Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

SIRS 2014—Epidemiology Digs Deeper Into the Infection-Psychosis Link

12 May 2014

May 13, 2014. The immune system is again gaining attention in the etiology of schizophrenia and other psychotic disorders (see SRF related news report; SRF news report), and has featured prominently in many recent scientific meetings. The 2014 meeting of the Schizophrenia International Research Society in Florence, Italy, was no exception. An April 7 session chaired by Christina Dalman of the Karolinska Institutet in Stockholm, Sweden, described recent epidemiological studies investigating the link between psychosis and different types of infections.

Viral pathogens

Seropositivity for the Epstein-Barr virus (EBV), a member of the herpes virus family, has recently been linked to psychotic experiences during adolescence. In the first talk of the session, Golam Khandaker, University of Cambridge, UK, presented data on the first longitudinal study to examine this association. Khandaker and colleagues used data from a subset of the Avon Longitudinal Study of Parents and Children (ALSPAC), a general population birth cohort of more than 14,000 people born in the UK in the 1990s. Antibodies to EBV were measured at age four, which revealed seropositivity in 25 percent of the sample. Khandaker reported that those with exposure to the virus were five times more likely to have had psychotic events (PEs) (17 percent of sample) at age 13 than those who were not seropositive (11 percent).

Exposure to herpes simplex virus 1 (HSV1), another member of the herpes virus family, has been linked to cognitive dysfunction in both control subjects and people with schizophrenia (Dickerson et al., 2003), so Khandaker and colleagues reasoned that IQ may mediate the effect of EBV exposure on later PEs. However, although EBV-exposed children did have lower IQs at age nine than controls, this effect was not significant when accounting for confounding factors and did not explain the increased risk for later PEs.

In a follow-up from the 2012 SIRS meeting (see SRF related conference report), Alan Brown, Columbia University, New York, discussed his recently published data on the influence of maternal influenza exposure on bipolar disorder in the Child Health and Development Study population-based birth cohort (Canetta et al., 2014). A strong link between a mother's influenza exposure and a greater risk of schizophrenia in her offspring has been found (see SRF related news report), but the diagnostic specificity of this finding has not been conclusively assessed, Brown said.

Although Brown and colleagues did not find an association between serologically documented maternal influenza and bipolar disorder as a whole in the offspring, a mother's infection did raise the risk of bipolar disorder with psychotic features fivefold, consistent with an earlier study in the same birth cohort. Combined with the schizophrenia data, the results suggest that prenatal influenza infection may raise the risk for psychosis more generally, though Brown noted that the inability to distinguish influenza infection during pregnancy from that which occurred prior to conception is a limitation of the study.

Hakan Karlsson, of the Karolinska Institutet described a large Swedish study that examined the effect of chronic maternal infections on the newborn immune system (Gardner et al., 2013). He reported that babies born to mothers with a Toxoplasma gondii infection have higher levels of acute proteins (an important component of the innate immune system) shortly after birth, demonstrating an immune reaction to her infection. In contrast, babies who later develop non-affective psychosis do not show this same elevation, suggesting that they don't mount a proper immune response. Karlsson and colleagues observed a similar effect when looking at babies of mothers with cytomegalovirus, but not with herpes simplex virus 1 or 2.

Beyond specific pathogens

Asa Blomstrom, also from the Karolinska Institutet, provided an update on data she had presented at the 2012 SIRS meeting (see SRF related conference report). In the current study, she investigated the link between infections (more generally) and psychosis by examining hospital admissions for infections in children born in Sweden from 1973 to 1997. After correcting for multiple confounding factors, she observed a small but significant association between hospital admissions for any infection up to age 13 and a subsequent diagnosis of non-affective psychosis (including schizophrenia) (Blomstrom et al., 2013). Bacterial infection had the strongest effect on risk for subsequent psychosis. Similar results were obtained when schizophrenia patients alone were examined.

An audience member presented an alternative take on the general hypothesis that infections themselves increase risk for psychosis, instead speculating that perhaps the elevated risk for non-affective psychosis was due to alterations in the microbiome resulting from the antibiotics used to treat the infections.

Urs Meyer, Swiss Federal Institute of Technology in Zurich, presented an overview of the use of animal models of immune activation, focusing on prenatal immune activation. He emphasized the role of animal studies in providing "experimental support" for human epidemiological studies. The elevated risk of schizophrenia conveyed by maternal infection does not seem to be pathogen specific, Meyer added, and highlighted four animal models of general maternal immune activation during pregnancy: influenza, polyinosinic/polycytidylic acid, lipopolysaccharide, and turpentine. "We should also think beyond the brain," said Meyer, reviewing data on metabolic changes resulting from maternal immune activation. All of these animal models are neurodevelopmental in nature and thus may be useful in a number of psychiatric illnesses thought to have a developmental origin, he added.

Discussant Robert Yolken, Johns Hopkins University, Baltimore, Maryland, concluded the session with an overview of the field of immune activation in psychosis. He highlighted the need for an improvement in the measurement of immune and inflammation molecules, pointing to mass spectroscopy as a potentially useful tool. In terms of risk factors for psychosis, Yolken emphasized low birth weight as one that needs to be further explored (see SRF related news report). It will be important to "get the message out that some aspects of these diseases are preventable," he said. Several strategies such as the monitoring and treatment of infections during pregnancy, immunizations, and safe sexual practices should be communicated to both doctors and patients, Yolken added.—Allison A. Curley.