Depression and Cognitive Outcomes: Linking the Mayo Clinic Study of Aging and Rochester Epidemiology Project.

Depression was associated with a significantly increased risk for progression to mild cognitive impairment.

• Hazard ratio for MCI was 1.37 (95% CI=1.13 to 1.65)
• Models used Cox proportional hazards models adjusted for pertinent covariates
• Sample included 1,805 community-dwelling adults with mean age 74.5 ± 7.3 years
• Participants were cognitively unimpaired at baseline

Depression was associated with a significantly increased risk for progression to dementia.

• Hazard ratio for dementia was 1.33 (95% CI=1.04 to 1.70)
• Models used Cox proportional hazards regression adjusted for pertinent covariates
• Depression was classified via EMR diagnosis or antidepressant medication use

In sensitivity analyses, only participants taking antidepressants (not those with depression diagnosis alone) were at increased risk for progressing to MCI and dementia.

• Participants were stratified into those with antidepressant use versus depression diagnosis without antidepressants
• Those without depression served as the reference group
• This finding suggests the relationship between depression and cognitive outcomes may differ by depression subgroup or treatment status

Depression was associated with lower baseline cognitive z scores in all cognitive domains except memory.

• Linear mixed-effects models were used to examine global and domain-specific cognitive z scores
• The association with lower cognitive scores at baseline was observed across multiple domains
• Memory was the only domain that did not show a significant association with depression at baseline

Associations between depression and longitudinal cognitive trajectories were only observed when the timing of depression was taken into account.

• Depression timing was categorized as: no depression, depression before MCSA baseline only, depression at MCSA baseline only, or depression at both time points
• Without considering timing, depression was not significantly associated with longitudinal cognitive trajectories
• This suggests the temporal relationship between depression onset and cognitive assessment is important for detecting longitudinal effects

The study population consisted of 1,805 community-dwelling older adults recruited through the Mayo Clinic Study of Aging with linked Rochester Epidemiology Project data.

• 51% of participants were men
• Mean age was 74.5 ± 7.3 years
• All participants were cognitively unimpaired at baseline
• Depression data were acquired from electronic medical records via the REP, defined by a depression diagnosis or antidepressant medication use