areas of immediate need.Alzheimers Dement. 2019; 15: 292-312Summary Full Text Full Text PDF PubMed Google Scholar, 8Glymour MM Manly JJ Lifecourse social conditions and racial and ethnic patterns of cognitive aging.Neuropsychol Rev. 2008; 18: 223-254Crossref PubMed Scopus (224) Google Scholar Recruitment of research participants should strive to represent (or over-represent) the populations subject to greater burdens of AD–ADRD risk factors, including environmental toxicant exposures, social stressors, trauma, and violence. These populations include racialised and minoritised groups, groups with reduced access to education and economic opportunity, sexual and gender minorities, and people who have greater traumatic brain injury including athletes and survivors of interpersonal violence.9Gilmore-Bykovskyi A Croff R Glover CM et al.Traversing the aging research and health equity divide: toward intersectional frameworks of research justice and participation.Gerontologist. 2022; 62: 711-720Crossref PubMed Scopus (8) Google Scholar Equipped with measures developed from samples that represent populations with high AD–ADRD burden, a next generation of brain age research can take a consequentialist approach10Galea S An argument for a consequentialist epidemiology.Am J Epidemiol. 2013; 178: 1185-1191Crossref PubMed Scopus (117) Google Scholar that centres science on improving health. In a world marred by structural racism with catastrophic consequences for the healthy ageing of individuals and communities, this approach means using longitudinal study designs to establish if changes in exposures or interventions can meaningfully change brain age, including in younger adults for whom AD–ADRD can still be prevented. Funders and investigators should prioritise the development of data resources that make such consequentialist research possible. The value of biomarkers like brain age is that they enable research on AD–ADRD in new settings, including studies of programmes and policies that modify structural determinants of health. The field should hold proposed new biomarkers of brain ageing to this translational standard. We declare no competing interests. Life course, genetic, and neuropathological associations with brain age in the 1946 British Birth Cohort: a population-based studyBrain-PAD was associated with cardiovascular risk, and imaging and biochemical markers of neurodegeneration. These findings support brain-PAD as an integrative summary metric of brain health, reflecting multiple contributions to pathological brain ageing, and which might have prognostic utility. Full-Text PDF Open Access