Direct and Indirect Predictors of Medication Adherence With Bipolar Disorder: Path Analysis

Abstract

Background Despite the efficacy of treatment and severity of symptoms, medication adherence by many with bipolar disorder (BD) is variable at best. This poses a significant challenge for BD care management. Objective For this study, we set out to identify psychosocial and psychiatric predictors of medication adherence with BD. Methods Using microtargeted social media advertising, we recruited an international sample of young and older adults with BD living in North America (Canada and the United States), Western Europe (eg, United Kingdom and Ireland), Australia and New Zealand (N=92). On average, participants were 55.35 (SD 9.65; range 22-73) years of age, had been diagnosed with BD 14.25 (SD 11.14; range 1-46) years ago, and were currently prescribed 2.40 (SD 1.28; range 0-6) psychotropic medications. Participants completed questionnaires online including the Morisky Medication Adherence Scale. Results Medication adherence did not significantly differ across BD subtypes, country of residence, or prescription of lithium versus other mood stabilizers (eg, anticonvulsants). Path analyses indicate that alcohol misuse and subjective or perceived cognitive failures are direct predictors of medication adherence. BD symptoms, psychological well-being, and the number of comorbid psychiatric conditions emerged as indirect predictors of medication adherence via perceived cognitive failures. Conclusions Alcohol misuse did not predict perceived cognitive failures. Nor did age predict medication adherence or cognitive failures. This is noteworthy given the 51-year age range of participants. That is, persons in their 20s with BD reported similar levels of medication adherence and perceived cognitive failures as those in their 60s. This suggests that perceived cognitive loss is a facet of adult life with BD, in contrast to the assumption that accelerated cognitive aging with BD begins in midlife.