Subjective cognitive decline and anxious/depressive symptoms during the COVID-19 pandemic: what is the role of stress perception, stress resilience, and β-amyloid?

Abstract

Background: The COVID-19 pandemic may worsen the mental health of people reporting subjective cognitive decline (SCD) and therefore their clinical prognosis. We aimed to investigate the association between the intensity of SCD and anxious/depressive symptoms during confinement and the underlying mechanisms.

Methods: Two hundred fifty cognitively unimpaired participants completed the Hospital Anxiety and Depression Scale (HADS) and SCD-Questionnaire (SCD-Q) and underwent amyloid-β positron emission tomography imaging with [¹⁸F] flutemetamol (N = 205) on average 2.4 (± 0.8) years before the COVID-19 confinement. During the confinement, participants completed the HADS, Perceived Stress Scale (PSS), Brief Resilience Scale (BRS), and an ad hoc questionnaire on worries (access to primary products, self-protection materials, economic situation) and lifestyle changes (sleep duration, sleep quality, eating habits). We investigated stress-related measurements, worries, and lifestyle changes in relation to SCD. We then conducted an analysis of covariance to investigate the association of SCD-Q with HADS scores during the confinement while controlling for pre-confinement anxiety/depression scores and demographics. Furthermore, we introduced amyloid-β positivity, PSS, and BRS in the models and performed mediation analyses to explore the mechanisms explaining the association between SCD and anxiety/depression.

Results: In the whole sample, the average SCD-Q score was 4.1 (± 4.4); 70 (28%) participants were classified as SCD, and 26 (12.7%) were amyloid-β-positive. During the confinement, participants reporting SCD showed higher PSS (p = 0.035) but not BRS scores (p = 0.65) than those that did not report SCD. No differences in worries or lifestyle changes were observed. Higher SCD-Q scores showed an association with greater anxiety/depression scores irrespective of pre-confinement anxiety/depression levels (p = 0.002). This association was not significant after introducing amyloid-β positivity and stress-related variables in the model (p = 0.069). Amyloid-β positivity and PSS were associated with greater HADS irrespective of pre-confinement anxiety/depression scores (p = 0.023; p < 0.001). The association of SCD-Q with HADS was mediated by PSS (p = 0.01).

Conclusions: Higher intensity of SCD, amyloid-β positivity, and stress perception showed independent associations with anxious/depressive symptoms during the COVID-19 confinement irrespective of pre-confinement anxiety/depression levels. The association of SCD intensity with anxiety/depression was mediated by stress perception, suggesting stress regulation as a potential intervention to reduce affective symptomatology in the SCD population in the face of stressors.

Keywords: Alzheimer’s disease; Anxiety; COVID-19 confinement; Depression; Stress; Subjective cognitive decline.

Fig. 1

Confinement-related worries and lifestyle changes in the whole sample and in the subsample with SCD. A In the whole sample, the percentage of participants showing specific worries and lifestyle changes during the confinement is displayed by SCD status. ^aWith SCD N = 62, without SCD N = 167. ^bWith SCD N = 62, without SCD N = 166. ^cWith SCD N = 62, without SCD N = 165. ^dWith SCD N = 60, without SCD N = 162. ^eWith SCD N = 60, without SCD N = 162. ^fWith SCD N = 60, without SCD N = 162. B In the subsample with SCD, the percentage of participants showing specific worries and lifestyle changes during the confinement is displayed for the participants who sought medical help and the participants who did not seek medical help. ^aWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 45. ^bWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 45. ^cWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 45. ^dWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 43. ^eWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 43. ^fWith SCD, seeking medical help N = 17; with SCD, not seeking medical help N = 43. Abbreviations: SCD, subjective cognitive decline

Fig. 2

Schematic illustration of the mediation analyses investigating the link between SCD and anxiety/depression scores. Regression coefficients for paths a, b, c, c′, and ab are reported in the figure. The total effect of SCD-Q MyCog on HADS (path c) is shown as the summation of the direct effect adjusted by the given mediator (c′) and the indirect effect (ab). The analyses were adjusted by age, sex, years of education, pre-confinement HADS scores, and time difference between the pre-confinement and confinement assessments. *p < 0.05, **p < 0.01, ***p < 0.001. A The results of the mediation analysis investigating the role of PSS as a mediator between SCD-Q MyCog scores and HADS scores. The indirect effect (path ab) was statistically significant (p = 0.01) corresponding to 63% of the total association between the two variables. B The results of the mediation analysis investigating the role of Aβ as a mediator between SCD-Q MyCog and HADS scores. The indirect effect (path ab) did not show a statistically significant effect (p = 0.45). Abbreviations: SCD-Q, Subjective Cognitive Decline Questionnaire; PSS, Perceived Stress Scale; HADS, Hospital Anxiety and Depression Scale; Aβ, amyloid-beta; CL, Centiloid