Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

Hippocampal Site Seen Revving Up as Psychosis Emerges

24 Sep 2009

25 September 2009. Researchers have some idea of the broad areas of the brain that act up in schizophrenia, but finer resolution in these regions has been hard to achieve. When a team of investigators, led by Scott Small of Columbia University in New York City, used state-of-the-art functional magnetic resonance imaging to scrutinize the usual suspects, an area in one stood out: the CA1 subfield of the hippocampus. In the September Archives of General Psychiatry, the researchers report that this subfield becomes overactive in schizophrenia and that this dysfunction foretells which high-risk subjects will develop psychosis. Their evidence also suggests that psychotropic drugs are not to blame.

Recent advances in high-resolution fMRI, with gadolinium as the contrast agent, made the study possible. They enabled first author Scott Schobel of Columbia University and his colleagues to map cerebral blood volume in small brain areas. This let them gauge oxygen use and hence, activity levels, in these subregions.

Schobel and colleagues shone their fMRI searchlight on regions of interest fingered in prior studies of psychosis (see SRF related news story; related news story; related news story; related news story; related news story; related news story). These include the hippocampus, basal ganglia, frontal lobes, and amygdala (Shenton et al., 2001; for more recent reviews, see Gur et al., 2007; Allen et al., 2008).

Snapshots of psychosis

The researchers used a multipronged approach to pinpointing the brain regions that might underlie schizophrenia and related psychoses. First, they imaged the brains of 18 subjects who met criteria for schizophrenia or schizoaffective disorder. They compared their cerebral blood volume in the different subregions with that of an equal number of healthy subjects who were matched to them on age and sex.

The findings suggest that, in schizophrenia, oxygen use increases in the hippocampal CA1 subfield and the orbitofrontal cortex. In contrast, it seems to decrease in the dorsolateral prefrontal cortex.

However interesting, these differences could reflect the harm done by chronic psychosis, and Schobel’s team wanted a snapshot of early disease processes. This led them to study 18 prodromal subjects who met clinical criteria for being at ultra high risk of developing psychosis. These subjects presented symptoms that did not quite rise to the level seen in psychosis. For example, prodromal subjects might profess unusual ideas that seem less compelling than delusions.

During the two years of follow-up, seven of the prodromal subjects became psychotic. Of the three dysfunctional areas seen in subjects with more established disease, only the CA1 subfield appeared awry in these subjects. Notably, high activity in this area predicted the eventual emergence of psychosis. In fact, it did so with a positive predictive value of 71 percent and a negative predictive value of 82 percent. Not only do these findings yield insight into the earliest stages of psychosis, they raise hopes for a marker to flag those at greatest risk with the idea of someday preventing the disease from taking its toll.

While something had gone wrong in the CA1 area in the two clinical groups, its relationship to symptoms remained unknown. To assess symptom severity, Schobel and colleagues used the Positive and Negative Symptom Scale in subjects with schizophrenia and the Scale of Prodromal Symptoms in the high-risk group. For positive symptoms, the two scales rely on similar items, enabling the researchers to combine them into one measure.

Regression analyses showed that positive symptom scores, particularly delusions, correlated with CA1 activity in the two groups combined (β = .53, P = .01). However, these symptoms appeared unrelated to activity in the orbitofrontal cortex or the dorsolateral prefrontal cortex.

Unfortunately, the two scales differed too much in their assessment of negative symptoms to combine them, limiting Schobel and colleagues to looking at these symptoms in each group of subjects alone. They found that, in those with schizophrenia, negative symptoms appeared independent of activity in the three regions. A different story unfolded in the prodromal group. Their CA1 blood volume did relate to negative symptoms (β = .59, P = .03), including avolition and social dysfunction.

Saying no to drug confounds

In addition to teasing out early versus later disease processes, Schobel and colleagues faced the usual challenge of ruling out possible medication effects on brain activity, since most diagnosed subjects receive psychotropic drugs. However, in this study, only one prodromal subject who developed psychosis was undergoing treatment with antipsychotic medication at baseline. Even so, to boost confidence that medication had not influenced their findings, the researchers compared CA1 blood volume in the three prodromal subjects who were taking antipsychotic medication with that in the 15 subjects who were not. They found no differences between them. Other analyses discounted antidepressant use as a contributor.

Of course, animal studies give researchers ultimate control over subjects’ medications, so the researchers turned to mice. Each day for three weeks, they gave five mice the atypical antipsychotic risperidone, while five others received the drug vehicle alone. They imaged their brains before and after treatment. In the CA1 subfield and, indeed, all regions of the hippocampus, the two groups showed similar activity. Thus, the possibility of a medication confound appears, in this instance, to be less likely.

Schobel and colleagues assert that their study suggests that schizophrenia and related psychoses target the CA1 site in particular. The abnormal activity they observed in established psychosis, which also predicts progression from the prodrome to outright disease and correlates with psychotic symptoms, taken together, are evidence that schizophrenia involves “a basal hypermetabolic state in the hippocampus.” This conclusion contrasts with the traditional view that the hippocampus does too little in schizophrenia, although recently the notion of a hyperactive hippocampus has been gaining a foothold (see SRF related news story). The new findings may further its momentum.—Victoria L. Wilcox.


Schobel SA, Lewandowski NM, Corcoran CM, Moore H, Brown T, Malaspina D, Small SA. Differential targeting of the CA1 subfield of the hippocampal formation by schizophrenia and related psychotic disorders. Arch Gen Psychiatry. 2009 Sept;66(9):938-46. Abstract