A powerful argument for discontinuation of the use of subtypes in schizophrenia is that administrative psychiatric practice data collected in the US and Europe show that most are rarely used diagnostically (<5%), with the exception of paranoid schizophrenia (50-75%) and, to a lesser extent, undifferentiated schizophrenia. It could be argued, however, that subtypes may show genuine epidemiological variation and therefore should be retained. The larger question, therefore, is whether there is evidence that subtypes are valid. This question was subject of a systematic review described in detail elsewhere, focusing on subtypes within schizophrenia and within psychotic disorders as a whole¹. 

Researchers who have endeavored to examine the question of whether taxonic subtypes exist in schizophrenia and broader groupings of psychotic disorders, have used one or more of four statistical approaches. We identified 24 relevant papers describing 38 analyses of 28 participant cohorts that met a priori inclusion criteria. Face-value reading of these studies suggests that a variety of latent-class models may underlie a rationale for subtyping in schizophrenia: 

Two-class interpretations

                Fourteen analyses concluded with two-class interpretations, for example between neurodevelopmental and complement groups, in those with schizophrenia

Three-class interpretations

                In five cases, findings have been interpreted as evidence of three latent classes, for example, twelve-month follow-up ratings of negative symptoms in two cohorts yielded remitted, partially remitted, and worsened outcome-related classes, or neurodevelopmental, paranoid, and schizoaffective classes

Four-class interpretations

                For example, studies suggested a four-class solution of psychotic, mixed positive-negative, schizomania, and schizodepression subtypes.

Five-class interpretations

                Two studies have identified five-class solutions, for example schizoaffective, negative symptom schizophrenia, chronic delusional, paranoid, and remitting/relapsing catatonic schizophrenia classes.

Six-class interpretations

                Two studies concluded with six-class interpretations, for example classic schizophrenia, major depression, schizophreniform, bipolar schizomania, schizodepression, and hebephrenia classes. 

Limitations undermining evidence quality

The results of the systematic review do not point to a single system of subtyping. More important, however, is that there are a variety of important limitations affecting the internal and external validity of these findings.

                By far the most substantial problem affecting this body of work is the application of statistical methods that are biased in favor of finding latent classes.

                Second, the key result from modeling methods is not the absolute quality or fit of the final model. Instead, the key result is how the final model (e.g., of K classes) compares to the null model (K = 1 class) and to potential competing models in light of substantive considerations. In many of the studies, one or more of the null or adjacent models (i.e., K – 1 classes, K + 1 classes) was not explicitly examined.

                Third, in many instances, the analysis approach required the assumption of within-class independence among the class indicators. Failure of this assumption is a significant threat to the validity of findings from LCA and latent regression analysis, whereas CCK methods can accommodate small to moderate conditional dependence. In a large number of studies, adherence to the assumption of conditional independence went unchecked. Yet in some studies, it seems highly likely that this assumption was violated.

                Fourth, problems with conditional dependence are sometimes addressed with preliminary analyses during the indicator selection stage, along with other aspects of indicator quality, such as item validity and multidimensionality of measures. However, indicator quality received attention, albeit sometimes very little, in only 11 of the 38 analyses.

                Fifth, CCK and other modeling methods commonly used are not intelligent systems. Thus, recruitment and sampling procedures that contribute systematic error to observations will, at a minimum, obscure the true latent structure, and when more extreme, may create artefactual latent class structures. One critical source of systematic error variance prominent among the reviewed studies was the use of commingled samples.

                Sixth, a second critical source of systematic bias was the imposition of arbitrary constraints on the sample population. This occurred frequently. For example, including only or excluding those who meet DSM-III-R criteria for schizophrenia or a related disorder places an arbitrary suppositional constraint on latent structure outcomes. The consequences of such constraints are unclear, but may include under- or over-extraction of classes, depending on where the criteria fall with respect to the true latent structure, or misspecification of the prevalence rates of latent classes.

                Seventh, just as sample recruitment and screening methods can introduce artifacts, so too assessment methods and indicator construction. First, the modal method of assessment of indicators involved interviewer ratings. In many instances, these were made in the course of clinical assessment or in the context of diagnostic interviews where it may be expected that the raters’ conceptualizations of psychopathology, including diagnostic boundaries, or the imperatives of the assessment device introduced perceptual or rating biases. Although useful in some contexts, such biases have been demonstrated to modify the latent structure of data.

                In summary, fundamental limitations in statistical methods, measurement, and design suggest the face-value interpretations of reported subtypes are largely undermined if not unfounded, suggesting their use in DSM should be discontinued.                 

1.            Linscott RJ, Allardyce J, van Os J. Seeking Verisimilitude in a Class: A Systematic Review of Evidence That the Criterial Clinical Symptoms of Schizophrenia Are Taxonic. Schizophr Bull 2009.