Exploring longitudinal relationships between psychological flexibility and medication adherence, mood and general functioning in people with long-term health conditions.

Introduction: Many people with long-term physical health conditions (LTCs) are non-adherent to prescribed medications and therefore have an increased risk of morbidity and mortality. Several psychological models have attempted to understand why people with LTCs do not adhere, but all have limited explanatory power and interventions stemming from them show modest effects. Few studies have explored the utility of the psychological flexibility model (PF), the transdiagnostic theory underlying Acceptance and Commitment Therapy (ACT), in this context. A small number of preliminary trials for ACT show promising efficacy but have been conducted in specific LTC groups with small samples. Ecological momentary assessment methods (EMA) may build on these studies because they could examine temporal relationships and account for within-individual variability across different contexts and are less prone to recall biases. However, few momentary PF measures have been validated in LTCs samples.
Method: The primary aim of this online longitudinal study (n=701) was to examine relationships between validated measures of PF and self-reported intentional and unintentional non-adherence and appointment attendance in people with LTCs at baseline and three months follow-up using binomial regressions. The second aim (not reported in the current thesis) was to preliminarily validate new momentary measures of PF, adherence and mood for future EMA studies to better understand within-individual and group-level variability.
Results: PF variables explained a significant, albeit modest, amount of the variance in intentional and unintentional non-adherence and appointment attendance. However, confirmatory factor, internal consistency and test-re-test analyses indicated MARS-5 items failed to meet established criteria for construct validity and demonstrated poor stability over time. This was supported by the instrument’s poor convergence with new appointment attendance scales. PF shared medium to strong relationships with mood and general functioning in expected directions.
Discussion: This project has improved our understanding of the potential applicability of the PF model and ACT in understanding and treating intentional and unintentional non-adherence and appointment non-attendance. However, further clarification of the utility of PF in understanding treatment non-adherence is warranted using prospective or experimental designs in conjunction with more objective valid and reliable adherence measures.