Models Constructed to Predict Decline Over Time in MCI, Mild Dementia

Hypothetical patients with MCI, baseline MMSE of 28, CSF Aβ1-42 of 925 pg/mL predicted to reach MMSE of 20 after 6.0 years
Alzheimer's frailty dementia
Adobe Stock
Published on
Updated on

WEDNESDAY, July 10, 2024 (HealthDay News) -- Models can be constructed that predict cognitive decline using Mini-Mental State Examination (MMSE) scores over time for mild cognitive impairment (MCI) and mild dementia, according to a study published online July 10 in Neurology.

Pieter J. van der Veere, M.D., from the VU University Medical Center in Amsterdam, and colleagues constructed clinically applicable prediction models of cognitive dementia in amyloid-positive patients with MCI or mild dementia from the Amsterdam Dementia Cohort. Linear mixed modeling was used to predict the MMSE over time.

Data were included for 961 participants: 310 had MCI and 651 had mild dementia with 4 ± 2 MMSE measurements over two years. The researchers found that over time, there were increases in cognitive decline rates for both MCI and mild dementia. Backward selection retained age, sex, cerebrospinal fluid (CSF) β-amyloid (Aβ)1-42, and pTau as time-varying effects altering the MMSE trajectory for MCI. Time-varying effects of Aβ1-42, age, APOE ε4, and baseline MMSE were retained for mild dementia. In internal cross-validation, R2 was 0.15 and 0.26 for the MCI and mild dementia models, respectively. For a hypothetical patient with MCI, a baseline MMSE of 28, and CSF Aβ1-42 of 925 pg/mL, an MMSE of 20 was predicted after 6.0 years and after 8.6 years with a hypothetical treatment.

"There is a need among patients and care partners for prognostic information on their cognitive trajectory," the authors write. "These models can provide such information, although our results also emphasize that the heterogeneity in cognitive trajectories can only be partially captured."

Several authors disclosed ties to the biopharmaceutical industry.

Abstract/Full Text

Related Stories

No stories found.
logo
www.healthday.com