Dementia Disease Initiation: Demographically adjusted norms based on Scandinavian samples and comparison with published norms from North America

Abstract

<p><i>Background:</i> Norms are necessary for interpreting neuropsychological test scores. In Norway and Sweden, there is a lack of local norms on neuropsychological tests. Consequently, Scandinavians are frequently norm-referenced to participants from North America tested decades prior. However, factors such as time of assessment, language, population differences in average performance on cognitive tests, and cultural differences in education and health related factors are sources of heterogeneity in norms. Thus, there is a need for neuropsychological test norms based on Scandinavian populations. In this thesis, the primary aim was to develop norms on the Trail Making Test (TMT), Rey Auditory Verbal Learning Test (RAVLT) and Delis-Kaplan Executive Functions System Color-Word Interference Test (D-KEFS CWIT) based on Scandinavian samples of healthy adults. The secondary aim was to assess frequently used norms from North America in Scandinavian samples. <p><i>Methods:</i> Based on healthy adult participants from Norway and Sweden, we modelled pertinent effects of age, education, and sex using regression-based norming procedures in Papers 1-3. We provide normative calculators to aid clinicians and researchers in the computation of normed scores. We assessed overall distributions and performed multiple regression analyses with demographical variables as predictors to assess whether North American norms adequately adjusted for demographical variables in T-scores. We calculated test-retest reliability indices (Paper 2 & 3) and assessed differences in the estimated rate of impairment using local norms and North American norms (Paper 3). <p><i>Results and conclusions:</i> In the Scandinavian samples, age, education, and sex was significantly related to scores on TMT, RAVLT and D-KEFS CWIT. Results indicated adequate reliability for the most prominent subtests. Compared to previous studies, education appeared to explain less variance in scores in the Scandinavian samples. Compared to previously published norms from North America, results indicated less difference between young and old participants, participants with high and low educational attainment, and less difference between men and women on RAVLT. These discrepancies may be in part due to cultural differences in education and health-related factors. In future research, there is a need for harmonized representative samples to assess whether these results represent generalizable differences characterizing healthy Scandinavians. The North American norms were too lenient for elderly individuals and/or those with low educational attainment, and too strict for individuals with high educational attainment and/or younger individuals. Results from Paper 3 suggest that this may have clinical implications for the accurate assessment of cognitive functions in Scandinavian samples.