A number of deficits have been consistently identified in both adolescent and adult populations that perhaps reflect disease traits. Future research needs to focus on these and employ multimodal tests in pristine patient groups, with a view to identifying reliable biomarkers.

A succinct review of the literature pertaining to cognition in both adult and childhood populations is synthesised following Medline and PsychINFO searches using key-terms including ‘cognition’, ‘bipolar disorder’, ‘neurocognitive’ ‘child’, ‘adolescent’ and a range of neuropsychological domain names. In addition, literature known to the authors was scrutinised and relevant references further pursued.

In the ongoing quest for improved diagnostic markers of bipolar illness, the focus of research has gradually shifted to examining the onset of mood difficulties early in life and investigating the potential corollaries of such early onset such as cognitive impairment, disruption of social and emotional functioning, and constriction of quality of life. This article considers the disruptions to cognitive functioning that accompany bipolar disorder and compares adult and child profiles to ascertain the likelihood of identifying a neurocognitive biomarker of the illness.

Introduction

Classic bipolar disorder (BD) is a recurrent, episodic, cyclical illness characterised by episodes of major depression, interspersed with periods of hypomania or mania and periods of relative mood stability called ‘euthymia’. Its strong heritability suggests a significant genetic loading. However, the clinical heterogeneity of the disorder remains a constant challenge to diagnosis and research. It is likely that its aetiology, in common with many psychiatric disorders, is varied, combining biological, social and psychological factors. As with many disorders, there is a strong desire on the part of researchers and clinicians to identify a specific ‘marker’ or diagnostic sign, and in the case of BD, this has been accompanied by a shift in the focus of research towards examining younger populations.

In this article, the literature pertaining to this search for a marker in BD is reviewed, with particular emphasis on juvenile populations.

Genetic studies are increasingly employed in researching neuropsychiatric disorders to assist with determining causality. In BD, findings such as high concordance rates (40–97%) in identical twins, compared with the much lower rates (14%) in dizygotic twins, drive the search for better identification of transporter, as well as susceptibility, genes (Kieseppa et al. 2004). Genes that have been investigated in BD as potential susceptibility candidates include COMT, DAT and HTR4, amongst others (Kieseppa et al. 2004; Edvardsen 2008). Specific gene loci have been associated with subgroups of patients, such as those who respond to lithium (MacQueen 2005; Consoli 2007), and those with psychotic or suicidal features (Cheng 2006). As yet, however, a single or dominant gene contribution to bipolar disorder has not been identified and studies continue to focus predominantly on the exclusion of specific genes as transporters.

The lack of consensus surrounding the diagnostic boundaries of BD and the lack of a single genetic marker, has resulted in the expansion of studies into a search for endophenotypic markers for bipolar disorder. The definition of an endophenotype has been broadened from that described by Gottesman & Gould (2003), to include a replicable state and trait feature of the disorder. To date, several cognitive impairments, identified in the profiles of adult populations with BD, have already been proposed as endophenotypes. These include sustained attention deficits, difficulties in executive functioning and verbal memory deficits (Clark et al. 2002; Martinez-Aran et al. 2002; Martinez-Aran et al. 2004; Martinez-Aran et al. 2004; Clark et al. 2005; Olley et al. 2005; Malhi et al. 2007). MacMann and Barnett (1997; 1999) have highlighted many of the errors that potentially occur with profile analysis. The strengths and weaknesses of a profile analysis approach are frequently highlighted in relation to breaking down Wechsler scales into their subtest scores (McDermott 1989; Watkins 2005), and the caveats raised by such discussion should be kept in mind when interpreting the neuropsychological profiles of adolescents. Significant examples which relate to the population under discussion include correlation with associated skills (such as language) and motivation to perform.

Recently, Burdick et al (2006) investigated neurocognitive profile, and the potential for this to be a stable BD endophenotype, in a follow-up of 16 patients with schizophrenia and 16 patients with bipolar disorder who were participating in a larger, long term study. In the investigation, patients first completed neuropsychological assessment after 14/15 years of follow-up and they repeated the assessment at 19/20 years follow-up. The test battery used for assessment included measures of executive functioning, attention and short and long term memory. Interestingly, patients with BD were found to demonstrate improved performance in both short and long term delayed free recall and in number of perseverative errors and verbal fluency. No improvement was observed in measures of attention and therefore the authors hypothesised that attention per se may represent a ‘stable’ marker in bipolar disorder. Attention deficits are certainly found across mood states of bipolar disorder (Brand and Jolles 1987; Trichard et al. 1995; Clark et al. 2002) and may contribute to reduced executive functioning in BD (Sapin et al. 1987; Sweeney et al. 2000; Martinez-Aran et al. 2004; Malhi et al. 2007) as many tests evaluating this domain rely on sustained attention to a stimulus.

Memory deficits have also been consistently found in BD (van Gorp et al. 1999; Krabbendam et al. 2000; Rubinsztein et al. 2000), with verbal memory, in particular, shown to be impaired across mood states (Martinez-Aran et al. 2004; Malhi et al. 2007) and, akin to attention, this clearly warrants further investigation. However, different aspects of memory need to be studied in finer detail, as it has already been proposed that executive function and working and declarative memory may be the most viable endophenotype in subtypes of bipolar disorder, based on a comparison of BD patients with and without psychotic symptoms (Glahn 2006).

Verbal fluency is also compromised in bipolar disorder (Ali 2001), although there are indications that this fluctuates according to mood state (Martinez-Aran et al. 2004). Overall, it appears that verbal fluency, executive functioning (in particular perseveration), memory and selective attention are the domains most impaired in adults with bipolar disorder (Cahill 2007). Findings that these deficits persist even during euthymia (Martinez-Aran et al. 2004; Olley et al. 2005; Malhi et al. 2007) seem to support the proposal that these could serve as trait markers of the illness at least after the initial episode.

A recent meta-analysis has posited that response inhibition and fronto- temporal/fronto limbic cognitive impairments are the most prominent endophenotypic markers of bipolar disorder (Bora 2009). Increasingly, studies are attempting to discern more subtle cognitive impairments in bipolar disorder. Reward processing, sensitivity to negative feedback, short term spatial storage and response consistency have all been found deficient in research studies to date but have not yet been widely replicated across mood states and populations (Yechiam 2008; Roiser 2009).

Efforts to discern an endophenotypic marker in studies of adult BD often run into difficulties when disease related characteristics are considered as putative causes of cognitive deficits (Savitz 2005). For example, it is reasonable to link the long term and wide-ranging use of medication, frequently associated with bipolar disorder, to the cognitive deficits that are observed when patients are undergoing treatment (Donaldson 2003). In addition, bipolar disorder is associated with a high level of comorbid substance abuse, particularly cannabis, which, in and of itself, is frequently associated with cognitive impairment (Cahill 2006). Further, disease characteristics that may account for cognitive impairment include symptomatic changes that affect functioning, such as reduced motivation and concentration in the case of depression, and increased disinhibition in mania.

Essentially, as a person with BD gets older, environmental factors and illness-related sequelae, such as scarring and medication effects, begin to exert a stronger influence in parallel with a reduction in the proportional influence of genes. Partly as a result, research aimed at identifying markers of the illness is increasingly examining younger populations, based on the logic that a marker of BD will be more readily identifiable in this age group. The emergence of a genetic loading influencing age of onset has encouraged investigations in this area (Faraone 2004) as has the increased association in preliminary linkage studies of early onset bipolar disorder (Mick 2009)