Amyloid‐positive individuals with subjective cognitive decline present increased CSF neurofilament light levels that relate to hippocampal volume

• Published in: Alzheimer's & dementia Vol. 16
• Main Authors: Sánchez‐Benavides, Gonzalo, Suárez‐Calvet, Marc, Milà‐Alomà, Marta, Arenaza‐Urquijo, Eider M., Grau‐Rivera, Oriol, Operto, Greg, Gispert, Juan Domingo, Vilor‐Tejedor, Natalia, Sala‐Vila, Aleix, Crous‐Bou, Marta, González‐de‐Echávarri, José Maria, Minguillón, Carolina, Fauria, Karine, Simon, Maryline, Kollmorgen, Gwendlyn, Zetterberg, Henrik, Blennow, Kaj, Molinuevo, Jose Luis
• Format: Journal Article
• Language: English
• Published: 01.12.2020
• ISSN: 1552-5260; 1552-5279
• DOI: 10.1002/alz.045715

Background Neurofilament light chain (NfL) is a biomarker of neurodegeneration. While cerebrospinal (CSF) NfL is increased in individuals with AD dementia as compared to controls and mild cognitive impairment, studies in cognitively unimpaired individuals with subjective cognitive decline (SCD) are scarce. SCD may represent the first clinical symptom of AD. We aimed at investigating the association between presence of SCD and CSF NfL levels, the interaction between SCD and amyloid‐β (Aβ) status on NfL, and their relationship to hippocampal volume. Method We analyzed 78 SCD and 200 controls from the Alfa+ cohort [mean age 61(4.6)] with Roche NeuroToolKit. We defined SCD as a positive answer to the question “Do you perceive memory or cognitive difficulties?” from the SCD‐Questionnaire. We assessed vascular risk factors and mood with the CAIDE score and the Hospital Anxiety and Depression Scale (HADS), respectively, and calculated hippocampal volumes using FreeSurfer V6.0. ANCOVA models accounting for covariates (age, gender, and HADS) were used to test the association between log10(NfL) levels and SCD status, and between log10(NfL) and bilateral total intracranial volume (TIV)‐adjusted hippocampal volume. The interactions SCD*Aβ42/Aβ40 and SCD*hippocampal*dichotomized amyloid status [Aβ+ (n=88) vs Aβ‐ (n=190), defined by the cut‐off of 0.071 for Aβ42/Aβ40] were also tested, and within group analysis performed when interactions were significant. Result SCD subjects displayed higher NfL (p=0.002, Partial Eta2=0.033) and lower Aβ42/Aβ40 (p=0.044, Partial Eta2 =0.015) than controls, after adjusting by age, gender and HADS (Table 1, Figure 1). Adding the cardiovascular risk factor score CAIDE as a covariate did not change these associations. We observed a significant SCD*Aβ42/Aβ40 interaction (p=0.028) on NfL. A sensitivity analysis stratifying for Aβ‐status showed that the association between SCD and NfL was significant only in Aβ+ individuals (p=0.006, A‐ p=0.21) (Figure 2). Further, the association between hippocampal volume and NfL was only significant in the Aβ+ individuals with SCD (p=0.04, age‐adjusted r=0.38, n=30) (Figure 3). Conclusion Aβ+ individuals with SCD displayed increased levels of CSF‐NfL. This association was independent of age, gender, anxiety/depression and vascular risk, and was related to hippocampal atrophy. The presence of SCD in Aβ+ may represent an early symptom related with neurodegeneration.