Relationship between the number of subjective cognitive decline domains involved and the neuroimaging biomarkers of Alzheimer’s disease: The SILCODE Study: Neuroimaging / Optimal neuroimaging measures for tracking disease progression

Xiaoqi Wang,Yu Sun,Qiuhui Bi,Ni Shu,Ying Han

DOI: https://doi.org/10.1002/ALZ.038513

2020-01-01

Alzheimer s & Dementia

Abstract:Background Subjective cognitive decline (SCD), one of the earliest manifestations in preclinical Alzheimer’s disease (AD), may involve one or more cognitive domains including memory, language, executive and others. Converging evidence has revealed the association between SCD and AD biomarkers such as amyloid deposition and neurodegeneration alterations. However, most of the studies about SCD limited to memory domain and whether the increasing number of SCD domains involved is related to the neuroimaging biomarkers of AD remains largely unknown. Method Multimodal imaging techniques, including 18 F‐florbetapir PET (AV45‐PET), 18 F‐fluorodeoxyglucose PET (FDG‐PET), structural and resting‐state functional MRI, were performed in 63 SCD subjects and the regional imaging measures of 12 selected regions of interests (ROIs) were calculated, namely standard uptake value ratio (SUVr) of AV45‐PET and FDG‐PET, cortical thickness and fractional amplitude of low‐frequency fluctuation (fALFF) values. The total number of cognitive domains with reported complaints and the severity of memory complaints were assessed for each SCD subject. Partial correlation analyses were conducted to test the relationship between these SCD features and the neuroimaging markers of AD. Result SCD who reported multidomain complaints (multidomain SCD, md‐SCD, n=34) had greater amyloid burden, elevated fALFF values and thickened cortical thickness in the posterior cingulate gyrus and middle temporal gyrus than SCD who reported merely memory complaints (single domain SCD, sd‐SCD, n=29) (Figure 1). Even when regarded as continuous variable, the number of SCD domains involved was positively correlated with the degree of amyloid load in posterior cingulate, fALFF values in middle temporal, posterior and anterior cingulate and cortical thickness in precuneus (Figure 2). Additionally, the severity of memory complaints measured by the SCD questionnaire (SCD‐Q9) was only related to cortical thickness, but no significant association with other imaging markers was found (Figure 3). Conclusion An increased number of SCD domains involved was correlated with amyloid accumulation and may also be indicative of an underlying compensatory mechanism medicated by brain structures and functions alterations. Therefore, the reported number of SCD domains may be a potential valuable feature of SCD‐plus for AD risk and is promising for predicting AD pathologies.