184 Chapter 8 GENERAL DISCUSSION Methodological considerations Subjective cognitive decline All studies in this thesis were conducted in individuals with SCD. Individuals with SCD report the subjective experience of cognitive deterioration, while objectively, their cognitive performance is within normal limits (1). SCD can be caused by many different factors. In a subset of individuals, it can be the first expression of an underlying neurodegenerative disease (2, 3). As we found in chapter 2 however, most individuals will not progress to MCI or dementia, making an underlying neurodegenerative disease as cause of cognitive complaints unlikely for the majority. In order to correctly identify individuals at risk, it is important to study risks of future cognitive deterioration and dementia in this group of individuals. However, there is a certain degree of heterogeneity in the characterization of SCD across SCD cohorts, which potentially limits generalizability of the results. The method of recruitment of SCD patients is important. Indeed, it is thought that individuals who visit a physician because of their cognitive complaints differ from individuals who report SCD on a questionnaire, yet did not visit a physician for these complaints. As such, it matters whether an SCD cohort is memory clinic based or population based (1, 4). Previous studies show memory clinic setting besides older age were associated with a higher frequency of amyloid positivity (5). Furthermore, there are variations in the definitions of SCD used across different cohorts. In our own center, individuals are generally referred by their general physician or medical specialist because of their cognitive complaints. When they perform within normal limits on an extensive neuropsychological examination, the complaints are judged to be of a subjective nature, hence they receive the label of SCD (2, 6). In other cohorts, SCD are defined as cognitively normal individuals reporting memory complaints on a questionnaire, or expressing subjective complaints and worries to a physician (7, 8). For these reasons, findings in our relatively young memory clinic based population are not directly generalizable to other (community based) populations. Longitudinal study design We used longitudinal data in all chapters, which enabled us to investigate associations between biomarkers and longitudinal cognitive trajectories in addition to crosssectional associations. At presentation, individuals with SCD are cognitively normal by definition, which means that large variability in cognitive performance at baseline is not expected. As a result, we hardly found any cross-sectional associations between
RkJQdWJsaXNoZXIy MjY0ODMw