Advancing the Use of Brief Cognitive Tests - Establishing Norms, Clinically Relevant Changes and Predictive Models

Introduction: It is important to understand how to interpret and utilize cognitive
assessment results for diagnosis, treatment, and inclusion in clinical studies. As
treatments for neurodegenerative diseases advance, the need to identify cognitive
decline in its earliest stages is becoming increasingly important, both for the timely
initiation of treatment and for assessing the efficacy of interventions in clinical
trials. For early identification, accurate cognitive test cut-offs derived from a
suitable population are essential. It is also important to identify a clinically
meaningful change in cognitive test scores, which is essential when following
patients in clinic with repeated assessments, as well as when using cognition as an
outcome in clinical trials. This is especially relevant as clinical trials increasingly
feature novel composites of cognitive tests. We also need methods to predict which
individuals seeking healthcare are at high risk of progressing to dementia in the near
future and which individuals are at low risk.

Methods: Participants from The Malmö Food and Diet, BioFINDER-1, and the
Alzheimer’s Disease Neuroimaging Initiative studies have been included in this
thesis. These studies all include individuals with and without cognitive impairment,
facilitating research in early diagnostic strategies for cognitive decline.

Results: In this thesis, we established Swedish MoCA cut-offs for cognitive
impairment for the primary assessment of cognitive impairment. We presented a
new approach to establish normative data for brief cognitive assessments for
identifying early cognitive changes in preclinical dementias. We have also identified
potential minimal clinically important differences (MCIDs) for cognitively
unimpaired individuals and individuals with mild cognitive impairment on a range
of cognitive test outcomes. Furthermore, we explored methods to predict a
composite cognitive measure for predicting a cognitive decline and to predict
progression to dementia for those with mild cognitive symptoms. Finally, we
created a two-step prediction model for predicting overall dementia for individuals
with mild cognitive symptoms.

Discussion: In our ageing population with increasing education levels and various
comorbidities, it is important to update guidelines for test norms, MCIDs and
methods for predicting cognitive decline. This can aid in optimal management and
early treatment, including timely referral to specialized units for enhanced
diagnostics of high-risk patients.