Lower spinal bone density linked to faster cognitive decline in aging study

Lower bone density in the spine was associated with faster cognitive decline and changes in selected brain regions in a study published September 22 in Radiology. The findings add evidence that skeletal and neurological aging can occur together, raising the possibility that information already present in chest scans could help researchers identify overlapping health risks.

The work does not establish that osteoporosis causes dementia or that increasing bone density prevents memory loss. Its immediate contribution is an observational link between an accessible imaging measurement and subsequent changes in brain structure and cognitive performance.

Following bone and brain health

Researchers led by Sara Momtazmanesh at Johns Hopkins University conducted a secondary analysis of the prospective Multi-Ethnic Study of Atherosclerosis, known as MESA. The analysis included 715 participants with a median age of 69 who had bone-density measurements, brain MRI and cognitive testing.

A deep learning algorithm estimated volumetric bone mineral density in thoracic vertebrae from noncontrast chest CT scans. The investigators then examined whether lower baseline density tracked with changes in cognition and two MRI measures: white matter hyperintensities and fractional anisotropy, which reflects properties of white matter microstructure.

The underlying observations spanned September 2016 through March 2024. Statistical models accounted for demographics, APOE genetic risk, cardiometabolic factors, lifestyle and medications. Available follow-up differed by endpoint: 639 participants contributed to the global cognitive composite analysis, while approximately 400 contributed to each longitudinal MRI analysis.

A regional signal, with an important qualification

Lower baseline bone density was associated with faster decline in global cognitive performance. Regional imaging analyses also linked it to greater white matter hyperintensity accumulation in the corpus callosum and declining fractional anisotropy in the anterior limb of the internal capsule.

The broader finding for total white matter fractional anisotropy was less robust. Although initially statistically significant, it did not remain significant after correction for multiple comparisons. That distinction limits any claim that lower bone density consistently predicted deterioration throughout the brain’s white matter.

The authors characterized the regional changes as modest. These findings concern group-level associations with test scores and imaging measurements; they do not provide an individual forecast of dementia.

What existing scans might contribute

The imaging approach builds on earlier work. A separate MESA study published in March 2025 evaluated automated three-dimensional vertebral measurements from conventional chest CT. It found that incorporating these measurements improved vertebral-fracture prediction compared with a fracture-risk tool without bone-density information.

That earlier result supports investigating additional uses for existing scans. It does not validate the same measurements as a clinical test for future cognitive decline. Measuring bone reliably and demonstrating that a measurement improves neurological care are separate steps.

In its September 22 announcement, RSNA described the researchers’ interest in extracting bone information from chest scans obtained for other clinical purposes. Whether that information improves decisions about brain health remains to be tested.

Shared risks remain an explanation

The researchers suggested that metabolic and hormonal factors could contribute to both skeletal and brain changes. Those possibilities remain interpretations rather than mechanisms established by this analysis.

Adjustment for measured risk factors cannot remove all confounding. The smaller groups with repeat imaging also leave questions about selection and how well the findings apply beyond participants who completed follow-up.

For healthspan research, the useful next question is whether bone measurements add meaningful predictive information beyond established risk factors—and whether acting on that information improves outcomes. This study supports investigating those questions without establishing a new dementia-screening strategy or a cognitive benefit from osteoporosis treatment.