Relationship between depressive symptoms and cognitive function in the U.S. POINTER study

Abstract

AbstractBackgroundDepressive symptoms play an important role in cognitive health of older adults, yet the exact nature of their influence has not been well established. There is strong evidence that depressive symptoms are independently associated with cognitive decline and dementia in older adults. Depressive symptoms have also been associated with cognitive dysfunction in the absence of dementia.MethodsWe examined the association of depressive symptoms and cognitive functioning in U.S. POINTER participants at baseline. U.S. POINTER is a large multi‐site, 2‐year study testing the impact of interventions on cognition in 2000 older adults at risk for cognitive decline. This analysis included 1,814 study participants (aged 60‐79). Exclusion criteria included major depressive disorder and bipolar disorder. We measured cognition with the Neuropsychological Test Battery modified for U.S. POINTER, or PmNTB. The battery provides a global composite score and three domain‐specific composite scores for episodic memory, executive function, and processing speed. Depression was measured with the 15‐item Geriatric Depression Scale (GDS). Simple linear regression models were run to evaluate the relationship between cognitive function and GDS after controlling for age, sex, education, race, ethnicity, and site. In secondary analyses, simple linear regression models were used to evaluate the relationship between cognitive functioning and GDS, using a median split (0‐1, 2+) and the conventional/clinical GDS cut‐off of 4 (0‐4 vs 5+).ResultsWe found greater endorsement of depressive symptoms was not associated with the global composite score, episodic memory, or executive function. However, higher GDS scores were associated with slower processing speed (estimate = ‐0.023; 95% CI: ‐0.05, 0.00, p‐value = 0.05). In secondary analyses using the GDS median split, the association between GDS and processing speed was mildly attenuated (p‐value = 0.06) and the association between GDS and executive function became marginally associated (p‐value = 0.06). When using conventional GDS clinical cut‐offs, no associations between cognitive function and GDS were found.ConclusionAt baseline, subclinical depressive symptoms may be associated with decreased processing speed and executive function in U.S. POINTER participants.