Diagnostic accuracy of neuropsychological tests for classification of dementia

Although numerous studies have shown that each neuropsychological test is effective for diagnosing mild cognitive impairment (MCI) or Alzheimer’s disease (AD), studies comparing diagnostic accuracies of various neuropsychological tests are relatively rare and practical cutoff values are not available. The present study aimed to investigate the validity of neuropsychological tests and develop cutoff values for each in differentiating healthy control (HC), MCI and AD groups. A total of 84 HC, 187 with MCI and 195 with AD were evaluated by the selected seven neuropsychological tests using receiver operating characteristic (ROC) curve analysis. Logical Memory (LM) delayed recall (cutoff, 7) and Rey Auditory Verbal Learning Test (RAVLT) delayed recall (cutoff, 6) were effective for differentiating HC from MCI. To distinguish MCI and AD, Rey Osterrieth Complex Figure Test (ROCFT) 3 mindelayed recall (cutoff, 6) and LM immediate recall (cutoff, 4) were excellent. Delayed recall of verbal materials, as indexed by LM and RAVLT was sensitive for discriminating MCI from HC. Handling visual memory traces, as indexed by ROCFT and immediate verbal information by LM were sensitive for differentiating MCI and AD. Neurology Asia 2016; 21(1) : 47 – 54 Address Correspondence to: Daisuke Ito M.D. Ph.D., Department of Neurology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. Tel: +81-3-5363-3788; E-mail: [email protected] INTRODUCTION Early and accurate detection of dementia through screening methods may benefit risk assessment and care management, and may eventually contribute to substantial cost savings. Screening methods should also be important in the future for the expected disease-modifying and preventive treatments for Alzheimer’s disease (AD). One of the basic problems is that there are no standard neuropsychological tests to differentiate among normal aging, mild cognitive impairment (MCI), and AD. Despite ongoing efforts in recent years to evaluate proper cognitive tests that help detect or diagnose the cognitive impairment continuum from normal aging through subjective memory complaints and MCI to AD, establishing appropriate classification of cognitive impairment with neuropsychological tests has remained challenging. The mini mental state examination (MMSE) is the most popular screening test used to assess patients with cognitive impairment in the memory clinic. The MMSE is administered in a short time, and is acceptable to both patients and healthcare workers. However, MMSE has some limitations for detecting MCI and also is not sensitive for detecting subtle declines in cognitive function. It is unclear how effective other neuropsychological tests are to differentiate healthy controls (HC), MCI and AD patients; there is a need to identify suitable and effective, validated neuropsychological screening measures. Although numerous studies have shown that each neuropsychological test is effective for diagnosing MCI or AD, studies comparing diagnostic accuracies of various neuropsychological tests are relatively rare and practical cutoff values are not available. The aim of the present study was, therefore, to determine which of the selected neuropsychological tests in current use are more effective to distinguish HC, MCI, and AD. This study identified neuropsychological tests effective for classification of HC, MCI, and AD through an analysis of diagnostic accuracy, and provided practical cutoff values for each test for diagnostic use. Neurology Asia March 2016