Science1 publisher2 min readPublished
Neighborhood deprivation tracks with older-looking brains in 2,826 clinical MRIs
Radiologists at the University of Wisconsin matched 2,826 routine brain scans to a census-based deprivation index and found larger brain-age gaps, smaller total volumes and more white matter damage in the most deprived areas.
The Scientist · Science desk
What happened
- Researchers at the University of Wisconsin School of Medicine and Public Health analyzed 2,826 clinical brain MRI scans from patients aged 18 to 96 in a retrospective study published in Radiology.
- Those same residents of highly deprived areas showed greater white matter hyperintensity volume, which radiologists read as an indicator of chronic microvascular damage.
- The scans came from six months of routine imaging, January to June 2024, at University of Wisconsin Hospitals and Clinics and community healthcare partners.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint Every patient counted had a clinical reason to be scanned, so the result cannot be converted into a risk rate for the general population of a deprived ZIP code.
- decision A health system weighing whether to put address-derived deprivation scores into imaging risk models needs a magnitude, and the reported summary gives only the direction.
- capability The inputs are a patient address and a volumetric MRI pipeline, both of which most large health systems already hold, so other archives can be checked without collecting new data.
- contradiction The announcement claims an imprint at the molecular and cellular level while the three measurements are macroscopic imaging numbers, which changes how far the biological claim can be carried.
A brain age gap is a model's output. Software trained on structural MRI features predicts an age from the scan, and the gap is that prediction minus the patient's chronological age [12]. In the Wisconsin cohort that gap was larger among patients from the most deprived areas, and their total brain volume, normalized to intracranial volume, was smaller [5][8]. Elevated brain age has been tied in earlier work to cardiovascular disease, cognitive impairment and psychiatric disorders [13].
The account of the study gives the direction of each finding but no size for the brain age gap in years or the volume difference, and it does not identify the age-prediction model [18]. What follows depends on how large the difference is. A three-month difference in predicted brain age across the full range of the deprivation index is a finding for the literature. Three years would be worth acting on in a clinic.
Routine care supplied the sample: 2,826 scans over six months, about 471 a month [1][3][17], with women accounting for 1,732 of the patients, or 61% [2][16]. Everyone counted had a clinical reason for a brain MRI. Excluding scans with visible neurological disease and adjusting for age and sex tightens the comparison [8]. Who reached the scanner is a separate question. Access and referral patterns are plausibly related to deprivation themselves, and a retrospective design [14] cannot separate that from biology.
The exposure measure is the Area Deprivation Index, built at census block level from the 2023 American Community Survey out of 17 markers of income, housing quality, employment and educational attainment [7]. Patients were mapped to it by ZIP code [7]. Those are two different units of geography, and I would expect the coarser one to blur a real association.
Senior author John-Paul J. Yu said, "We've shown very robust, measurable metrics that our living environment has an outsized impact on the brain at the molecular and cellular level" [10]. The three reported metrics are a model-predicted age, a normalized total volume, and a white matter hyperintensity volume [5][6]. None of them images a cell. Hyperintensity burden is read as a marker of chronic stress-induced microvascular damage [6], an inference from millimetre-scale contrast on a scan.
Yu also said, "We've long suspected that your living environment affects your brain, but until now it's not been proven at scale in clinical populations" [9]. RSNA, which published the announcement, calls the work the first to establish a scaled link between environment and neuroimaging biomarkers in a real-world clinical cohort [15]. The paper is in Radiology [4].
What to watch
- The Radiology paper's effect sizes and confidence intervals for brain age gap, total volume and hyperintensity burden.
- Whether the association survives geocoding patients at census block level instead of by ZIP code.
- Repeat scans on the same patients, testing whether the brain age gap widens over time in the most deprived areas.