Imaging studies of attention processes in people with schizophrenia have reported widespread functional abnormalities in the attention network (Andreasen, 1995 1; Goldman-Rakic, 1988 2). For example, using an oddball paradigm to assess sustained attention, a functional magnetic resonance imaging (fMRI) study found decreased activity in superior temporal and frontal gyri in addition to the cingulate cortex and thalamus 3. Similarly vast areas of decreased activity during a target detection task were found in patients with schizophrenia that included areas in the limbic cortex, striatum, anterior cingulate gyrus and prefrontal cortex 4. During a selective attention task, decreased activation was also found in several areas of the dorsolateral prefrontal cortex (DLPFC), anterior cingulate in addition to parietal areas 5. Evoked related potential recordings have also shown reduced hippocampal activation, in addition to reductions in the anterior cingulate gyrus and basal ganglia during orienting of attention 6.

Because attention deficits are a key and a stable feature of schizophrenia 7, those who share the same genes for schizophrenia may also show similar patterns. However, findings from neuropsychological studies of those who are at genetic risk for schizophrenia have been inconsistent 89, partly because these studies often rely on the assumption of elevated genetic risk for schizophrenia based on population levels of chance. For example, a study of 20 subjects with 1 relative with schizophrenia will effectively deal with a risk increase from 1% to 2–4%, which does not necessarily reflecting strong genetic load. Another possibility for these inconsistencies may be due to the subtlety of the deficits that make them undetectable behaviourally. Neurophysiological measures of attention, such as eye tracking, may be a better indicator of attention deficits as they are more proximal to the biological mechanisms. For example, eye tracking studies have suggested that abnormalities in antisaccades and smooth pursuit are possible endophenotypes for schizophrenia 1011. In addition, electrophysiological abnormalities (i.e., late P300) during target detection despite normal task performance has been reported in first-degree relatives of people with schizophrenia 1213. This is evidence to suggest that these enduring traits related to attentional processing are also present in those who are genetically related to people with schizophrenia.

Neuroimaging studies of high-risk populations have recently begun to investigate a subgroup of relatives of patients with schizophrenia, sometimes referred to as presumed obligate carriers (POCs) in order to segregate a more homogeneous group 1415. These individuals are the first-degree relatives of people with schizophrenia who, although they do not manifest the disorder, are in direct lineage of it. Thus, POCs are presumed to transmit the genes for schizophrenia. Structural neuroimaging findings have suggested volumetric differences between POCs and healthy controls 1416. For example, it was found that in contrast to healthy, non-carrier siblings, POCs have reduced amygdalohippocampal complex volumes 14. At present, we are aware of only one published functional neuroimaging study of POCs. Using positron emission tomography (PET), Spence et al. tested neural function associated with verbal fluency. The authors reported that the POCs had a pattern of frontal activity similar to that of the individuals with schizophrenia (i.e., reduced bilateral activation of the dorsolateral prefrontal cortex), and that it is qualitatively different from that of controls. The authors suggested that this abnormality may be due to either a loss of asymmetry or 'hyperinnervation' between the prefrontal cortices 17.

The present study attempts to extend the literature on POCs of schizophrenia by examining the neural correlates of attentional abnormalities using fMRI. fMRI provides the ability to investigate the underlying processes that may be dysfunctional despite intact cognitive behaviour. The aims of this study were: (1) to contribute to the growing literature on neural abnormalities present in people who have unexpressed genetic predisposition to schizophrenia, and (2) to demonstrate the utility of fMRI in characterising intermediate phenotypes for schizophrenia.

Because attention deficits have been widely reported as possible intermediate phenotypes between genotype and clinical symptoms of schizophrenia (i.e., present in those who are at high genetic risk), we chose to investigate the neural mechanisms that are associated with attention processes in the POCs of schizophrenia. Using fMRI, we measured sustained, selective and divided attention processes in order to investigate whether the genetic predisposition to schizophrenia is associated with differential neural activity that could be used as biomarker for schizophrenia. Based on previous studies, we expected that abnormalities in a widely distributed network of brain regions including the parietal, prefrontal and temporal cortices, the hippocampus, the reticular formation, and the striatum that underlie attention would show a differential response in POCs 18. Prefrontal areas, particularly the DLPFC and ventrolateral prefrontal cortex are important areas for attention processes (i.e., sustained, selective, divided), in addition to executive functioning processes (i.e., inhibition and working memory) 19202122. In addition to the prefrontal cortex, the anterior cingulate gyrus has also been suggested to be involved in target detection, response selection and inhibition, error detection, performance monitoring, and motivation 232425. The reticular formation, which sends input to the thalamus, appears to modulate attention and filter interfering stimuli 26, while the hippocampus has been suggested to play a role in attention, possibly as a 'comparator' by processing varieties of input and resolving conflict 18.

We included POCs in this study who are healthy relatives who lie between two affected generations (e.g., have both a parent and a child with schizophrenia), thus increasing the likelihood that these individuals carry genes linked to schizophrenia. Because the existing literature on POCs report that this population shares similar neural abnormalities present in those with schizophrenia, we expected to find abnormalities in neural activity associated with attention processes in our group of POCs 172728.

Table 2 summarises the mean reaction times for each condition per group. The POCs were significantly slower in their reaction times than the normal controls (p = 0.03) during the letters task. The groups did not differ in the balls or dual task.

This study reports preliminary findings from a cohort of individuals who are the presumed obligate carriers of genes linked to schizophrenia. The performance of this group was compared to that of healthy age/gender/IQ matched individuals on three measures of attention: sustained, divided and dual attention. We hypothesised that if these individuals are carriers of genes that code for schizophrenia then their performance on these tasks would differ from healthy participants and be more similar to the reported findings in those with schizophrenia. As predicted, the POCs showed altered neural activation patterns relative to the healthy participants during sustained, selective and divided attention as measured by the BOLD response. Compared to the healthy participants, the POCs demonstrated a general decreased pattern of activation during the three attention tasks despite unimpaired task performance.

Our findings showed that POCs have altered neural function in the attention network that is similar to the abnormalities previously reported in patients with schizophrenia. While differences in activation appeared to be widespread in all three tasks of attention, the POCs had attenuated response in prefrontal areas and parahippocampal gyrus/entorhinal cortex during all three tasks. Prefrontal abnormalities, including those in the anterior cingulate gyrus, have been suggested to be the best indicator of genetic vulnerability to schizophrenia as they have been found in the unaffected relatives of those with schizophrenia 3435. Temporal lobe abnormalities, particularly in the medial portion, have also often been reported in people with schizophrenia and have been associated with the psychotic and neuropsychological symptoms of schizophrenia 36. Both structural 37 and functional 3438 temporal lobe abnormalities have also been reported in the relatives, therefore, indicating that these may be an indicator of risk for schizophrenia. In the present study, decreased activity in the parahippocampal gyrus of the POCs was observed during all three conditions of attention. The parahippocampal gyrus surrounds the hippocampus and lies in the medial temporal lobes of the brain. It has been argued that a disordered 'hippocampal formation' (i.e., hippocampus and parahippocampal gyrus) is the fundamental abnormality in the pathophysiology of schizophrenia based on decreased size, reduced N-acetyl aspartate, altered metabolic/synaptic activity, presence of memory deficits (attributable to hippocampal dysfunction), and alterations in receptors 39404142. In addition to the role of the hippocampus in memory formation, it has also been suggested that it plays a crucial role in integrating multimodal sensory inputs, and in resolving conflicts between expectancies and the current perception 4344. This is in accord with the role of the hippocampus in encoding attention proposed by Mirsky et al. 18.

The widespread nature of disrupted neural activity in the present group of POCs provides further support for the idea that schizophrenia is an inter-region disconnectivity syndrome 45. Since frontal abnormalities are the cardinal feature in the pathophysiology of schizophrenia, it is not surprising that parietal and temporal abnormalities exist because reciprocal connections exist between the PFC, parietal, temporal and cingulate cortices 46. In a recent study that investigated functional connectivity in schizophrenia, it was found that even during rest (i.e., without a directed task), disconnectivity was found globally in those with schizophrenia 45. Of the 177 connectivities investigated, 158 showed decreased activity in the individuals with schizophrenia compared to the healthy participants. Moreover, functional connectivity abnormalities have also been found in individuals with high genetic risk for schizophrenia, such that altered connectivity was found between the parietal and prefrontal network, and the prefrontal cortex and cerebellum 47.

Our findings also showed that during sustained and selective attention, the POCs had less right hemispheric activation in the fronto-temporo-parietal network. Posner and Petersen posited the notion of a right-lateralised vigilance system that is due to greater innervation of ascending noradrenergic pathways in the right hemisphere 48. In this context, decreased activity in the POCs may reflect hypofunction of noradrenergic pathways leading to hypovigilance and distractibility. Dysfunctional noradrenergic pathways, in addition to abnormalities in dopaminergic pathways, have also been proposed in schizophrenia 49 and have been targets for more novel pharmacotherapies (e.g., quetiapine) 50. Reduced BOLD signal in these regions is in accord with the frequently reported hypofrontality in schizophrenia. Despite inconsistencies in the literature, studies have generally provided evidence for attenuation of neural activity in the frontal cortex of people with schizophrenia 51. Thus, our findings of significantly decreased activity in POCs in several areas of the prefrontal cortex during all three conditions are consistent with previous findings 52. Of note, although disturbances in the frontal lobes have been the hallmark of schizophrenia, hypofunction has also been reported in the parietal and temporal lobes 53.

While we observed a general decrease in regions that underlie attention in the POCs during attention processes, these findings must be interpreted in light of some areas that show increased response (i.e., temporal regions). Enhanced activity in conjunction with hypofunction in the attention network despite unimpaired behaviour performance may indicate recruitment strategies in the POCs. Recruitment or compensatory mechanisms in schizophrenia have been previously reported in the literature 1354. For example, it was found that during a working memory task, depending on task demands, greater activation in the PFC was found along with decreased activation in the posterior parietal cortex and vice versa 55. It was suggested that function in the PFC was dependent on availability of other areas within the network. Additionally, the idea that neuroplastic processes may also be responsible cannot be discounted. It has been proposed that structural abnormalities may trigger regeneration or re-organisation of function 56. Network modelling or similar techniques were beyond the scope of this report, but are important for future studies.

Lastly, although our findings confirm our hypothesis and are in accord with the current literature on POCs of schizophrenia, interpretation must be made with caution given our small sample size. Although the implications of having a small sample size pertains to generalisability of these findings, our findings are in concordance with previous neuroimaging studies in POCs with very small sample sizes, e.g. 6 for Steel et al. (2002) 14; 11 for Chua et al. (2000) 57; 9 for Sharma et al. (1999) 58; 10 for Spence et al. (2000) 17. Nevertheless, because this is the first fMRI study of POCs and the first to use this task, replication of these findings in a larger cohort of POCs is necessary in order to validate these results.

These preliminary findings suggest the presence of abnormal patterns of neural activity in the POCs despite unimpaired performance during tasks of attention. Altered BOLD response was observed in fronto-temporo-parietal regions that have been shown to subserve attention processes. Our findings in the POCs during attention processes parallel those previously reported in people with schizophrenia, and may reflect compensatory mechanisms necessary to successfully attend to stimuli. In addition to replication of these findings, future research should also examine whether this abnormality is the cause, consequence, or compensation for the pathophysiology of schizophrenia 59 and how this may translate to functional disconnectivity.

The authors declare that they have no competing interests.

FF conceived of the study, collected, analysed and interpreted the data, and drafted the manuscript. RGM made contributions to the design of the study. TR provided assistance in the image analyses. CMcD provided the clinical diagnoses, screening and categorisation of the participants. RMM made substantial contributions in the drafting and revision of the manuscript. All authors read and approved the final manuscript.