Midlife sensory and motor measures among best predictors in parsimonious models of long-term cognitive decline and incidence of cognitive impairment in aging adults.

A parsimonious LASSO model identified six top predictors of 10-year cognitive decline in midlife adults.

• The top cognitive decline predictors were age, income, fine-motor skills, olfaction, peripheral artery disease, and serum neurofilament light chain protein (NfL).
• The model yielded an AUC of 0.80 (95% confidence interval [0.76-0.83]).
• The study included N=1,529 Beaver Dam Offspring Study participants with mean age 49 years, 54% women.
• Baseline measures were used to predict 10-year cognitive outcomes using LASSO logistic regression.

A separate parsimonious LASSO model identified six top predictors of 10-year cognitive impairment incidence in midlife adults.

• The top cognitive impairment predictors were sex, fine-motor skills, olfaction, self-rated vision, alcohol consumption, and NfL.
• The model yielded an AUC of 0.73 (95% confidence interval [0.69-0.77]).
• Cognitive impairment prediction used a different set of predictors than cognitive decline, notably including self-rated vision and alcohol consumption instead of age, income, and peripheral artery disease.

Sensory measures including olfaction and self-rated vision appeared among the best midlife predictors of long-term cognitive outcomes.

• Olfaction was selected as a top predictor in both the cognitive decline and cognitive impairment models.
• Self-rated vision was selected as a top predictor specifically in the cognitive impairment model.
• These sensory measures were identified through LASSO variable selection from a broader set of health measures assessed at baseline.

Fine-motor skills were selected as a top predictor in both the cognitive decline and cognitive impairment models.

• Fine-motor skills appeared in both parsimonious models, making it one of the most consistent predictors across outcomes.
• Fine-motor skills were assessed at baseline when participants had a mean age of 49 years.
• Motor function was identified alongside sensory measures as among the best midlife predictors of neurodegeneration and cognitive decline.

Serum neurofilament light chain protein (NfL) was selected as a top predictor in both cognitive decline and cognitive impairment models.

• NfL appeared in both final parsimonious models, representing a biomarker of neurodegeneration among the best predictors.
• NfL was measured at midlife baseline (mean age 49 years) and predicted outcomes 10 years later.
• NfL was one of two variables (along with fine-motor skills and olfaction) selected in both models.

The study used the Beaver Dam Offspring Study longitudinal cohort with a 10-year follow-up period.

• N=1,529 participants were included with mean age 49 years and 54% women.
• Several health measures were assessed at baseline and 10-year cognitive decline and cognitive impairment were assessed as outcomes.
• Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression models were used to construct parsimonious risk prediction models.
• The authors noted that external validation in other studies would be needed before results could be used to identify those at risk.