Clinical Manifestations.

Abstract

Oldest old adults (90+) are the fastest growing population with high burden of cognitive impairment and depression. Though depression is a known risk factor for cognitive impairment, it is unclear how each of the predominant depression dimensions correspond with cognitive outcomes. We used baseline data from LifeAfter90 study, a racially and ethnically diverse cohort of community-dwelling adults aged 90 and above. Participants are longtime members of Kaiser Permanente Northern California Health System without a dementia diagnosis in their medical record. Verbal episodic memory, semantic memory, and executive function were assessed with the Spanish & English Neuropsychological Assessment Scale and averaged as a composite measure of global cognition. Depression was measured by the Geriatric Depression Scale (GDS) and the dimensions (Dysphoria, Withdrawal-Apathy -Vigor (WAV), Anxiety, Hopelessness and Subjective Memory Complaint) calculated based on GDS-items. Generalized linear mixed models estimated associations between global and domain-specific cognition and depression dimensions adjusting for age, sex, education, and race. Participants (n = 960, mean age = 92.4 ± 2.3 yrs, 61.56% female) were 24% Asian, 23% Black, 19% Latino, and 27% White. The most common reported GDS items were "decreased energy" (52.45%), preference to "stay home" (46.31%) and "dropped activity" (40.0%) and 25% reported memory complaints (Figure 1). Hopelessness was associated with worse executive functions (P< 0.001) but not with other cognitive domains nor with global cognition. Dysphoric mood, WAV, and anxiety were not associated with any of the cognitive outcomes (Table 1). Conversely, subjective memory complaints was associated with worse global cognition, executive function and verbal episodic memory, but not with semantic memory. Characterizing the dominant depression dimensions may be informative in predicting pattern of cognitive deficits. Perceived memory impairment in the oldest old was supported by objective deficit in episodic memory and executive functions suggesting that subjective reports should be taken as a valid early sign of impairment in this population. Further exploration of the longitudinal association of depression dimensions with cognitive domains is warranted.