Association between symptoms and quality of life in patients with schizophrenia: A pooled analysis of changes over time

https://doi.org/10.1016/j.schres.2011.09.021Get rights and content

Abstract

Quality of life is an important outcome in the treatment of patients with schizophrenia. It has been suggested that patients' quality of life ratings (referred to as subjective quality of life, SQOL) might be too heavily influenced by symptomatology to be a valid independent outcome criterion. There has been only limited evidence on the association of symptom change and changes in SQOL over time. This study aimed to examine the association between changes in symptoms and in SQOL among patients with schizophrenia. A pooled data set was obtained from eight longitudinal studies that had used the Brief Psychiatric Rating Scale (BPRS) for measuring psychiatric symptoms and either the Lancashire Quality of Life Profile or the Manchester Short Assessment of Quality of Life for assessing SQOL. The sample comprised 886 patients with schizophrenia. After controlling for heterogeneity of findings across studies using linear mixed models, a reduction in psychiatric symptoms was associated with improvements in SQOL scores. In univariate analyses, changes in all BPRS subscales were associated with changes in SQOL scores. In a multivariate model, only associations between changes in the BPRS depression/anxiety and hostility subscales and changes in SQOL remained significant, with 5% and 0.5% of the variance in SQOL changes being attributable to changes in depression/anxiety and hostility respectively. All BPRS subscales together explained 8.5% of variance. The findings indicate that SQOL changes are influenced by symptom change, in particular in depression/anxiety. The level of influence is limited and may not compromise using SQOL as an independent outcome measure.

Introduction

Quality of life is regarded as a relevant outcome criterion in the treatment of patients with schizophrenia (Priebe and Fakhoury, 2008). There are objective and subjective indicators of quality of life, and patients' rating of their own quality of life (referred to as subjective quality of life, SQOL) are widely used in research. The US Food and Drug Administration (US FDA, 2009) explicitly referred to it when underlining the importance of using patient reported outcomes in trials. Whilst there is no universal consensus on its precise definition, several scales are based on Lehman's (Lehman, 1996) approach of measuring SQOL as satisfaction with life in general and major life domains.

Many research studies and evaluations of routine care have found symptomatology to be associated with SQOL scores (Priebe and Fakhoury, 2008). Numerous cross-sectional studies, including meta- and pooled analyses (Eack and Newhill, 2007, Vatne and Bjorkly, 2008, Priebe et al., 2011), have reported that higher symptom levels are associated with less favourable SQOL ratings. These findings have led to suggestions that SQOL ratings in patients with schizophrenia might be too heavily influenced by symptoms and therefore not a valid independent outcome criterion (Atkinson et al., 1997, Katschnig, 1997).

However, treatment effects are commonly evaluated by investigating changes over time. For assessing whether SQOL can be used as an outcome criterion independent of symptoms, evidence is required on how changes in symptoms and changes in SQOL over time are associated. Cross-sectional associations can either overestimate or underestimate the extent of associations of changes over time. Only a few studies have investigated the association of symptom change and change in SQOL in a longitudinal design (Kaiser and Priebe, 1998, Fakhoury et al., 2002). These studies had small sample sizes and produced inconsistent findings. While Kaiser and Priebe (1998) reported an association between change in anxiety and depressive symptoms, symptoms of hostility, and change in SQOL, there was no evidence of such an association for general psychopathology and other types of symptoms. Further, Fakhoury et al. (2002) even found an absence of evidence of an association of changes in SQOL with changes in general psychopathology and depression/anxiety symptoms.

Using a pooled analysis of individual patient data from several studies, we aimed to study the association of changes in psychopathological symptoms and changes in SQOL in patients with schizophrenia. A pooled analysis considers both studies and individual patients as the unit of analysis and has several advantages compared to a conventional meta-analysis of aggregate data sets: it enables a more precise estimate of the effects of influential factors; allows for controlling of confounding factors including heterogeneity across studies at the patient level; and reduces effects of heterogeneity from aggregation of methodologically diverse studies by using the same statistical model (Blettner et al., 1999, Reininghaus and Priebe, 2007).

Section snippets

Methods

In a pooled analysis, linear mixed models were applied to individual patient data from samples of patients with schizophrenia, with SQOL change scores as the dependent variable.

Included studies and samples

Eight studies (Roder-Wanner and Priebe, 1998, Priebe et al., 2002, Priebe et al., 2007, Priebe et al., 2009, Ruggeri et al., 2002, Slade et al., 2006, Kallert et al., 2007), including one unpublished study (Junghan, 2009, unpublished data), were included (see Table 1). Five were prospective observational and three randomised controlled trials. From these 8 studies,BPRS and SQOL change scores were available for a total of 886 patients with schizophrenia, schizotypal, or delusional disorders. The

Discussion

After controlling for heterogeneity of findings across studies in linear mixed models, a reduction in symptoms was associated with improvements of SQOL scores. In univariate analyses, changes in each BPRS subscale were associated with changes in SQOL scores. Only associations of depression/anxiety and hostility subscales and SQOL change scores remained significant in a multivariate model. Whilst these associations were statistically significant in a large sample, the shared variance of symptom

Role of funding source

This work was supported by a Research Training Fellowship funded by the National Institute for Health Research (UK) to U.R. The report is an independent research and the views expressed in this publication are those of the authors and not necessarily those of the NHS, the National Institute for Health Research or the Department of Health.

Contributors

SP designed the study and had overall responsibility for completion. UR conducted the literature searches and managed the data. SP, RM, UJ, TK, MR, and MS took each responsibility for at least one of the data sets in the analysis. UR and SP analysed the data. All authors contributed to and have approved the manuscript.

Conflict of interest

None of the authors reported potential conflicts of interests.

Acknowledgements

None.

References (31)

  • H. Katschnig

    How useful is the concept of quality of life in psychiatry?

    Curr. Opin. Psychiatry

    (1997)
  • S. Kay et al.

    The positive and negative syndrome scale (PANSS) for schizophrenia

    Schizophr. Bull.

    (1987)
  • A. Lehman

    Measures of quality of life among persons with severe and persistent mental disorders

    Soc. Psychiatry Psychiatr. Epidemiol.

    (1996)
  • J. Oliver et al.

    Measuring the quality of life of severely mentally ill people using the Lancashire Quality of Life Profile

    Soc. Psychiatry Psychiatr. Epidemiol.

    (1997)
  • J. Overall et al.

    The Brief Psychiatric Rating Scale (BPRS): recent development in ascertainment and scaling

    Psychopharmacol. Bull.

    (1988)
  • Cited by (69)

    • Quality of life after involuntary psychiatric admission

      2022, International Journal of Law and Psychiatry
      Citation Excerpt :

      To our knowledge, only three published studies have examined QoL after involuntary admission (Priebe et al., 2011; Rüsch et al., 2014; Shannon et al., 2015). Better QoL several months after admission was associated with baseline satisfaction with treatment (Priebe et al., 2011), lower scores on the Brief Psychiatric Rating Scale (BPRS) and absence of psychosis (Rüsch et al., 2014). Shannon et al. (2015) found that coercive events were not associated with QoL scores at one-year post-discharge in a population including both voluntarily and involuntarily admitted patients.

    View all citing articles on Scopus
    View full text