BASE RATES OF IMPAIRED SCORES ARGUE AGAINST CORRECTION FOR MEDICAL CONDITIONS WITHIN A BRIEF NEUROPSYCHOLOGICAL BATTERY

Abstract

Normative data are usually created by making demographic adjustments based on correlations between single neuropsychological tests with demographic variables such as age and education. However, this traditional approach has some limitations. For example, it does not detail the implications for clinical practice such as likelihood of misclassification of cognitive impairment. It also does not elucidate the impact on decision making when using a neuropsychological battery. We hypothesize that baserates analyses, particularly, differential baserates of impaired scores between theoretical and actual baserates, would be better suited to create demographic adjustments within normative data. Differential baserates empirically detail the potential clinical implications of failing to create an appropriate normative group. Thus the aim of the present study is to explore whether adjustments for age and medical conditions are warranted based on differential baserates of spuriously impaired scores. The present study uses telephone-administered data collected from over 20,000 participants of the Canadian Longitudinal Study on Aging (CLSA). The CLSA has recruited a large national sample of approximately 50,000 French- and English-speaking men and women between the ages of 45 and 85 across 10 Canadian provinces; over 20,000 responded through telephone interview and 30,000 to an in-home interview and physical assessments at a data collection site. Differential baserates are calculated using the results of a short neuropsychological battery administered by telephone to a neurologically healthy sample (but only the English-speaking sub-sample is used in the present research). Theoretical baserates underestimate the frequency of impaired scores in older adults, providing an evidence base for the creation of age corrected normative data. Interestingly, several medical conditions such as cancer, chronic obstructive pulmonary disease, cardiovascular (e.g., high blood pressure, angina, acute myocardial infarction), hormonal (e.g., hypo- and hyperthyroidism) and metabolic (e.g., diabetes) conditions are not associated to differential baserates of impaired scores. Despite a small magnitude correlation between number of medical conditions and each neuropsychological variable, normative adjustments for number of medical conditions does not appear warranted.