Chapter 2
Psychiatric comorbidities in schizophrenia: the size of the problem
This is a book about psychiatric comorbidities in schizophrenia. There are significant conceptual issues relating to the essence of what one considers ‘comorbid’ and how one classifies, assesses, and enumerates putative comorbid conditions. These puzzles are addressed in other chapters of this book, notably Chapters 1 and 3. In this chapter we sidestep most of those considerations and accept the conventional use of the term ‘comorbid’, namely another psychiatric condition occurring in a person with schizophrenia and not directly explicable as part of the manifestations of the schizophrenia itself. The established diagnostic criteria for the comorbid diagnosis need to be met, with the accepted exclusions of symptoms due to an organic process (although this is complex when it comes to substance use disorders, as detailed in Chapter 11). This allows a review of the pertinent literature regarding how often such conditions occur in people with schizophrenia as opposed to people without schizophrenia. We follow the convention that schizophrenia hierarchically ‘trumps’ other psychiatric disorders (see Chapter 3).
General considerations
Overall rates of psychiatric comorbidities in people with schizophrenia are dependent upon a number of methodological issues, summarized in Box 2.1. The sampling frame (i.e. general population vs. a treated population) is critical, as treated samples introduce biases related to severity of illness as well as Berkson’s bias, namely the bias related to ascertainment of two or more disease entities through pathways relevant to each, thus overenumerating their co-aggregation.
Box 2.1 Methodological issues relevant to ascertaining rates of comorbidity in schizophrenia
• Setting (e.g. general population vs. a treatment setting)
• Skill and experience of interviewers
• Reporting bias (e.g. illicit drugs)
• Pathways and barriers to care for each condition
Another critical issue is the diagnostic criteria applied: this is relevant to both schizophrenia and the comorbid condition. Schizophrenia itself has undergone many alterations in its definition over the last century. A full exposition of these changes is beyond the scope of this book and the reader is referred to the companion book in this series for details (Castle and Buckley 2015). A summary is provided in Box 2.2. Perhaps of most importance to this chapter are the variations in age at onset, as psychiatric comorbidities affect individuals at different life phases, as outlined below. Feighner criteria for schizophrenia loaded towards early onset cases (under 40 years), DSM-III specified an onset before the age of 45 years, but DSM-IV dropped any age specification. Another related issue is the impact on sex ratios of the different sets of criteria. For example, Feighner’s criteria preference males, with an estimated male:female ratio of 2.5:1, whilst DSM-III returns a ratio of 2.2:1, and more ‘liberal’ criteria with age at onset specifications, such as ICD-9, estimate males and females to be roughly even in terms of schizophrenia risk (Castle et al. 1993). Sex affects psychiatric comorbidities in important ways, again as detailed below. Finally, quirky criteria within certain diagnostic sets will serve to prejudice against finding high rates of psychiatric comorbidities, for example the loading for a family history of schizophrenia in Feighner’s criteria.
Box 2.2 Selected overview of changes in the schizophrenia construct, over time, relevant to psychiatric comorbidities
• Kraepelin (1896): early onset, male preponderant ‘neurodevelopmental’-type illness
• Research Domain Criteria (RDC) (1976): broader criteria
• ICD-9 (1970s): broad definition not operationalized
• DSM-III (1980): age at onset stipulated as under 45 years
• ICD-10 (1990): no age at onset stipulation
• DSM-IV (1987): abandoned age at onset stipulation
• DSM-5 (2013): moved away from emphasis on Schniederian ‘first rank’ symptoms; abandoned subtypes