Association between antibodies to multiple infectious and food antigens and new onset schizophrenia among US military personnel

https://doi.org/10.1016/j.schres.2013.10.004Get rights and content

Abstract

Introduction

Multiple studies have documented immune activation in many individuals with schizophrenia suggesting that antigens capable of generating a prolonged immune response may be important environmental factors in many cases of this disorder. While existing studies have found single-agent associations of antibodies to food and neurotropic infectious agents with schizophrenia, a simultaneous examination of multiple agents may shed light on agent interactions or possible etiopathogenic pathways.

Methods

We used traditional regression and novel statistical techniques to examine associations of single and combined infectious and food antigens with schizophrenia. We tested 6106 serum samples from 855 cases and 1165 matched controls.

Results

Higher antibody levels to casein were borderline significant in the prediction of schizophrenia (HR = 1.08, p = 0.06). Study participants with higher cytomegalovirus (CMV) IgG antibody levels had a reduced risk of developing schizophrenia (HR = 0.90; p = 0.02). While IgG antibodies to gliadin, Toxoplasma gondii, vaccinia, measles, and human herpesvirus-6 (HHV-6) showed no significant independent associations with schizophrenia, the increase in antibody levels to several combinations of agents, to include casein, measles, CMV, T. gondii and vaccinia, was predictive of an 18–34% increase in the risk of developing schizophrenia.

Conclusion

Certain patterns of antibodies, involving some agents, were predictive of developing schizophrenia, with the magnitude of association rising when the level of antibodies increased to two or more agents. A heightened antibody response to a combination of several infectious/food antigens might be an indicator of an altered immune response to antigenic stimuli.

Introduction

Schizophrenia is a pervasive neuropsychiatric disorder with uncertain etiology and pathogenesis, variable clinical presentations and outcomes. The etiology of schizophrenia is complex and is likely to involve both genetic and environmental components, which are very challenging to disentangle (O'Donovan et al., 2003, Maki et al., 2005, Tandon et al., 2008). Multiple studies have documented immune activation in many individuals with schizophrenia, suggesting that antigens capable of generating a prolonged immune response may be important environmental factors in this disorder (Moscavitch et al., 2009). These antigens are likely to interact with human leukocyte antigens (HLA) and the products of other immune response genes that have recently been found to increase the risk of schizophrenia in several different populations (Carter, 2009, Stefansson et al., 2009).

Existing studies examining single-agent associations of antibodies to food and neurotropic infectious agents with schizophrenia have yielded noteworthy but modest associations (Leweke et al., 2004, Yolken, 2004, Wang et al., 2006, Niebuhr et al., 2008a, Niebuhr et al., 2008b, Dickerson et al., 2010a, Severance et al., 2010, Niebuhr et al., 2011, Jin et al., 2012). Problems analyzing single-agent associations are compounded by the heterogeneity of IgG levels depending on the temporal proximity of the specimen collection to the onset of schizophrenia and subjects' demographic characteristics. One limitation in the analysis of environmental factors in a complex disorder such as schizophrenia is the lack of existing statistical techniques to evaluate the possible direct or indirect involvement of multiple environmental agents at different time points before and after symptom onset.

The use of regression techniques with multiple biomarkers is particularly difficult when the sample size is limited. Stepwise regression is a well-known method for dimension reduction that is widely applied and available in most commonly used statistical software packages. One of the major limitations of this algorithm is its inability to address multicollinearity. In practice, multiple independent variables used in a regression may have a high degree of correlation, making it difficult, if not impossible, to distinguish the effect of each agent on the dependant variable. As a result, the estimation and the test of statistical significance become unreliable due to a violation of the assumption of independence required for these tests. In the past several decades, a small number of regression methods that adopt regularization have been introduced: ridge regression (Hastie et al., 2001), subset selection, and principle component analysis. Recently, there has been an increasing interest in replacing the sample covariance with some sparse estimates of the true covariance or its inverse for high-dimensional regression problems (Witten and Tibshirani, 2009). Li and Niebuhr (2011) have recently developed a gradient noise orthogonal (GNO) method based on finding the most significant linear combination of the biomarkers in order to reduce the dimension.

We used traditional regression and novel GNO statistical techniques (Li and Niebuhr, 2011) to examine associations of single and combined infectious and food antigens with schizophrenia. Based on the biological plausibility and previously reported associations, four neurotropic viruses [cytomegalovirus (CMV), human herpesvirus 6 (HHV-6), measles virus, and vaccinia virus], an apicomplexan protozoa (Toxoplasma gondii) and two food antigens (casein and gliadin) were selected for this study (Albrecht et al., 1980, Leweke et al., 2004, Yolken, 2004, Wang et al., 2006, Dickerson et al., 2010a, Prasad et al., 2011, Jin et al., 2012, Severance et al., 2012).

Section snippets

Data collection

Data for US military service members who received medical discharges from the military with a diagnosis of schizophrenia from 1992 to 2005 were obtained from the US Army Physical Disability Agency, the Secretary of the Navy Council of Review Boards, and the Air Force Personnel Center/US Air Force Physical Disability Division. The rest of the data were obtained using the Defense Medical Surveillance System (DMSS). The diagnostic process leading to medical discharge from military service and

Statistical analyses

The samples from each case–control group were analyzed on the same microplate making within group comparisons more valid. In order to control any systematic bias between plates, all agent measurements were normalized. Robust random effect normalization considering both standard errors between and within plates was performed. To minimize the differences in the ranges of antibody levels to the examined agents, their values were transformed into normal Z scores. The standard deviation was chosen

Results

We tested a total of 6106 serum samples from 855 cases with schizophrenia and 1165 controls. As shown in (Table 1), the majority of cases were men, white, younger than 25 and had less than 3 years of service. Fewer than 3% of the cases had only one serum specimen available, about 30% had two or three, and approximately 40% had four or more specimens. About 90% of the pre-onset specimens were collected within 4 years before diagnosis and 96% of the post-onset specimens were collected within 2 years

Discussion

In this study, we found that schizophrenia cases had significantly lower CMV IgG antibody levels compared to matched controls. A combination of genetic and environmental factors resulting in some immune dysregulation in general and the lack of the CMV IgG antibodies in particular could be the most plausible speculations. While there are no specific research results supporting this finding in existing literature, the activation of the immune system and prolonged neuroinflammation are well

Role of funding source

This work was supported by the Stanley Medical Research Institute, Bethesda, Maryland (Research Grant # 03-NV-005) and the Department of the Army.

Contributors

Drs. Niebuhr, Cowan, Yolken, Li and Weber designed the study and wrote the protocol. Drs. Yolken, Weber, Mr. Fisher and Ms. Larsen managed the literature searches. Drs. Weber, Yolken, Niebuhr, Li, Cowan, Mr. Fisher and Ms. Larsen developed introduction and discussion. Dr. Li undertook the statistical analysis and wrote statistical analysis and result sections of the manuscript. All authors contributed to and have approved the final manuscript.

Conflicts of interest

The views expressed are those of the authors and should not be construed to represent the positions of the Department of the Army or Department of Defense. No authors have any actual or potential conflict of interest including any financial, personal or other relationships with other people or organizations within three (3) years of beginning the work submitted that could inappropriately influence, or be perceived to influence, their work.

Dr. Yolken is a member of the Stanley Medical Research

Acknowledgments

The authors would like to thank Walter Reed Army Institute of Research, Preventive Medicine Program staff member Ms. Janice K. Gary, AAS, for her work in careful review and administrative support in preparation of the manuscript and Dr. Kevin R. Porter, Director of the Infectious Diseases Directorate at the Naval Medical Research Center for his expertise. The authors also recognize the contribution of the Armed Forces Health Surveillance Center personnel, particularly Dr. Angie Eick, for

References (68)

  • D.W. Niebuhr et al.

    Association between bovine casein antibody and new onset schizophrenia among US military personnel

    Schizophr. Res.

    (2011)
  • P.H. Patterson

    Maternal infection: window on neuroimmune interactions in fetal brain development and mental illness

    Curr. Opin. Neurobiol.

    (2002)
  • S. Potvin et al.

    Inflammatory cytokine alterations in schizophrenia: a systematic quantitative review

    Biol. Psychiatry

    (2008)
  • A. Schattner et al.

    No evidence for autoimmunity in schizophrenia

    J. Autoimmun.

    (1996)
  • E.G. Severance et al.

    Subunit and whole molecule specificity of the anti-bovine casein immune response in recent onset psychosis and schizophrenia

    Schizophr. Res.

    (2010)
  • E.G. Severance et al.

    Gastrointestinal inflammation and associated immune activation in schizophrenia

    Schizophr. Res.

    (2012)
  • X.Q. Song et al.

    The interaction of nuclear factor-kappa B and cytokines is associated with schizophrenia

    Biol. Psychiatry

    (2009)
  • R. Tandon et al.

    Schizophrenia, “just the facts” what we know in 2008. 2. Epidemiology and etiology

    Schizophr. Res.

    (2008)
  • S. Theodoropoulou et al.

    Cytokine serum levels, autologous mixed lymphocyte reaction and surface marker analysis in never medicated and chronically medicated schizophrenic patients

    Schizophr. Res.

    (2001)
  • T. Wang et al.

    A potential association between Toxoplasma gondii infection and schizophrenia in mouse models. Experimental parasitology

    Exp. Parasitol.

    (2013)
  • K. Weigelt et al.

    TREM-1 and DAP12 expression in monocytes of patients with severe psychiatric disorders. EGR3, ATF3 and PU.1 as important transcription factors

    Brain Behav. Immun.

    (2011)
  • J. Xiao et al.

    Serological pattern consistent with infection with type I Toxoplasma gondii in mothers and risk of psychosis among adult offspring

    Microbes Infect.

    (2009)
  • R.H. Yolken et al.

    Serological evidence of exposure to herpes simplex virus type 1 is associated with cognitive deficits in the CATIE schizophrenia sample

    Schizophr. Res.

    (2011)
  • X.Y. Zhang et al.

    Elevated interleukin-2, interleukin-6 and interleukin-8 serum levels in neuroleptic-free schizophrenia: association with psychopathology

    Schizophr. Res.

    (2002)
  • W. Beumer et al.

    The immune theory of psychiatric diseases: a key role for activated microglia and circulating monocytes

    J. Leukoc. Biol.

    (2012)
  • S.L. Buka et al.

    Maternal infections and subsequent psychosis among offspring

    Arch. Gen. Psychiatry

    (2001)
  • C.J. Carter

    Schizophrenia susceptibility genes directly implicated in the life cycles of pathogens: cytomegalovirus, influenza, herpes simplex, rubella, and Toxoplasma gondii

    Schizophr. Bull.

    (2009)
  • S.J. Chen et al.

    Prevalence of autoimmune diseases in in-patients with schizophrenia: nationwide population-based study

    Br. J. Psychiatry

    (2012)
  • F.M. Coelho et al.

    Increased serum levels of inflammatory markers in chronic institutionalized patients with schizophrenia

    Neuroimmunomodulation

    (2008)
  • J. Doorduin et al.

    Neuroinflammation in schizophrenia-related psychosis: a PET study

    J. Nucl. Med.

    (2009)
  • R.C. Drexhage et al.

    The mononuclear phagocyte system and its cytokine inflammatory networks in schizophrenia and bipolar disorder

    Expert. Rev. Neurother.

    (2010)
  • R.C. Drexhage et al.

    An activated set point of T-cell and monocyte inflammatory networks in recent-onset schizophrenia patients involves both pro- and anti-inflammatory forces

    Int. J. Neuropsychopharmacol.

    (2011)
  • W.W. Eaton et al.

    Association of schizophrenia and autoimmune diseases: linkage of Danish national registers

    Am. J. Psychiatry

    (2006)
  • W.W. Eaton et al.

    Autoimmune diseases, bipolar disorder, and non-affective psychosis

    Bipolar Disord.

    (2010)
  • Cited by (17)

    • Toxoplasma gondii, Herpesviridae and long-term risk of transition to first-episode psychosis in an ultra high-risk sample

      2021, Schizophrenia Research
      Citation Excerpt :

      The three prospective studies that measured T. gondii IgG in young people and examined the association with later diagnosis of schizophrenia-spectrum disorders did find similar effects to those observed in the present study. Two case control studies were conducted using clinical records and routinely collected serological data of US military recruits: one study (n = 712) found a positive correlation for T. gondii IgG level and later schizophrenia diagnosis (HR 1.24; p < 0.01) (Niebuhr et al., 2008), while the other (n = 2020) did not find a significant association (HR 1.04; p = 0.27) (Li et al., 2013). In a Danish cohort of women whose serostatus was tested when they gave birth (n = 45,609) the relative risk of developing schizophrenia-spectrum psychosis later in life associated with seropositivity for T. gondii was 1.31 (p = 0.049, after adjusting for age and observation time) (Pedersen et al., 2011).

    • Targeting the microbiome-gut-brain axis for improving cognition in schizophrenia and major mood disorders: A narrative review

      2021, Progress in Neuro-Psychopharmacology and Biological Psychiatry
      Citation Excerpt :

      Patients with schizophrenia also show increased seropositivity to Toxoplasma Gondii, a replicated risk factor for schizophrenia that causes long-term microanatomical changes in the intestines (Torrey et al., 2012). Other gut-related autoimmune findings in schizophrenia include an increased prevalence of coeliac disease, which is associated with inflammation of the gut mucosa (Cascella et al., 2011; Li et al., 2013), and the association between bovine casein antibody and new-onset schizophrenia (Niebuhr et al., 2011). At the same time, some comorbid inflammatory conditions related to the gut microbiota have been associated with bipolar disorders (Gondalia et al., 2019).

    • Neutrophil/lymphocyte ratio: Can a non-specific marker of inflammation helps to confirm the inflammatory hypothesis of the serious mental diseases? A case-control study

      2019, Medical Hypotheses
      Citation Excerpt :

      As far as the infectious etiology is concerned, there are already antecedents that try to associate the Herpes simplex virus [15,16], the Epstein-Barr virus [17], the Toxoplasma gondii [18], the virus of the disease of Borna [19], Cytomegalovirus [20], Chlamydophila pneumoniae [21], Borrelia burgdorferi [22] or Candida albicans infections [23] with pathology of the mental sphere. With regard to allergic reactions, there have already been published studies in which food antigens [22–25] or mast cells [26] are considered factors to be taken into account for the presentation of psychiatric symptoms. Finally, among the autoimmune processes associated with psychosis, it is worth mentioning the positive association established between autoimmune diseases and schizophrenia [27]; some studies find a significantly higher prevalence of positive titers for twenty different autoantibodies in patients with schizophrenia compared to controls, highlighting the anti-cardiolipin IgG and IgM, the nerve growth factor, the ANA and the anti-NMDA [28].

    • Monocyte activation detected prior to a diagnosis of schizophrenia in the US Military New Onset Psychosis Project (MNOPP)

      2018, Schizophrenia Research
      Citation Excerpt :

      A low-level systemic inflammation is a hypothesized harbinger of schizophrenia and other psychoses, as dysregulated inflammatory markers are often found in early vs. later stages of these disorders (Bechter, 2013; Li et al., 2013; Niebuhr et al., 2011; Niebuhr et al., 2008; Severance et al., 2012; Suvisaari and Mantere, 2013).

    • Combined Toxoplasma gondii seropositivity and high blood kynurenine - Linked with nonfatal suicidal self-directed violence in patients with schizophrenia

      2016, Journal of Psychiatric Research
      Citation Excerpt :

      Thus, T. gondii seropositive subjects with KYN in the upper 25th percentile, at higher risk for NF-SSDV, may have chronic gastrointestinal inflammation that contributes to NF-SSDV risk via mechanisms involving the gut microbiota and increased gut permeability. Altogether, these findings are consistent with current models relating T. gondii infection, possibly acting in concert with other antigens (Li et al., 2013), to gut microbial dysbiosis and schizophrenia (Severance et al., 2014; Severance et al., 2015). Indeed, the higher trait impulsivity that has been described in subjects with schizophrenia, particularly younger men (Cook et al., 2015), may increase exposure to other risk antigens, either through increased sexual risk behavior (Pack and Crosby, 2001) (e.g. vaccinia, cytomegalovirus [CMV; Li et al., 2013]), or altered diet (Jasinska et al., 2012).

    • Fungal Forces in Mental Health: Microbial Meddlers or Function Fixers?

      2023, Current Topics in Behavioral Neurosciences
    View all citing articles on Scopus
    View full text