Clinical and Cognitive Predictors of Response to Psychotherapy in Late Life Depression

Introduction Late life depression (LLD) is a common and heterogeneous condition that is associated with increased disability, cognitive impairment, and all-cause mortality in older adults. Anti-depressant medication and psychotherapy have been identified as effective treatments for LLD, though as many as 50% of patients do not respond to treatment. Several clinical (age, age of MDD onset, MDE severity, anxiety, perceived stress, social support, sleep quality) and cognitive (executive dysfunction, global cognitive impairment) factors have been identified as predictors of response to anti-depressant treatment. Less is known about predictors of psychotherapy treatment response in LLD, but cognitive dysfunction has been linked to poor treatment outcomes and may negatively impact the effectiveness of behavioral interventions that are subserved by various cognitive processes. Methods Participants included 83 individuals with LLD (HAM-D > 17) enrolled in a manualized 12-week psychotherapy with a licensed psychologist. All participants completed diagnostic interviews, comprehensive neuropsychological assessments and clinical evaluations before and after treatment. Separate multiple linear regression models were used to assess independent effects of baseline MDD features (age at onset, severity, treatment history, frequency and chronicity of MDEs), related clinical factors (comorbid anxiety and hoarding behaviors, perceived stress and health, social support, sleep disturbance), and cognitive performance (global cognition, executive functioning, episodic memory, processing speed, language, visuospatial skills, motor speed and dexterity) as predictors of treatment response, with percent change in HAM-D as the outcome variable. Cognitive performance was grouped by domain using composite Z-scores based on group means and standard deviations (all Chronbach's a's > .70). Results Forty-five percent (n=38) of participants showed treatment response (>50% reduction in HAM-D post-treatment). Responders were significantly younger (mean age = 69.76) than non-responders (mean age = 73.11, t = 7.10, p = .009). Level of education, gender distribution, and baseline depression characteristics (severity, length of episode, anti-depressant use) were equivalent (all p’s > .05). depression features accounted for 23.9% of the variance in HAM-D change, and showed a trend toward significance for older age of MDD onset associated with increased treatment response (t = 1.85, b = 0.38, p = .06) after accounting for age (t = 2.10, b = 0.31, p = .017). Non-depression clinical features accounted for 31.2% of the variance in HAM-D change, and greater sleep disturbance was associated with worse treatment response after accounting for age (t = 2.53, b = 0.34, p = .015). Cognitive performance accounted for 10.1% of the variance in HAM-D change, with no significant predictors after accounting for age (t = 2.36, b = 0.25, p = .021 Conclusions Our findings show that older age and sleep disturbance predict poorer response to treatment with 12 weeks of psychotherapy in a sample of individuals with LLD. Older age may negatively impact treatment response via increased risk for underlying neurodegenerative pathologies that are not affected by treatment with psychotherapy or medication. Similarly, sleep dysfunction in older adults has been implicated in the pathogenesis of Alzheimer's disease and related dementias and may reflect the presence of an underlying neurodegenerative process that is not targeted by psychiatric intervention. A preliminary exploration of cognitive factors suggests that cognitive dysfunction may play a more limited role in response to psychotherapeutic relative to pharmacologic intervention. Future research should assess the effect of neurodegenerative pathology on response to treatment with psychotherapy and medication in LLD. Funding This project was supported by the National Institute of Mental Health grant 1R01MH101472.