Early Infections of Toxoplasma gondii and the Later Development of Schizophrenia

Abstract

Early exposure to several infectious agents has been associated with the later development of schizophrenia. Two recent studies assessed in utero or early postnatal exposure to Toxoplasma gondii. In one study of 63 individuals, who developed schizophrenia spectrum disorders, maternal sera obtained during pregnancy showed an increased risk (OR 2.61) of having IgG antibodies to T. gondii. In the other study of 71 individuals who developed schizophrenia, sera obtained shortly after birth also showed an increased risk (OR 1.79) of having IgG antibodies to T. gondii. Causal linking mechanisms are at present speculative but include possible direct effects of maternal IgG on the developing central nervous system (CNS) of the offspring. Additional studies are underway.

Introduction

There is a widely held belief that many factors of etiological relevance to schizophrenia act during gestation or early infancy.¹ It is therefore not surprising that much of the research associating infectious agents with schizophrenia has focused on exposure during this period of life. Since the work of Mednick and others linking intrauterine exposure to the influenza virus to the risk for schizophrenia in adult life,^2–4 early exposure to several other viruses has also been associated with the development of schizophrenia. These include polio,⁵ measles,⁶ varicella-zoster,⁶ rubella,⁷ and herpes simplex virus type 2,⁸ as well as exposure to unspecified CNS viral infections during childhood.⁹ Periconceptional maternal genital infections have also been implicated as a risk factor.¹⁰ Much of the research linking infections to schizophrenia has been reviewed elsewhere.¹¹

In addition to these viruses, 2 recent studies have addressed the possible association of in utero or perinatal exposure to Toxoplasma gondii, a protozoan parasite, and the development of schizophrenia or other adult-onset psychiatric disorders.^12,13 This article will compare these studies and speculate on some additional directions for such research.

METHOD

Whereas many previous studies of the early exposure to infectious agents in schizophrenia had to rely on national statistics on reported cases of infections or data from case notes, both studies of T. gondii discussed in this article had the advantage of access to maternal sera or blood samples drawn from the child a few days after birth. Both studies also had the possibility of linking these samples to records of psychiatric care during adult life. The details of these studies are given in the original publications and are summarized in table 1.

Table 1.

Serology-based Studies of Early T. gondii Infections and Schizophrenia Risk

	Brown et al12	Mortensen et al13
Study population	63 cases of schizophrenia and schizophrenia spectrum disorders	71 cases of schizophrenia
		342 cases with other psychoses or affective disorders
	In most cases, diagnoses based on research interview	Clinical diagnoses made by treating senior psychiatrist
	123 controls	684 controls
Setting	1959–1967 birth cohort from the Kaiser Permanente Health plan, California, including treatment records and maternal sera drawn during pregnancy	All children born in Denmark 1981 or later, including treatment records for cases, controls, and their parents and sibs; neonatal blood samples from the national PKU-screening program
Design	Nested case-control study	Nested case-control study
Exposures measured	Toxoplasma gondii IgG and IgM titres from the maternal sera	Toxoplasma gondii IgG (maternal origin) and IgM titres from the child's neonatal blood samples
Confounders examined or controlled for	Age, gender, maternal age, maternal SES, ethnicity and education, timing of maternal sample	Age, gender, maternal age, psychiatric treatment history of parents and sibs, year, and place of birth
Association with elevated anti-Toxoplasm gondii IgG	OR 2.61 (1.00–6.82)	OR 1.79 (1.01–3.15)
		No significant association with schizophrenia-related disorders or affective disorders

Note: PKU, Phenylketonuria; SES, Socio-economic status.

Both studies measured Toxoplasma IgG and IgM titres. In Brown et al, the sera were of maternal origin. In the Danish study, by contrast, the source of the sera was from the child within a few days of birth¹²; thus, the IgG measured in these samples would have been maternal in origin because IgG passes through the placenta, whereas the IgM would have been from the child. Maternal IgG can persist in a child for several months, with a half life of 1 month, whereas an infected newborn child will only begin producing T. gondii-specific IgG at approximately 3 months.¹⁴ It should be noted that, in both studies, the presence of IgG does not mean that the mother had been infected during pregnancy but rather that she had been infected with T. gondii at some point during her life before the sample was drawn. Other studies have suggested that only a small percentage of such women were, in fact, infected during pregnancy.¹⁵ An increased IgM titre in the Danish study, by contrast, would indicate that the child had been infected.

Several confounders were examined and controlled for in both studies. Because both studies are relatively small, however, and because a genetic predisposition toward schizophrenia could not be controlled for in the US study and could only be controlled for using crude indicators in the Danish study, it cannot be ruled out that genetic or other risk factors may explain an association between maternal Toxoplasma IgG titres and schizophrenia.

Results and Discussion

Both studies found a modest association between IgG antibodies to T. gondii and schizophrenia-related disorders. In the study of Brown et al, the results achieved statistical significance for the entire sample of schizophrenia-related diagnoses (OR 2.61; 1.00–6.82). The result for schizophrenia alone was not reported in that article. In the study by Mortensen et al,¹³ the association was significant for schizophrenia (OR 1.79; 1.01–3.15). The association with schizophrenia-related diagnoses and affective disorders, respectively, was much weaker and not close to being statistically significant.

No increase in IgM antibodies was found in either study, suggesting that there was no evidence of acute infection in either the mother or the child. Such acute infections are, of course, known to have deleterious effects, usually evident shortly after birth, including congenital malformations, retinal damage, cognitive deficits, etc. The study by Lebech et al,¹⁵ which screened 90 000 pregnant women and births for T. gondii-specific IgG and IgM, reported that <1% of the cases had actually been exposed to an acute T. gondii maternal infection in utero; so it is not surprising that no increase in IgM was found in these 2 studies. At present, therefore, the evidence suggests a linkage between maternal prevalence of past infection with T. gondii and an increased risk of the development of schizophrenia in her offspring but not a direct effect of an acute maternal infection.

Causal mechanisms that could explain such an association are speculative. One possibility is a direct effect of the maternal IgG on the fetus, or an effect of a latent infection, because it is known that women with high IgG may have latent infections. If there is a direct effect of circulating anti-T. gondii IgG, one could speculate that the pathogenic mechanism is similar to what is hypothesized to occur in some paraneoplastic CNS manifestations. Here symptoms, including psychiatric symptoms, are related to the levels of circulating tumor-specific antibodies, and a causal hypothesis is that the tumor expresses proteins that normally are expressed only in the brain and are recognized by the immune system as foreign when expressed outside the immunologically privileged brain.¹⁶

In addition, there is evidence that autoantibodies can affect cognition, emotion, and behavior and are crucial in the pathogenesis of CNS changes in systemic lupus erythematosus, partly due to the binding of antiantibodies to the N-methyl-D-Aspartate (NMDA) receptor.^17–19 This could, at least by analogy, suggest possible mechanisms through which anti-T. gondii IgG could affect the brain in a way relevant to schizophrenia.

Another possible explanation is that the presence of maternal IgG reflects a lifestyle that increases the probability of the child being exposed to T. gondii after birth. For example, seropositive women may be more likely to have cooking practices that are conducive to food-borne infection with T. gondii, or more likely to expose her children to cats, the prime vector of the infection.

Another example of the effects of a maternal infection with an intracellular organism on the developing CNS of the offspring is cytomegalovirus (CMV). Infections with CMV in utero have been shown to reduce the expression of the NMDA receptor in the offspring of mice exposed to CMV through their mother²⁰; this receptor is known to be important in normal brain development, including development of the hippocampus.²¹ Similarly, infections in adult mice with T. gondii can affect neurotransmission through increased levels of dopamine and homovanillic acid.²² One could speculate that similar neurotransmitter mechanisms may be operant, eg, through the reactivation of a maternal T. gondii infection. It is also possible that maternal antibodies against T. gondii are an indicator of increased infections with other organisms that spread through similar mechanisms, ie, as a zoonosis from cats or through undercooked meat.

Another possible mechanism through which exposure to infectious agents, including T. gondii, could contribute to schizophrenia risk is through early sensitization of the immune system, as hypothesized by Müller and Schwarz.²³ This might produce an imbalance of the immune response found in some patients with schizophrenia; this immune constellation, in turn, could lead to the accumulation of kynurenic acid in the CNS, which, due to its NMDA antagonistic properties and could lead to cognitive dysfunction and psychotic symptoms.

Although these explanations are at present speculative, the association between maternal antibodies against T. gondii and schizophrenia is consistent with studies identifying T. gondii as a plausible candidate organism to affect the brain. As reported elsewhere in this issue, many studies have found a higher frequency of T. gondii antibodies in schizophrenia patients compared with controls; exposure was measured more directly than is sometimes the case when recording exposure in epidemiological studies of schizophrenia. In addition, studies have suggested that growing up in an environment with cats is associated with an increased risk of schizophrenia, suggesting a role for early infections.²⁴ It is likely that an early infection with T. gondii in itself may not be a sufficient cause of schizophrenia, but it may be one factor that interacts with a genetic predisposition to schizophrenia, possibly, acting at particularly vulnerable stages of neurodevelopment.

We are pursuing this research in an expanded study using the same Danish data sources as described in this review. The study, now in progress, includes more than 1300 individuals with schizophrenia. In this study, we are including risk genes for schizophrenia (and other relevant gene variants) and assessing the infection status in the neonatal samples by measuring specific IgG levels, mainly of maternal origin, and IgM, indicating that the newborn has been exposed to an infectious agent. We will also measure cytokines and other inflammatory markers; such markers can indicate both the response to an infection and exposure to hypoxia. These findings will be assessed in the context of other known risk factors for schizophrenia, which may provide valuable new insights into the etiology of this disease.

The task of identifying the causal pathways of schizophrenia will not be easy, however, even with the advantage of resources provided by large population-based registers and biobanks. If, eg, the timing of the infection is crucial, it may be difficult to obtain the most informative samples, which would include information and biological material from conception to the onset of disease. Despite the limited current empirical evidence identifying T. gondii as a risk factor for schizophrenia, we believe that it is a promising candidate and that its role as a risk factor and possible interaction with other risk factors warrants further study.