Can we predict who will develop Parkinson’s disease before symptoms develop?

DuBois Bowman, a professor of biostatistics at Columbia’s Mailman School of Public Health, is creating new methods of analyzing brain imaging data to predict which patients will develop Parkinson’s disease.

The disease begins with the death of dopamine-producing neurons in the area of the brain controlling movement. Symptoms do not manifest immediately—there is a delay before outward signs emerge. To enable early detection, Parkinson’s researchers are attempting to create tools that spot neurodegeneration before external symptoms appear.

Bowman has developed an advanced statistical learning technique to analyze brain imaging data, drawing from a large cohort of patients. This imaging data captures detailed information about brain structure and function, and Bowman’s analysis allows him to identify biomarkers that indicate early signs of the disease.

When patients are diagnosed sooner, as Bowman’s work will make possible, they can begin a regimen of neuroprotective treatments that slow the disease's progression. By postponing the onset of damaging motor symptoms, they can live longer and healthier lives.