Multiple illness perceptions in older adults: Effects on physical functioning and medication adherence

Abstract

Objective: Previous research on illness perceptions has focused on single illnesses, but most adults over 65 suffer from multiple illnesses (multimorbidity). This study tests three competing operationalisations of multiple illness perceptions in predicting physical functioning and adherence: (1) main effects and interactions model; (2) peak model with effects of the most prominent illness perception; and (3) combined model with averaged illness perceptions over multiple illnesses.

Design: Longitudinal study in N = 215 individuals (65–86 years) with multimorbidity at two measurement points over six months. Participants filled in two Brief Illness Perception Questionnaires (B-IPQ) on their two most severe illnesses.

Main outcome measures: Physical functioning, medication adherence.

Results: Factor analyses suggest that the B-IPQ contains three dimensions; consequences, control and timeline. Multiple regression models fit the data best for (1) and (3). Timeline (β = −.18) and control (β = .21) predict adherence in (1); consequences (β = .16) and control (β = .20) in (3). Physical functioning was significantly predicted by interacting control beliefs in (1; β = .13), by peak consequences in (2; β = −.14) and by consequences (β = −.15) in (3).

Conclusions: Individuals with multimorbidity hold both distinct and combined perceptions about their illnesses. To understand individual responses to multimorbidity, perceptions about all illnesses and multimorbidity as entity should be examined.

Keywords:

• illness perceptions
• multimorbidity
• older adults
• physical functioning
• medication adherence
• illness representations

Acknowledgements

The content is the sole responsibility of the authors. We would also like to gratefully acknowledge the helpful comments of two anonymous reviewers and the editor, which substantially improved this article.

Notes

1. Test–retest reliabilities for the four-point response format were comparable to those reported in Broadbent et al. (2006) who examined a shorter time span. For example, in our study, the six-month test–retest reliabilities for participants indicating osteoarthritis as most severe illnesses included .40 (coherence)–.70 (emotional response). For participants indicating diabetes as most severe illness, the test–retest reliabilities included .35 (coherence)–.90 (consequences).

2. Note that the original scale ranges from ‘never’ to ‘very often’; we have changed the response format based on the think-aloud pilot study, in which the ‘never’ response increased item difficulty considerably (i.e. participants found this response option too restrictive).