UC San Diego

Individuals diagnosed with Mild Cognitive Impairment (MCI) are at a higher risk for conversion to dementia and, therefore, may be particularly motivated/well-suited to participate in interventions to slow cognitive decline. Cognitive and lifestyle interventions targeting modifiable risk factors (e.g., physical activity [PA] and sleep) can slow the rate of cognitive decline. Psychometrically validated subjective indices of PA and sleep offer a practical and economical method of assessing these cognitively salient lifestyle factors in the context of longitudinal studies of older adults with MCI. Moreover, the burgeoning availability of inexpensive, accurate wearable devices to measure lifestyle behaviors makes valid, unobtrusive objective measurement of these outcomes possible. Thus, this dissertation study examined the efficacy of an 8-week Motivationally Enhanced Compensatory Cognitive Training (ME-CCT) intervention, compared to Goal-focused Supportive Contact (SC), in improving subjectively and objectively measured lifestyle factors in older Veterans with MCI.Self-reported sleep disturbance and PA levels were examined at baseline, mid-treatment, and post-treatment in 74 Veterans with MCI enrolled as part of a larger randomized controlled trial. Sleep and PA were objectively measured via the Fitbit Charge 2 in a subset of the sample (n=23). Mixed-effects models examined whether (1) ME-CCT, compared to SC, was associated with greater improvements in self-rated PA levels and sleep disturbance; and (2) baseline levels of self-rated PA and sleep disturbance moderated treatment effects on these outcomes. Exploratory analyses examined baseline-level correlates of Fitbit-measured PA and sleep, and the efficacy of ME-CCT in improving objective PA/sleep. There was no differential treatment-related improvement in either subjective or objective measures of PA or sleep, with no treatment-moderating effect of baseline PA/sleep (ps>.05). At baseline, greater self-reported engagement in PA (rs=0.57, p=.005) and less executive dysfunction (r=-0.43, p=.043) were associated with greater Fitbit step counts. Moreover, higher pain intensity (r=-0.49, p=.048) was associated with shorter Fitbit sleep duration. ME-CCT did not differentially improve sleep or PA levels, nor did baseline levels of sleep/PA influence treatment outcomes. The influence of pain intensity and executive deficits, such as planning and organizing, in moderating trajectories in cognitive and health behavior engagement should be considered in current rehabilitation efforts in aging.