Cognitive Performance in People with Type 2 Diabetes and Comorbid Depression

Abstract

Cognitive Performance in People with Type 2 Diabetes and Comorbid Depression¶

Background: Worse cognitive performance is often under-diagnosed among adults with type 2 diabetes (T2D). Vulnerability factors that can exacerbate cognitive dysfunction among at-risk individuals with T2D are poor glycemic control, increased age, depression severity, and anticholinergic and/or sedative drug burden. This can be conceptualized as a syndemic, or co-occurring risk-factors that can be related to each other and combine to lead to compounded negative effects on patient outcomes. However, few studies assess more than one of these vulnerabilities in relation to cognitive function in T2D. Focusing on the contribution of these factors to increased risk for worse cognitive performance will improve our understanding of their additive effects and will help inform interventions to improve cognitive outcomes in these patients. Given that diabetes treatment and self-management demands are cognitively demanding, the multiple vulnerabilities that represent this syndemic for worse cognitive performance in T2D may negatively impact self-management and the success of behavioral therapeutic interventions aiming to improve depression and diabetes self-management. The proposed study examined associations between these cognitive vulnerabilities and cognitive performance among adults with T2D who were recruited for participation in a cognitive behavioral therapy (CBT) trial for depression and diabetes self-management. The potential moderating effects of baseline cognitive performance on the effects of this intervention was also explored. ¶

Methods: The current study was a secondary analysis of a randomized control trial examining the efficacy of a cognitive behavioral therapy for adherence and depression (CBTAD) in patients with sub-optimally controlled T2D and comorbid depression. Cognitive performance was assessed by a neuropsychological testing battery and calculated as an overall aggregate score across these tests, which was supported by an Exploratory Factor Analysis. Cognitive vulnerability factors were measured continuously and dichotomously using clinically meaningful cut offs. Linear regressions were used to assess predictors of cognitive performance. Additive risk was assessed by quantifying syndemic vulnerability factors as a count score (quantified 0-4). A generalized linear model was used to explore potential moderating effects of cognitive performance on CBT-AD treatment outcomes (i.e., clinician-rated depression severity and electronically monitored medication adherence). It was expected that those with worse overall cognitive performance would show less benefit from the intervention on these outcomes as compared to those with better cognitive performance.¶

Results: Higher number of complications was significantly associated with lower cognitive performance (β=-.33, p=.002), remained significant after adjustment for the syndemic vulnerability factors (β=-.22, p=.048), but was attenuated to non-significance when controlling for total number of medications (β=-.18, p=.115). Age was the only statistically significant vulnerability factor associated with lower cognitive performance (β=-.35, p=.001) and remained significant when controlling for total medications (β=-.35, p=.001). Age remained a significant predictor when dichotomized to reflect older age (≥65 years; β= -.29, p=.005). Accumulative risk, indicated by higher syndemic vulnerability count score, was significantly associated with lower cognitive performance (β=-.29, p=.006) and remained significant after controlling for total number of medications (β=-.25, p=.022). Moderation analyses showed there was no significant interaction between cognition and treatment condition over time for depression severity (F(1,139)=.042; p=.839) or for electronically monitored medication adherence (F(1,113)=.270; p=.604).¶

Discussion: Out of the four proposed vulnerability factors, older age was the only statistically significant predictor of worse cognitive performance in individuals with T2D and comorbid depression, exhibiting a large effect size (f2 =.34), which is consistent with prior literature. The overall model including the syndemic sum score resulted in statistically significant increase in predicted variance, explaining 7.8% of variance in overall cognitive functioning when controlling for number of complications (i.e., retinopathy, nephropathy, neuropathy, stroke, angina, and heart attack). This corresponds to a small effect size ( f2 =.24) for additive syndemic risk. However, this effect appears to have been driven primarily by age. This study adds to the syndemic literature on diabetes and represents the first study to our knowledge to examine accumulative risk of worse cognitive performance in a vulnerable subset of the T2D population. Additionally, despite our insignificant findings regarding the moderating effects of cognitive performance on treatment outcomes, our study was the first of its kind to examine cognitive performance as a potential moderator of CBT outcomes among depressed adults with T2D and will pave the way for future research.