A statistical algorithm for detecting cognitive plateaus in Alzheimer's disease

Hyonggin An, Roderick J.A. Little, Andrea Bozoki

Research output: Contribution to journalArticle

Abstract

Repeated neuropsychological measurements, such as mini-mental state examination (MMSE) scores, are frequently used in Alzheimer's disease (AD) research to study change in cognitive function of AD patients. A question of interest among dementia researchers is whether some AD patients exhibit transient "plateaus" of cognitive function in the course of the disease. We consider a statistical approach to this question, based on irregularly spaced repeated MMSE scores. We propose an algorithm that formalizes the measurement of an apparent cognitive plateau, and a procedure to evaluate the evidence of plateaus in AD using this algorithm based on applying the algorithm to the observed data and to data sets simulated from a linear mixed model. We apply these methods to repeated MMSE data from the Michigan Alzheimer's Disease Research Center, finding a high rate of apparent plateaus and also a high rate of false discovery. Simulation studies are also conducted to assess the performance of the algorithm. In general, the false discovery rate of the algorithm is high unless the rate of decline is high compared with the measurement error of the cognitive test. It is argued that the results are not a problem of the specific algorithm chosen, but reflect a lack of information concerning the presence of plateaus in the data.

Original languageEnglish
Pages (from-to)779-789
Number of pages11
JournalJournal of Applied Statistics
Volume37
Issue number5
DOIs
Publication statusPublished - 2010 May 1

Fingerprint

Alzheimer's Disease
Dementia
Linear Mixed Model
Repeated Measurements
Measurement Error
Alzheimer's disease
Simulation Study
Evaluate

Keywords

  • Alzheimer's disease
  • Cognitive plateau
  • False discovery rate
  • Linear mixed model
  • Longitudinal data
  • Nonlinear model

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

A statistical algorithm for detecting cognitive plateaus in Alzheimer's disease. / An, Hyonggin; Little, Roderick J.A.; Bozoki, Andrea.

In: Journal of Applied Statistics, Vol. 37, No. 5, 01.05.2010, p. 779-789.

Research output: Contribution to journalArticle

An, Hyonggin ; Little, Roderick J.A. ; Bozoki, Andrea. / A statistical algorithm for detecting cognitive plateaus in Alzheimer's disease. In: Journal of Applied Statistics. 2010 ; Vol. 37, No. 5. pp. 779-789.
@article{5fdf5dfcc2b440f2a7a2a6a6bac64994,
title = "A statistical algorithm for detecting cognitive plateaus in Alzheimer's disease",
abstract = "Repeated neuropsychological measurements, such as mini-mental state examination (MMSE) scores, are frequently used in Alzheimer's disease (AD) research to study change in cognitive function of AD patients. A question of interest among dementia researchers is whether some AD patients exhibit transient {"}plateaus{"} of cognitive function in the course of the disease. We consider a statistical approach to this question, based on irregularly spaced repeated MMSE scores. We propose an algorithm that formalizes the measurement of an apparent cognitive plateau, and a procedure to evaluate the evidence of plateaus in AD using this algorithm based on applying the algorithm to the observed data and to data sets simulated from a linear mixed model. We apply these methods to repeated MMSE data from the Michigan Alzheimer's Disease Research Center, finding a high rate of apparent plateaus and also a high rate of false discovery. Simulation studies are also conducted to assess the performance of the algorithm. In general, the false discovery rate of the algorithm is high unless the rate of decline is high compared with the measurement error of the cognitive test. It is argued that the results are not a problem of the specific algorithm chosen, but reflect a lack of information concerning the presence of plateaus in the data.",
keywords = "Alzheimer's disease, Cognitive plateau, False discovery rate, Linear mixed model, Longitudinal data, Nonlinear model",
author = "Hyonggin An and Little, {Roderick J.A.} and Andrea Bozoki",
year = "2010",
month = "5",
day = "1",
doi = "10.1080/02664760902889999",
language = "English",
volume = "37",
pages = "779--789",
journal = "Journal of Applied Statistics",
issn = "0266-4763",
publisher = "Routledge",
number = "5",

}

TY - JOUR

T1 - A statistical algorithm for detecting cognitive plateaus in Alzheimer's disease

AU - An, Hyonggin

AU - Little, Roderick J.A.

AU - Bozoki, Andrea

PY - 2010/5/1

Y1 - 2010/5/1

N2 - Repeated neuropsychological measurements, such as mini-mental state examination (MMSE) scores, are frequently used in Alzheimer's disease (AD) research to study change in cognitive function of AD patients. A question of interest among dementia researchers is whether some AD patients exhibit transient "plateaus" of cognitive function in the course of the disease. We consider a statistical approach to this question, based on irregularly spaced repeated MMSE scores. We propose an algorithm that formalizes the measurement of an apparent cognitive plateau, and a procedure to evaluate the evidence of plateaus in AD using this algorithm based on applying the algorithm to the observed data and to data sets simulated from a linear mixed model. We apply these methods to repeated MMSE data from the Michigan Alzheimer's Disease Research Center, finding a high rate of apparent plateaus and also a high rate of false discovery. Simulation studies are also conducted to assess the performance of the algorithm. In general, the false discovery rate of the algorithm is high unless the rate of decline is high compared with the measurement error of the cognitive test. It is argued that the results are not a problem of the specific algorithm chosen, but reflect a lack of information concerning the presence of plateaus in the data.

AB - Repeated neuropsychological measurements, such as mini-mental state examination (MMSE) scores, are frequently used in Alzheimer's disease (AD) research to study change in cognitive function of AD patients. A question of interest among dementia researchers is whether some AD patients exhibit transient "plateaus" of cognitive function in the course of the disease. We consider a statistical approach to this question, based on irregularly spaced repeated MMSE scores. We propose an algorithm that formalizes the measurement of an apparent cognitive plateau, and a procedure to evaluate the evidence of plateaus in AD using this algorithm based on applying the algorithm to the observed data and to data sets simulated from a linear mixed model. We apply these methods to repeated MMSE data from the Michigan Alzheimer's Disease Research Center, finding a high rate of apparent plateaus and also a high rate of false discovery. Simulation studies are also conducted to assess the performance of the algorithm. In general, the false discovery rate of the algorithm is high unless the rate of decline is high compared with the measurement error of the cognitive test. It is argued that the results are not a problem of the specific algorithm chosen, but reflect a lack of information concerning the presence of plateaus in the data.

KW - Alzheimer's disease

KW - Cognitive plateau

KW - False discovery rate

KW - Linear mixed model

KW - Longitudinal data

KW - Nonlinear model

UR - http://www.scopus.com/inward/record.url?scp=77951138068&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951138068&partnerID=8YFLogxK

U2 - 10.1080/02664760902889999

DO - 10.1080/02664760902889999

M3 - Article

AN - SCOPUS:77951138068

VL - 37

SP - 779

EP - 789

JO - Journal of Applied Statistics

JF - Journal of Applied Statistics

SN - 0266-4763

IS - 5

ER -