The prognostic role of perceived criticism, medication adherence and family knowledge in bipolar disorders

Abstract

Background

In schizophrenia, high levels of critical comments by significant others are associated with early relapse, especially if medication adherence is sub-optimal. Levels of criticism may be influenced by family knowledge about both the disorder and its treatment. No study has explored whether this combination factors influence outcome in adults with bipolar disorders.

Methods

Medication adherence was assessed in 81 individuals with bipolar disorder of whom 75 rated perceived criticism by an identified ‘significant other’ as well as their own perceived sensitivity. 33 (of the 75) had a close family member who agreed to completed an assessment of their knowledge and understanding of bipolar disorders. Psychiatric admissions were then recorded prospectively over 12 months.

Results

Perceived criticism and medication adherence were significant predictors of admission. In the patient–family member dyads (n=33), the odds ratio (OR) for admission was 3.3 (95% confidence intervals 1.3–8.6) in individuals with low levels of medication adherence, high perceived criticism, and a family member with poor knowledge and understanding.

Limitations

The small sub-sample of patient–family member dyads means those findings require replication. Sensitivity to criticism by professional caregivers may not equate to that by relatives.

Conclusions

Perceived criticism may be a simple but robust clinical predictor of relapse in mood disorders. High levels of perceived criticism, poor understanding of bipolar disorder by a significant other, and sub-optimal treatment adherence are risk factors for hospitalization in adults with bipolar disorders that are potentially modifiable through the use of strategic psychosocial interventions.

Introduction

Even with optimal pharmacological treatment, about 50% of individuals with bipolar disorders (BD) experience a relapse in the year after an index episode (e.g. Gitlin et al., 1995, Scott et al., 2006). This suggests that individual psychosocial and environmental variables may also be important risk factors for relapse (Reinares et al., 2006, Scott, 1995, Scott and Colom, 2005, Vieta and Colom, 2004). In schizophrenia, high levels of ‘critical comments’ within a family are associated with early relapse, especially if the patient is poorly adherent with medication (Vaughan and Leff, 1976). Furthermore, family ‘emotional environment’ may be influenced by their knowledge about schizophrenia and its treatment (Sellwood et al., 2003).

Research has highlighted how individual beliefs and expectations about BD and its treatment can directly influence the likelihood of medication adherence, which will modify relapse risk (Scott and Pope, 2002a, Clatworthy et al., 2009). The beliefs and attributions of a patient's family may also impact on the patient's adherence with medication and relapse rates (Tacchi and Scott, 2005, Velligan et al., 2009). Perlick et al., 2001, Perlick et al., 2004 showed that the family's beliefs and coping level significantly predicted family burden and that burden predicted the outcome of BD at follow-up. Furthermore, this effect was mediated by the affective response in the family and patient levels of medication adherence. However, research on emotional environment in BD and its relationship to BD medication adherence is less well developed than in schizophrenia or unipolar disorders (Butzlaff and Hooley, 1998).

In unipolar disorders, expressed emotion and a related construct, perceived criticism, are robust predictors of depressive relapse (Hayhurst et al., 1997, Hooley and Licht, 1997, Hooley and Teasdale, 1989). In BD, Yan et al. (2004) demonstrated that negative emotional environment was especially predictive of depressive relapses in 47 individuals with BD I. Other studies have demonstrated an association between negative affective style and a poor response to lithium prophylaxis or admission (O'Connell et al., 1991, Preib et al., 1989, Honig et al., 1997). Simoneau et al. (1998) showed that BD families with more ‘toxic’ emotional environments had more negative interpersonal interactions, whilst Tompson et al. (2000) noted that high expressed emotion families were less likely to engage in family therapy. Miklowitz et al. (2009) highlight that familial expressed emotion moderates the effects of family therapy for adolescents with BD. However, only one small study (n=17) has specifically explored the interaction between affective style, emotional environment, medication adherence and relapse rates in a homogenous clinical sample of BD cases (Miklowitz et al., 1988). The study showed that negative affective style and high expressed emotion predicted poor outcome, especially in non-adherent cases.

We have previously explored the relationship between likelihood of medication adherence and patient attitudes to both BD and BD treatment (Colom et al., 2005a, Colom et al., 2005b, Scott and Pope, 2002b, Tacchi and Scott, 2005, Clatworthy et al., 2009). We believe that it is now important to clarify in a clinical BD cohort (a) the correlations among medication adherence, family knowledge about BD and emotional environment (as assessed by patient ratings of perceived criticism) and (b) the association of each of these variables with the adverse outcome of psychiatric admission during the next 12 months.