Brains affected by Alzheimer's disease and mild cognitive impairment appear older than their owners' birth certificates suggest, and by a wider margin than brains affected by addiction or psychiatric illness. That is the central finding of research led by Shile Qi of the Nanjing University of Aeronautics and Astronautics, published July 21 in the open access journal PLOS Medicine.
The team used a measure called the predictive age difference, or PAD, which compares a person's actual age with the age predicted from structural MRI scans of their brain. A positive PAD means the brain looks older than it should. Drawing on several brain imaging banks, the researchers compared scans from 45,900 controls with 2,698 people who had conditions including ADHD, autism spectrum disorder, alcohol or tobacco addiction, Alzheimer's disease, mild cognitive impairment, schizophrenia, bipolar disorder or major depressive disorder. Neurodegenerative conditions topped the list for accelerated aging; addiction and psychiatric disorders also showed elevated PAD, while people with ADHD or autism were indistinguishable from controls.
The regional picture is where the study gets more interesting for anyone hoping to use brain imaging diagnostically. The prefrontal cortex looked prematurely aged across nearly all the disorders studied, but the patterns diverged beyond that. Dementia was tied to accelerated aging in the frontal and occipital cortex, psychiatric disorders to the frontal and temporal lobes, and addiction to the default mode network, the salience network, the putamen and the thalamus. The researchers also identified differences in gene expression that tracked with particular conditions.
Distinct regional signatures raise the possibility that brain age maps could eventually serve as biomarkers, helping distinguish between conditions that can look similar in the clinic or occur together. The authors are careful to note the limits: these findings are correlational rather than causal, and conditions such as addiction and psychiatric illness frequently co-occur, which complicates efforts to attribute a given pattern to a single diagnosis.