The role of worry in subjective cognitive decline

Background Subjective cognitive decline (SCD) is defined as self-report of memory or other cognitive decline with normal test performance, but there is lack of consensus regarding operational criteria for SCD. One fundamental question is whether perception of cognitive decline is sufficient to establish SCD, or whether accompanying concern or worry about the cognitive change is another core feature. The goal of the current study was to characterize clinical and neuroimaging characteristics of older adults with subjective memory decline with worry (SCD+W) and without worry (SCD-W). Methods Participants were diagnosed based on psychiatric/medical assessment, neuropsychological (NP) test performance, and responses to 1)“Do you feel like your memory is becoming worse?”; 2)“If so, are you worried?”. The sample included 71 SCD+W [49F, age 71.1(SD6.3)] and 20 SCD-W [15F, age 72.5(SD6.9)] based on normal NP scores and responses of yes/yes and yes/no respectively, 33 MCI [17F, age 72.3(SD7.6)] defined as subjective cognitive complaints, impaired NP performance and functional independence, and 16 CU [8F, age 69.2(SD7.2)]] based on normal NP and absence of subjective memory decline/worry. Participants completed subjective memory, depression, state-trait anxiety, and health anxiety scales. Resting state functional magnetic resonance imaging (rs-fMRI) was acquired in a subset [24 SCD+W, 12 SCD-W, 12 CU] to assess default mode network (DMN) connectivity. Group effects were evaluated using multivariate ANOVA followed by Sidak-adjusted post-hoc comparisons. Results Relative to CU, SCD+W and MCI participants endorsed poorer subjective memory (P’s<.001), greater trait anxiety (P’s<.05) and greater health anxiety (P’s<.02). SCD-W and CU did not differ on any measure. On rs-fMRI, a significant group effect was found for DMN-parahippocampal connectivity [F(2, 45)=5.35, P=.008], which was attributable to lower DMN-parahippocampal connectivity in SCD+W compared to CU (P=.006). This finding remained significant with trait anxiety or health anxiety included as covariates. Conclusions Similar to MCI, SCD+W participants showed poorer subjective memory ability and greater trait and health anxiety, while SCD-W did not differ from CU. Disrupted neural connectivity was found in SCD+W only, suggesting worry may define a subgroup of SCD patients with distinct neurobiology than those without worry. Longitudinal studies are needed to determine the prognostic significance of worry.