We wish to expand on the recent demonstration of distinct and functionally relevant cognitive profiles in bipolar disorder by Burdick et al. (Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014). Our group has previously shown similar cognitive heterogeneity in a sample of younger depressed subjects diagnosed with anxiety disorder, depressive disorder, bipolar disorder or first-episode psychosis (Hermens et al. Reference Hermens, Redoblado Hodge, Naismith, Kaur, Scott and Hickie2011). As shown in the study by Burdick et al. (Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014), our cluster analysis also revealed three distinct neuropsychological subgroups. Further, several cognitive measures used in discriminating group membership in Hermens et al. (Reference Hermens, Redoblado Hodge, Naismith, Kaur, Scott and Hickie2011) were also useful in longitudinally predicting functional disability early in the course of major depression, bipolar disorder and schizophrenia (Lee et al. Reference Lee, Hermens, Redoblado-Hodge, Naismith, Porter, Kaur, White, Scott and Hickie2013). Importantly, in both our cluster analysis (Hermens et al. Reference Hermens, Redoblado Hodge, Naismith, Kaur, Scott and Hickie2011) and longitudinal follow-up study (Lee et al. Reference Lee, Hermens, Redoblado-Hodge, Naismith, Porter, Kaur, White, Scott and Hickie2013) the cognitive changes were apparent irrespective of the specific diagnosis. Therefore, we propose that the available evidence indicates that neuropsychological subgroups are not only evident in, and have functional relevance for, the major psychoses, but across the spectrum of major mood and psychotic illnesses, even at the early stages of illness progression.

Of particular note, visual learning and memory deficits appear to be specific to subjects with global cognitive impairment across both studies (Hermens et al. Reference Hermens, Redoblado Hodge, Naismith, Kaur, Scott and Hickie2011; Burdick et al. Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014). Moreover, the globally impaired group with visual memory dysfunction was more occupationally disabled in Burdick et al. (Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014), similar to how visual learning and memory were more impaired in those who were more functionally impaired in Lee et al. (Reference Lee, Hermens, Redoblado-Hodge, Naismith, Porter, Kaur, White, Scott and Hickie2013). Thus, visual learning and memory may be particularly sensitive to overall cognitive and functional decline across the major psychiatric illnesses.

On balance, neuropsychological clusters across the major psychiatric illnesses are likely to provide more homogeneous cognitive phenotypes that are more useful than previous syndromally classified cohorts (Hickie et al. Reference Hickie, Scott, Hermens, Scott, Naismith, Guastella, Glozier and McGorry2013) in investigations of: (i) aetiological (e.g. genetic) and biological mechanisms (e.g. neurobiology); (ii) prognosis (e.g. functional outcome); and (iii) targeted interventions (e.g. cognitive remediation), which hitherto have largely been obscured by clinical heterogeneity.