Neuropsychological Assessment Reduces False Positives in Mild Cognitive Impairment

By Lisa D. Ravdin, PhD, ABPP

Associate Professor of Neuropsychology, Weill Cornell Medical College

Dr. Ravdin reports no financial relationships relevant to this field of study.

Synopsis: Traditional comprehensive neuropsychological testing has greater reliability, sensitivity, and specificity than bedside screening tests in the accurate diagnosis of mild cognitive impairment.

Source: Klekociuk SZ, et al. Reducing false positive diagnoses in mild cognitive impairment: The importance of comprehensive neuropsychological assessment. Europ J Neurol 2014;21:1330-1336.

Mild cognitive impairment (MCI) is a term initially used to describe reduced memory functioning as an intermediary stage between age-related memory changes and dementia. The classification of MCI has since been expanded to include non-memory cognitive changes, as well as various subtypes of presentation (i.e., single or multiple affected cognitive domains). Given the high level of function of those with MCI, neuropsychological assessment is better suited than bedside cognitive screening in terms of detection, distinguishing subtypes, and objectively measuring patterns of progression.

The aim of this study was to assess the ability of neuropsychological measures to predict true positive classification of MCI over time. Participants were examined at a baseline screening plus two additional time points approximately 9 months (± 3 months) and 11 months (± 1 month) later. Screening consisted of a battery of tests that differed in actual measures but tapped the same cognitive domains assessed at the follow-up visits. One hundred eighteen community-residing subjects had data from the three time points and were included in the analyses. MCI was defined as > 1.28 standard deviation (< 10th percentile) below the mean on one or more cognitive measures. Stability classifications were based on consistency of classifications across visits; subjects who met MCI criteria at all three time points or those who were initially unimpaired on screening but met criteria for MCI at visit 1 and 2 were given the stability classification of MCI. Those identified as MCI at screening, but performed within normal limits on subsequent visits, were classified as unimpaired, as were those who continually were unimpaired at all three test visits. Discriminant function analyses revealed test scores from multiple cognitive domains predicted group outcome, accounting for 83.9% of cases. A false-positive rate of 5.93% using this model was consistently lower than false-positive rates of misdiagnosis based on diagnostic criteria alone (23.73%).

The data suggest that memory testing in isolation is insufficient for recognizing MCI, and that reduced scores on tests of other higher order cognitive processes (i.e., executive function) enhance the ability to identify true positive cases. In fact, the strongest predictor of stability of MCI was a non-memory measure, Rapid Visual Processing, a formal test of sustained attention. Unlike other studies, the analyses did not examine different effects for those with single vs multiple domain MCI. Stability may vary depending on subtype, and others have shown the number of domains affected predicts outcome. Age is yet another factor not examined here that has been shown in other studies to be associated with lower diagnostic stability of MCI classification.

COMMENTARY

MCI is becoming a household word for educated medical consumers, but in practice there are discrepancies in the operational definition, with some clinicians considering the term to be virtually synonymous with early-stage dementia and others recognizing it as a risk factor for Alzheimer’s disease (AD) or generalized cognitive decline. There are likely a variety of etiologies of MCI, some of which may not necessarily be progressive in nature. The concept of "MCI due to AD" has been used to describe individuals with cognitive changes that are most likely secondary to pathophysiological processes associated with AD, and this group can theoretically be differentiated from those with a more static cognitive disturbance or the considerable number of cases that revert to normal. In practice, classification errors can arise due to clinical biases or insufficient methodology. "One test" methods have been criticized for their limited ability to identify cognitive decline, particularly in cases of subtle impairment, as is the case in MCI. The same can be said for bedside mental status screening that can show a ceiling effect and lack sensitivity in the high-functioning MCI patient. These data highlight that formal neuropsychological assessment is associated with fewer false-positive classifications of MCI. Of note, interpretation of scores on cognitive measures should account for the effects of age, education, and cultural factors on test performance. Efforts aimed at improving recognition and classification of MCI can have a significant impact on clinical care as well as better identification of appropriate individuals for future prevention trials.