Bayesian Analyses Showed More Evidence for Apathy than for Depression Being Associated With Cognitive Functioning in Nursing Home Residents.

Abstract Objectives Depression, apathy, and cognitive impairments are widespread in nursing home (NH) residents. Scarce research that explicitly compares apathy to depression suggests that the association between apathy and cognitive functioning is stronger than the association between depression and cognitive functioning. This study in Dutch NH residents aimed to use Bayesian methods for comparing the evidence for the relationship of cognitive performance with apathy to that with depression. Design Cross-sectional. Setting and Participants Sixteen NH somatic care units (N = 190 residents; mean age 77.2 years, standard deviation 12.9), and 17 dementia special care units (N = 243 residents; mean age 82.8 years standard deviation, 6.8]). Measures The Frontal Assessment Battery (FAB) and Mini-Mental State Examination (MMSE) were administered in residents for cognitive performance. Professional carers were interviewed for the Apathy Evaluation Scale and the Cornell Scale for Depression in Dementia. Results Regression models built with the BayesFactor package in R showed Bayesian factors (BFs) that implied extremely strong evidence in terms of the Jeffrey classification for the relationship of apathy with MMSE [standardized effect size, −0.57 (−0.66 to −0.48), BF = 3.4E+28], and with FAB [−0.50 (−0.59 to −0.42), BF = 3.0E+24]. Regarding depression, evidence was a minor fraction of that for apathy [MMSE, −0.17 (−0.27 to −0.06), BF = 15.45; FAB, −0.12 (−0.22 to −0.02), BF = 2.11]. The most evidence existed for the associations of apathy with MMSE orientation problems, especially orientation in time. Conclusions/Implications The study implies that cognitive assessments are important to differentiate apathy from depression in NH residents both with dementia and without dementia. More research is needed to clarify whether disorientation in time is a specific cognitive marker of apathy that may be used to reduce false positive depression diagnoses. [ABSTRACT FROM AUTHOR]