COGNITIVE PERFORMANCE AND SUBJECTIVE COGNITIVE DECLINE: DATA FROM THE HEINZ NIXDORF RECALL STUDY

Abstract

Subjective cognitive decline (SCD) is defined as subjective cognitive complaints in the absence of objective cognitive impairment. SCD is proposed to be the first clinical symptom of preclinical dementia even before onset of mild cognitive impairment (MCI) (Molinuevo et al.,2016) and to predict cognitive decline (Burmester et al.,2016; Mitchell et al.,2014). We compared the cognitive performance of participants with SCD either with or without additional cognitive concerns to a control group of participants without SCD in a subsample of cognitively normal participants of an established population-based cohort. We studied 2899 participants (68.5±7.3 years, 48.7% men) with complete data of a short cognitive performance assessment from the third examination of the Heinz Nixdorf Recall Study. Participants with objective cognitive impairment and/or elevated depressive symptoms were excluded from the analyses (final analysis sample n=1107). SCD was assessed with the question: “In comparison to two years ago would you rate your memory function as better, same or worse?”. The answer “better” or “same” was categorized as no SCD (control, n=793) and “worse” as SCD. Participants with SCD were then further categorized based on their cognitive concerns assessed with the question “Does this worry you?”: SCD with additionally reported concerns (SCD+C, n=150) and SCD without additionally reported concerns (SCD-C, n=164). A global cognitive score was obtained by adding the z-scores of all eight cognitive tests, with higher z-scores indicating better performance. Group comparison of global cognitive performance was performed with analysis of covariance (mean ± standard deviation (SD), B-value (95% confidence interval (CI)); unadjusted and adjusted for age, gender, education, APOE genotype). We found significant group differences although SCD+C showed the lowest performance compared to SCD-C and no SCD. There were no significant associations in the fully adjusted model but the SCD+C group again showed the lowest cognitive performance with a strong trend towards significance in comparison to the no SCD group (Table 1). In our population-based cohort study, cognitively normal participants differed in their global cognition regarding SCD when they expressed additional concerns but the association vanished in the fully adjusted model. Therefore, SCD did not seem to independently predict memory impairment.