Machine learning suggests polygenic contribution to cognitive dysfunction in amyotrophic lateral sclerosis (ALS)

Amyotrophic lateral sclerosis (ALS) is a multi-system disorder characterized by progressive muscular weakness and, in addition, cognitive/behavioral dysfunction in nearly 50% of patients. The mechanisms underlying risk for cognitive dysfunction, however, remain elusive. Using sparse canonical correlation analysis (sCCA), an unsupervised machine-learning technique, we observed that 26 single nucleotide polymorphisms collectively associate with baseline cognitive performance in 330 ALS patients from the multicenter Clinical Research in ALS and Related Disorders for Therapeutic Development (CReATe) consortium. We demonstrate that a polygenic risk score derived from sCCA relates to longitudinal cognitive decline in the same cohort, and also to in vivo cortical thinning (N=80) and post mortem burden of TDP-43 pathology in the middle frontal and motor cortices (N=55) in independent validation cohorts of patients with sporadic ALS. Our findings suggest that common genetic polymorphisms contribute to the manifestation of cognitive dysfunction and disease vulnerability in a polygenic manner in ALS.