Keywords

1 Introduction

“I was convinced that we were on the verge of a world war. The public defense sirens were tested every first Monday of the month. To me, those sirens meant that the cruise missiles had already been launched. We had just minutes to do what had to be done: find each other and wait to die. Sometimes I could actually feel the radiation taking effect. I felt nauseous and believed that my hair was falling out. The rest of the world pretended to carry on as normal, but I could see that everyone was afraid. They knew that we were all about to suffer a slow, painful death, but nobody knew how to prepare for it. […] The strange thing is that time did not exist for me in that situation. […] It is a kind of vacuum, no-man’s-land. Not in real life, but also not dead. […] Now I no longer see my psychoses as isolated psychopathology. […] My psychoses are my way of reacting to my life history. They are my response to the unpredictable abuse I had to face as a child. I hit my father back when I finally was angry enough, after years of submission. My father left the house after threatening to kill himself, after which the whole family turned against me. […] In the years to follow I lost all my strength and exchanged it for guilt, fear, and incomprehensible psychotic experiences. I became the problem that had to be solved. I don’t think that abuse itself is a strong cause for psychosis. […] I think that the threat and the betrayal that come with it feed psychosis. The betrayal of the family that says, “you must have asked for it, ” instead of standing up for you. That excuses the offender and accuses the victim. And forces the child to accept the reality of the adults. That forces the child to say that the air is green, while she sees clearly it is not green but blue. That is a distortion of reality that is very hard to deal with when you’re a child. You are forced to betray yourself [1].”

This piece is a first person account of a woman with psychosis and a history of childhood trauma. Psychosis is a mental health disturbance that can cause people to perceive and interpret the world differently from those around them. Symptoms can fall into one or more of two categories: positive and negative. Positive symptoms include hallucinations (i.e., seeing, hearing, and feeling things that are not there), delusions (i.e., false beliefs), and disorganized speech and/or behavior. Negative symptoms relate to functions that are diminished or absent in patients with psychosis, such as loss of motivation, flat affect, and apathy. If more than two symptoms are present during at least 6 months and cause functional dysfunction, the psychotic disorder is referred to as schizophrenia (Fig. 10.1).

Fig. 10.1
figure 1

Positive and negative symptoms of schizophrenia

Full size image

Evidence accrued in recent years that psychotic experiences are common in the general population with a lifetime prevalence of 7.2% [2]. Psychotic experiences can be fleeting hallucinations, suspiciousness, paranoia, and magical thinking. These are associated with a greater risk for later development of psychotic disorder (or other disorders) and have been reported in young children and adolescents, too [3]. It is, then, important to look into early childhood experiences that may be on the etiological pathway to psychosis. A possible causal role of childhood trauma has been proposed decades ago, not just because a significant proportion of people with psychosis reports a history of adversity [4]. Childhood trauma refers to a broad range of adverse experiences usually before the age of 16, including life events (e.g., death of a parent, environmental catastrophes) or single and/or repeated maltreatment (e.g., sexual abuse, physical abuse, physical neglect, emotional abuse, emotional neglect, peer bullying, i.e., maltreatment with the intention to harm). In this chapter, we will focus on the latter and discuss the different levels through which childhood trauma (which in the literature is sometimes also referred to as “adversity” or “maltreatment”) might exert its influence on the development of psychosis.

2 The Epidemiological Evidence for a Link Between Childhood Trauma and Psychosis

Evidence is amassing that consistently points in the same direction: childhood adversities, specifically with the intention to harm, are associated with an increased risk for psychosis. In a comprehensive meta-analysis, Varese and colleagues identified 36 studies with a combined sample size of 80,000, and concluded that childhood adversities were associated with a two- to fourfold increased risk for later psychosis [5]. The overall association was 2.78 increased odds, irrespective of study design. More specifically, zooming in on different forms of adversity, the odds rations reported were 2.4 for bullying, 3.4 for emotional abuse, 2.9 for neglect, 3.0 for physical abuse, and 2.4 for sexual abuse. Only parental death was non-significantly associated with psychosis (odds 1.7) [5]. There is, furthermore, some evidence for a dose–response relationship [6,7,8,9,10]. Research suggests that for each additional indicator of maltreatment, there would be a modest linear increase in risk for psychosis [11].

Since the review by Varese and colleagues, additional evidence has been published suggesting that a history of childhood trauma is not just common among those with a psychotic disorder, but also among those on the psychosis spectrum, including children and adolescents [7, 12,13,14,15,16,17,18,19,20,21,22,23,24,25,26]. In a prospective study using a large twin cohort, researchers assessed childhood trauma and psychotic experiences at age 5, 7, 10, and 12 years. They found that maltreatment by an adult, as well as peer victimization with intention to harm were associated with a 3.3 increased risk for psychotic experiences [27]. This is further underlined in a study conducted by Kelleher and colleagues who found that after cessation of trauma in adolescence, psychotic experiences may also stop [7]. Thus, when studying the association between childhood trauma and psychosis, it may be important to move away from categorical diagnoses and consider the whole psychosis spectrum.

3 The Question of Causality

It is generally assumed that childhood trauma is a causal risk factor in bringing on later psychosis. However, before such a claim can be substantiated, a number of issues need to be addressed. These include, among others, the question of robustness, temporality and genetic confounding when studies are summarized in a review [7, 28]. As evident from the meta-analysis by Varese and colleagues [5], there seems to be a well-established robust and strong association between childhood trauma and psychosis across different samples along the psychosis continuum. The same meta-analysis also summarized findings that point toward a dose–response relationship [5]. However, these studies were unable to determine the directionality of the association between childhood trauma and psychosis. While the assumption may be that childhood trauma increases the risk for psychosis, other explanations may also need to be considered [28, 29] (Fig. 10.2).

Fig. 10.2
figure 2

Etiological models explaining the association between childhood trauma (CT) and psychosis [28]

Full size image

A possible alternative could be the hypothesis of “reverse causation.” This would be the case if early symptoms of psychosis were to lead to an increased risk of exposure to childhood adversity. Psychotic symptoms would then have to be caused by other factors, and experiences of childhood adversity would only follow later in a person’s development [28]. To test this hypothesis, temporality needs to be determined, but obviously, this is very difficult in the context of childhood trauma since prospective studies are ethically difficult. Therefore, only a few studies to date have been able to successfully adopt this approach, and even fewer have assessed whether psychotic experiences were present before the exposure to childhood traumatic events [28]. In one study, researchers found that individuals who were sexually abused before the age of 16 had a twofold increased risk for developing any diagnosis of psychotic disorder, and a 2.6-fold increased risk for schizophrenia, compared to age- and gender-matched non-abused individuals [30]. Similarly, another prospective study found that those maltreated before the age of 12 had a 3.16-fold increased odds to report psychotic symptoms later in life [27]. Neither one of these studies, however, was designed to establish temporality of onset of psychotic symptoms. Recently, two studies have been published that prospectively examined both exposure to childhood trauma and psychotic symptoms. The findings exemplify that, whereas childhood traumatic experiences increased the odds for later psychotic symptoms, psychotic experiences also predicted a greater risk for exposure to traumatic events [7, 31]. The observation of a bidirectional relation calls to consider another, though not necessarily mutually exclusive, explanation.

Before causality, or any relationship can be established, the possibility that both childhood trauma and psychotic symptoms are caused by a third factor needs to be ruled out. It is conceivable that the exposure to early life adversity is not random, but may develop out of a preexisting vulnerability, such as genetic vulnerability. The hypothesis of a “gene-environment correlation” postulates that a genetic predisposition would lead to both an increased likelihood of exposure to traumatic events, as well as to the later expression of psychotic symptoms. In this case, individuals with an increased genetic risk for psychosis would be more likely to be exposed to adversity due to inborn traits associated with psychosis, such as impaired social functioning or cognitive abilities [32,33,34]. In this case, the association between childhood trauma and psychosis may seem statistically plausible, when in fact this association could be explained by underlying genetic vulnerability. In other words, the relation between childhood trauma and psychosis would be “genetically confounded” [19, 28, 29].

There are different ways of controlling for “genetic confounding.” An extensively applied approach is to use family psychiatric history as a proxy for genetic risk. Studies that adopted this approach did not find evidence that family liability confounded the association between childhood trauma and psychosis. Neither did they find evidence that combined family liability and childhood trauma increased the odds for later psychosis beyond the effect of each individually [8, 35], (for a review see [36]). Another method is to use direct measures of genetic vulnerability. A number of studies looked at specific candidate genes (i.e., genes that are thought to be either implicated in psychosis itself, or in possible mechanisms underlying psychosis onset) [37,38,39,40]. Again, others computed a polygenic risk score for schizophrenia. A polygenic risk score is a way to combine genome-wide genetic liability into one variable by summing up and weighting all risk alleles carried by one individual [41]. This approach could not confirm (passive) gene–environment correlation either [36]. However, neither of these concepts is currently able to capture all relevant genetic risks [19, 28, 29]. To overcome this limitation, the authors of three recently published studies used a monozygotic twin-differences approach. The idea behind this approach is that if two genetically identical (monozygotic) twins differ in environmental exposure, and that exposure is associated with psychotic symptoms, the association cannot be attributed to underlying genetic risk alone as monozygotic twins share 100% of their genes [19, 29, 31]. In agreement with the other approaches, these studies found that increased levels of exposure to childhood trauma were associated with more symptoms of psychosis, thus showing that the association cannot be solely attributed to genetic confounding.

A last consideration concerns the fact that not all individuals with a history of childhood trauma continue to experience psychotic symptoms. At the same time, whereas a great number of those with a psychotic disorder report having been victimized, not all persons with psychosis have. It is therefore crucial to consider that childhood trauma is neither sufficient nor necessary for the onset of psychosis [8, 28, 42]. Nevertheless, while other genetic and nongenetic factors might play a role in the etiology of psychosis, the current literature suggests that at least part of the association between childhood traumatic experiences and psychosis is genuine, and in agreement with a possible causal explanation [29].

4 Differential Sensitivity to Trauma: The Role of Genetic and Epigenetic Variation

While multiple studies have shown that the association between childhood trauma and psychosis persists even after controlling for genetic risk, it is currently unclear how genetic vulnerability plays a role in one’s individual risk to develop psychosis following exposure, and which genes are likely to be involved [28, 43, 44]. The current literature, however limited, points to genes that are not specific to psychosis, but associated with mechanisms such as neuroplasticity and stress regulation. Again, the genetic contribution to the association between childhood trauma and psychosis can be either directly or indirectly examined.

Studies that adopted an indirect approach used family history or a twin-design and focused on heritable traits that are thought to be associated with psychosis. Based on data from a twin sample from the general population, the authors reported that worse cognitive speed, as well as a lifetime diagnosis of depression, moderated the effects of childhood trauma on the development of psychosis [45]. However, other studies, using family psychiatric history by either looking at parental psychiatric history, or at co-twin’s symptoms, found no evidence of an interaction between indirect genetic liability and childhood trauma on the emergence of psychotic symptoms [27, 31, 35].

A number of studies have used a direct approach by looking at genetic variation to examine the interaction between childhood trauma and psychosis. Most of these studies examined candidate genes that are involved in regulating neurotransmitters (e.g., serotonin transporter gene SLC6A4/5-HTT/, specifically a functional polymorphism in the promoter region 5-HTTLPR) [46], neuroplasticity and cell survival following stress (brain-derived neurotrophic factor; BDNF) [40, 47], and the stress response system (FKBP5) [37]. For example, there is some evidence that a short- (s-) allele polymorphism in the promoter region of 5-HTTLPR is associated with several psychiatric disorders, possibly through its effects on the stress response system [48, 49]. One study found an interaction between 5-HTTLPR variations and physical neglect and abuse on cognitive functioning in patients with psychosis [46], in line with previous work suggesting that an increased stress response may be associated with cognitive impairments [50]. Another study looked at the moderating effects of the BDNF-Val66Met polymorphism between childhood trauma on positive and negative symptoms of psychosis [51]. Valine (Val) and methionine (Met) are amino acids that are used in the biosynthesis of proteins. The substitution of valine to methionine, by changes in a single base pair of the BDNF gene, may lead to differences in the activity of BDNF. The val/met polymorphism has been linked to changes in intracellular trafficking (i.e., distribution and release of macromolecules throughout and outside of a cell), as well as changes in activity-dependent secretion of BDNF [52]. Met carriers with a history of abuse reported more positive symptoms of psychosis compared to those with a Val/Val genotype [51]. However, others failed to replicate this interaction [40, 47]. A different gene of interest is FKBP5, which is thought to be involved in modulating the feedback loop determining glucocorticoid receptor sensitivity. Collip and colleagues [37] found an interaction between two FKBP5 single nucleotide polymorphisms and childhood trauma on psychotic symptoms and increased stress sensitivity in a twin sample from the general population. In agreement with these findings, a study of two independent samples from the general population found this polymorphism to be related to increased early stress sensitivity and the expression of positive psychotic symptoms [53]. Another replication study found an interaction between childhood trauma and the same polymorphism on schizotypy, psychotic experiences, as well as symptoms of anxiety and depression, in the general population [47]. Green and colleagues [38], albeit looking at a different polymorphism, found a variation in combination with maltreatment to affect cognition, in those with schizophrenia and healthy controls. For a more extensive overview of relevant gene–environment interaction studies, we also refer to the chapter by Araceli Rosa and colleagues in this handbook.

While the mentioned studies have looked at variations in DNA sequence, it is also possible that childhood trauma exerts its influence through epigenetics [44, 54]. Epigenetics refers to changes in gene expression rather than modifications of the genetic code itself [54]. While this process is not yet fully understood, vigorous scientific attention has been directed to DNA methylation as a mechanism through which early life stress may become “embedded” in the genome [55]. An animal study found that the level of postnatal maternal care in rodents was associated with alterations in the methylation of the glucocorticoid NR3C1 receptor [56]. A human postmortem study [57] reported increased methylation of the NR3C1 promoter and changes in mRNA (i.e., messenger RNA that communicates information from DNA to the ribosome about what amino acid sequence should be read, thereby specifying gene expression) in those with a childhood trauma history, compared to suicide victims and healthy controls. However, a second study failed to replicate these findings [56]. A number of in vivo studies have reported altered DNA methylation as a consequence of childhood trauma (for a review see [42]). To our knowledge, there is only one study that examined methylation in first-episode psychosis patients with a history of childhood trauma within a considerable sample (nfirst-episode-psychosis = 48; ncontrol = 48) [58]. Researchers found lower DNA methylation of LINE-1 (class I transposable elements in the DNA) sequences in those patients. However, in a recently published article on the E-Risk dataset (n = 1.658) (Environmental Risk longitudinal study), authors found very limited evidence for an association between childhood adversity and changes in epigenetic variation (in peripheral blood) [59]. They also found that methylation associations for adversities overlapped with those of other health behaviors such as tobacco smoking. Another study found similar results in that DNA methylation in whole blood was associated with tobacco smoking, and also BMI and additive genetic effects [60]. This highlights the methodological difficulties of disentangling biological effects of environmental influences from other health-related behaviors associated with a trauma history [59].

The evidence for genetic moderation of the association between childhood trauma and psychosis, thus, is mixed. Nevertheless, these studies exemplify the numerous genetic mechanisms through which childhood trauma could potentially influence the development of psychosis, and other related mental health disorders [54, 59, 61].

5 Biological Mechanisms Underlying the Link Between Trauma and Psychosis

The diathesis-stress model of mental illness postulates that major life stressors, such as childhood trauma, can be a serious neurodevelopmental insult on the developing brain and contribute to the onset of psychiatric disorders [62]. The idea that abnormalities in the limbic system could be related to the effects of childhood abuse on brain development dates back to van der Kolk and Greenberg in 1987 [63]. Since then, a number of studies have found structural, functional, and molecular changes in brain structures and persistent associations with childhood trauma, e.g., [63,64,66]. Below, we will discuss implicated alterations of the brain (more specifically hippocampus, amygdala, and findings from studies on cortical thickness), the stress response system and the immune system. Intact stress response and immune systems are of crucial importance for mental and physical health, behavioral adaptation, and brain development; the dysregulation of these systems may therefore have long-lasting effects on the developing brain and one’s mental and physical health [67].

5.1 The Brain

5.1.1 Hippocampus

The hippocampus may be the most likely place to reflect the effects of childhood trauma [68]. This brain area is densely packed with glucocorticoid receptors, which makes it highly susceptible to early life stress [69]. Structural imaging studies found reduced bilateral hippocampal volume in those with a history of childhood trauma, those with a psychotic disorder, and in individuals with both [67, 68, 70, 71]. Interestingly, volume reductions were also found in nonpsychotic siblings, individuals at risk for schizophrenia, as well as in people with psychotic bipolar disorder (for a review see [72]). Studies found that childhood trauma was associated with reduced hippocampal volume in patients with first-episode psychosis, which may be associated with worse cognitive performance compared to patients with a diagnosis of schizophrenia, but no history of adversity [39, 73, 74]. There is some evidence that the hippocampus is involved in regulating the hypothalamic-pituitary-adrenal (HPA) axis stress response and that smaller hippocampal volume in those with a first-psychotic episode may partially be explained by stress-related processes [39]. In animal models, decreased hippocampal volumes, specifically in dentate gyrus (DG) and cornu ammonis 3 and 1 (CA3;CA1), were found in mice exposed to maternal deprivation (i.e., traumatic event) [75]. This fits well with previous findings from human postmortem studies that suggest that the DG and CA3 subfields appear most affected by molecular and cellular changes when comparing tissue of previously healthy individuals with persons with schizophrenia [76]. Additionally, on a cellular level, it was shown that the hippocampus contains “place cells” and “grid cells” in the entorhinal cortex. These cells are responsible for the spatiotemporal representation of places, routes, and associated experiences [77]. As spatiotemporal disorientation may be a symptom of psychosis in some patients, it may not be farfetched that childhood trauma could exert its effects on the development of psychotic symptoms by affecting hippocampal development [78].

5.1.2 Amygdala

The amygdala is a key limbic structure for prioritizing and encoding emotionally salient information, and detecting and responding to facial expressions and potential threats [79, 80]. Given the altered emotional responses observed in patients with schizophrenia, an affective pathway to psychosis has been previously proposed [80,81,83]. This also makes sense based on the finding that, like the hippocampus, the amygdala is rich of glucocorticoid receptors, which would make it especially vulnerable to early stress, including childhood trauma. Most studies have examined emotional reactivity following childhood trauma on a behavioral level, for example, by examining reactivity to daily life stress [82, 83]. Other studies have examined structural and functional abnormalities of the amygdala in a variety of mental disorders, including psychosis [83,84,86]. Structural studies report contradicting results. Some found decreased amygdala volumes in those with a first-episode psychosis and a trauma history. Others found increased volumes and argue that this may be due to the stimulating effects of stress on pyramidal cells in this brain structure [87, 88]. The same discrepant findings were found on a functional level, in that some show either patterns of hypo- or hyper-activation during emotional tasks in patients with psychosis [89]. Therefore, Cancel and colleagues [90] suggested to examine functional connectivity between the amygdala and other related brain areas. They found that in patients with schizophrenia and a history of childhood trauma—specifically sexual abuse and physical neglect—connectivity between amygdala and posterior cingulate/precuneus was decreased [90]. These results are similar to previous studies that used a face recognition task and found decreased connectivity in response to fearful faces [91].

5.1.3 Findings from Studies on Cortical Thickness

There is some evidence that gray matter density is correlated with childhood trauma in patients with psychosis. Sheffield and colleagues [92] found that overall gray matter differed between psychotic patients with and without a history of trauma, and healthy participants. More specifically, they found that individuals with a history of sexual abuse and a diagnosis of psychosis differed from healthy controls in medial and inferior frontal, inferior and superior temporal, precentral gyri and inferior parietal lobe grey matter density. Yet, psychosis patients without sexual abuse only differed from healthy controls in cerebellar gray matter volume. Furthermore, within the psychosis patients group, those with and without a history of sexual abuse differed in gray matter volume in bilateral anterior cingulate cortices and left inferior frontal gyrus only [92]. In another study, authors reported anterior cingulate volumetric differences (compared to healthy controls) in patients with schizophrenia and antisocial personality disorder after controlling for childhood abuse [93]. Moreover, at least three studies demonstrated that prefrontal cortical deficits can be observed in victims of childhood trauma without mental health problems, too [93,94,96]. This fits well with cognitive models that propose that childhood trauma is associated with cognitive impairments in those with a psychotic disorder, suggesting that neurocognitive alterations mediate the development of psychosis following trauma exposure [39, 73, 97]. However, in a large prospective cohort study using the Dunedin dataset (Dunedin Multidisciplinary Health and Development Study, a longitudinal study that started in 1972), Danese and colleagues found that whereas those exposed to victimization had indeed impaired cognitive functions (including general intelligence, executive function, processing speed, memory, perceptual reasoning, and verbal comprehension), these deficits were explained by preexisting cognitive deficits prior to the victimization event(s) [98].

These studies underscore the effect of childhood trauma on brain development and also raise caution when trying to unravel the neurobiological correlates of psychopathology, as some may be specific to the disorder itself, and others may represent alterations associated with exposure to childhood trauma, or may even predate the exposure to trauma [68, 92].

5.2 The Stress Response System

One of the most commonly proposed idea is that childhood trauma increases the probability for later psychosis through its impact on the stress response system, which is in growing children, still under development. Early traumatic experiences may cause heightened sensitivity of the HPA-axis to subsequent stress, which eventually could increase the risk for psychosis (and related mental disorders) [99, 100]. Similar results were reported for the association between childhood trauma and subclinical psychotic experiences in a general population sample, through a pathway of heightened subjective stress appraisal [101].

At the behavioral level, studies found that individuals with a history of trauma reacted more strongly, with more negative affect, and more paranoia, to minor daily life stressors [82, 102, 103]. In an experimental study, Valmaggia and colleagues used virtual reality and reported stronger paranoid ideation to a neutral social environment in people at ultra-high risk for psychosis, and a history of bullying [104]. In agreement with these results, another virtual reality study indicated heightened social stress reactivity as a possible link between childhood trauma, psychosis liability, and paranoid ideation. Those with a trauma history were found to report more paranoid ideation and subjective distress, compared to individuals without such experiences [105].

In addition, there is a growing body of literature that links exposure to childhood adversities, and psychosis, to hyper-activation and sensitization of the HPA axis [67, 71, 100]. HPA-axis activity is often measured by examining salivary cortisol levels. There is some evidence for both elevated baseline cortisol secretion and blunted cortisol awakening response in patients with psychosis, including those with childhood trauma [105,106,107,109]. Others found that the pituitary gland was enlarged in those with a psychotic disorder [110, 111]. This fits in well with the “traumagenic neurodevelopmental model” which assumes that HPA-axis dysregulation may mediate the relationship between childhood trauma and the development of psychosis [97, 112]. However, there is also some evidence that stressful events reduce HPA-axis activation [113, 114] and that pituitary gland volume in children at risk for psychosis due to adversity is reduced [115]. Irrespective of these discrepancies, there seems to be a correlation between elevated dopaminergic brain response associated with psychosis and salivary cortisol, although the direction is still to be determined [108, 116]. It could be that cortisol exerts its influence on the expression of symptoms through its effects on dopaminergic pathways (“bottom-up” model). It is also plausible that underlying symptoms and corresponding neurotransmitter activity affect HPA-axis activity (“top-down” model) [117].

Childhood trauma may increase the risk for psychosis by affecting the mesolimbic dopamine system through a mechanism of exaggerated dopamine release to subsequent social stressors later in life [99, 118]. There may be a feedback loop at play: prolonged stress, including adversity, may increase glucocorticoid release, which in turn may increase dopamine secretion. This may start a positive feedback loop through which the increased dopaminergic activity elevates HPA activity and glucocorticoid release. Since it has previously been shown that dopaminergic hyperreactivity plays a role in psychosis, this may be a mechanism through which childhood trauma increases the risk for the subsequent development of psychotic symptoms [67]. Yet, most evidence that dopamine release is elevated following exposure to stress still comes mainly from animal models [119], but studies are beginning to emerge which provide some evidence in humans as well [97, 107, 117].

5.3 Immune System

There is accumulating evidence that alterations in the innate immune system are associated with psychotic disorders, and that childhood traumatic events are associated with a pro-inflammatory state in adulthood. These alterations include increased chemokines (e.g., CCL-11), acute-phase proteins (e.g., C-reactive protein [CRP]), and cytokines [120]. Some cytokines are pro-inflammatory, such as interleukin IL-6, IL-8, and tumor necrosis factor (TNF-α). Consequently, there are cytokines with a dampening effect as to keep a right balance [121]. A recent meta-analysis reported an association between childhood trauma and inflammatory markers. The biggest effect was found for TNF-α, followed by IL-6 and C-reactive protein [122]. Interestingly, Dennison and colleagues found that increased levels of pro-inflammatory markers were found only in patients with schizophrenia who also had a history of childhood trauma, as compared to patients without [123]. In another study, elevated CRP levels were shown in first-episode psychosis patients with a trauma history, but not in patients without a trauma, or healthy controls [124]. Yet, another study found the same group differences for elevated TNF-α levels [125]. These results have led to speculations that early childhood adversities “bring about epigenetic changes that lead to a pro-inflammatory phenotype in adulthood,” which has been associated with a range of mental and physical health issues, including psychosis [123]. While this sounds like a plausible interpretation, there is still little direct evidence, to date, supporting this hypothesis.

Interestingly, the association between HPA-axis (dys-)functioning and childhood trauma on the one hand, and the association between childhood trauma and the immune system on the other, has led to the exploration of interactive pathways [126]. It was found that receptors for one or more of the stress hormones are expressed on lymphocytes. Moreover, lymphocytes are also able to synthesize the adrenocorticotropic hormone (ACTH), which is secreted by the pituitary gland and is important in the regulation of the stress response system [127]. While much is still unknown, it may indeed be the case that early life stress is a neurodevelopmental insult on multiple, interacting systems [42, 68].

6 Psychological Mechanisms Linking Trauma and Psychosis

In addition to alterations at the biological level, it is likely that the exposure to severe traumatic events in childhood impact on one’s psychological development, and that psychological factors may co-determine a person’s vulnerability, or resilience, to these events. A number of possible psychological mechanisms linking exposure to psychopathological outcomes have been proposed, including dysfunctional cognitive schemas, affective dysregulation, insecure attachment styles, and dissociative mechanisms.

Cognitive models propose that negative beliefs about the self and others, and increased threat anticipation may mediate the association between childhood trauma and psychosis [127,128,130]. The social defeat theory postulates that it is not the experience of adversity itself, but the enduring feeling of defeat and the subordinate position experienced during adverse events that increase the risk for psychosis. Such feelings may stem from childhood trauma [130,131,133]. Animal studies report that long-term isolation can lead to reductions in whole-brain volume, hippocampus, or medial prefrontal cortex, and that social defeat can reduce neurogenesis [131]. However, Selten and colleagues note that to date, there are no studies that have examined possible brain changes in humans at psychosis onset in relation to proxy measures of social defeat (for a review see [131]). Preliminary support for the social defeat theory was found by a number of behavioral studies that showed that perceptions of defeat and entrapment were associated with positive symptoms in a clinical group of patients with schizophrenia [133,134,136]. Results from the NEMESIS-2 study (Netherlands Mental Health Survey and Incidence Study) indicate that self-reported feelings of social defeat may act as a mediator between childhood trauma and the expression of psychosis. This was found for individuals with a psychotic disorder, and also for psychotic experiences in the general population [137]. A virtual reality study found that those at ultra-high risk for psychosis had higher levels of social defeat and entrapment, and were more likely to react with paranoid appraisals to the virtual reality environment than their healthy counterparts [138]. However, there is some debate on the conceptualization of social defeat. The theory is built on observations in animals, in which animals are attacked by others and may actually die from this attack, even if they show subordinate behavior. This type of “subordination” is likely to be different compared to the subordinate position caused by social exclusion that humans can experience. Other concepts used in the context of social defeat studies in humans are discrimination, negative social evaluation, social adversity, social fragmentation, or social disadvantage [131], which is why the term “social defeat” in relation to human studies has previously been indicated as a “misnomer” (i.e., an inaccurate name or designation) [8]. Karlsen and colleagues looked at the association between (racial) discrimination and psychosis and found it to be strongest for discrimination involving physical assault. It has been proposed that not defeat itself, but hostility, threat, and violence in the context of social disadvantage and discrimination might explain the high rates of psychosis in migrants and minority groups [139].

Some researchers have also pointed to the possible relevance of attachment. Attachment styles are thought to reflect early cognitive-affective representations of the self and others, as well as strategies for regulating distress [140]. Children are dependent on their caregivers and adults around them to keep them safe. In his attachment theory, Bowlby [141] proposed that infants internalize experiences of interaction with their caregivers, and that this representation is carried forward into adulthood. Early trauma can affect a child’s attachment, which can consequently influence expectations and believes about the self and others in interpersonal interactions later in life [141, 142]. There is some evidence that an insecure attachment style (see Fig. 10.3) might act as mediator between childhood trauma and psychotic experiences in clinical samples and the general population [142,143,145]. On the other hand, an insecure avoidant attachment style mediated the association between neglect and paranoia in data from the US National Comorbidity Survey [146]. Overall, it looks like that an anxious and avoidant attachment style might mediate the relationship between childhood trauma and positive symptoms of psychosis [143,144,145,146,147,149]. Interestingly, this association was not found for negative symptoms [42].

Fig. 10.3
figure 3

Attachment styles

Full size image

Finally, it may be that a dissociative response to childhood trauma paves the way to psychosis. Dissociation can be defined as “a disruption in the usually integrated functions of consciousness, memory, identity or perception of the environment” [112, 150]. Research showed that individuals who apply a dissociative coping style following trauma are more likely to have impaired reality testing and may develop subsequent psychotic experiences [151]. In a clinical sample of psychotic patients, it was shown that those with a positive history of trauma had increased dissociative tendencies compared to patients without a trauma history [152, 153]. Moreover, it was found that the risk for psychosis through a history of physical neglect could be explained by increased dissociation in patients with psychosis [154]. Lastly, both in nonclinical and clinical groups, the association between childhood trauma and hallucination proneness was mediated by dissociative tendencies [23, 153].

7 Specificity of the Association Between Types of Trauma and Distinct Psychotic Symptoms

When moving away from broad diagnostic categories to a symptom-specific level, there is a discussion in the literature whether there are associations between specific forms of childhood trauma and specific symptoms of psychosis (e.g., 154). It has, for example, been proposed that different forms of adversities may exert differential influences upon affective and cognitive processes [6, 23, 155]. However, empirical findings have provided mixed support for this hypothesis. A number of studies found an association between sexual abuse and hallucinations and delusions in that the content may be related to patients’ past traumatic experiences [10, 16]. Other studies found emotional abuse and neglect to be most strongly associated with dissociative symptoms [152, 155,156,158]. Another study, however, found sexual abuse to be most strongly associated with dissociative symptoms [159], while other studies found no evidence for specificity at all [160, 161]. Previous work had some statistical limitations in that they merely relied on the presence or absence of a statistically significant associations between a given form of adversity and a specific symptom of psychosis, but did not formally test whether this association was significantly stronger than the association with another form of adversity, or another specific symptom. Van Nierop and colleagues formally tested specific associations and showed that no form of trauma had a statistically stronger association than the other with specific symptoms [162]. Only experiences with an intention to harm were more strongly associated with psychosis compared to experiences without such component. Authors concluded that, besides the presence of “intention to harm,” no other “specificity” exists [162]. In agreement with this finding, a study that looked at children’s risk of reporting psychotic symptoms found that the outcome was similar when the perpetrator was an adult or a peer, as long as there was an intention to harm (in agreement with the social defeat hypothesis) [132]. This suggests that the only specific predictor is an element of threat intentionally induced by others [162].

Moreover, it was found that about 45% of the general population’s attributable risk for childhood onset psychiatric disorders is accounted for by early maltreatment [163]. As convincingly outlined in this handbook, individuals with a history of early adverse experiences do not only have a higher risk for psychosis, but also show a higher prevalence of a number of disorders, including depression, anxiety, substance abuse, eating disorders, suicidal symptomatology, personality disorder, dissociative disorders, and posttraumatic stress disorder [4, 163,164,165,166,168]. Thus, childhood trauma seems to be characterized by an admixture of affective, anxious, and psychotic symptoms [17, 169, 170]. Investigating symptoms across these different, but related disorders (i.e., internalizing, externalizing, thought disorders) led to suggest that maltreatment was associated with greater “general psychopathology” (i.e., p-factor), but not with any specific endophenotype (i.e., measurable biomarkers that are correlated with a disorder, partly because of a shared genetic disposition) [171].

8 Treatment Considerations for Patients with a Trauma History

A history of childhood trauma is associated with a greater risk for psychiatric disorders, more comorbidity, worse social functioning, lower remission rates, and less favorable treatment outcomes in general [168, 169, 171,172,173,175]. There is some evidence, albeit limited, that greater (supportive) networks act as protective factors, but that social skills in those with trauma and/or psychosis are impaired, which may reduce the likelihood of having such networks [33, 176]. Therefore, social skills, as well as underlying mechanisms related to childhood trauma, such as stress sensitivity and cognitive biases, may be viable targets for psychotherapeutic interventions. Research on possible therapeutic outcomes in clinical groups converges to the necessity of distinguishing between patients with and without a history of childhood trauma [42]. It is not surprising, then, that treatments based on cognitive-behavioral therapy and exposure therapy, and focusing on reducing sensitivity to stress, readjusting cognitive biases, teaching social skills, psychoeducation, stabilization, and development of safe coping skills may be effective for those with a trauma history [176,177,179].

9 Conclusion

In this chapter, we have reviewed a body of research on the association between childhood traumatic experiences and psychoses. Evidence suggests that, at least in some people, exposure to childhood trauma with the intention to harm contributes to the onset of psychotic experiences and psychotic disorders. There is, however, also evidence suggesting that underlying traits, such as genetic predisposition, lower cognitive and social skills, may increase the risk for exposure to adversity. Whereas childhood trauma may be neither sufficient nor necessary to explain the onset of psychosis, at least part of that association seems to be causal. Nevertheless, it is important to note that this relationship is not restricted to psychosis only, but holds for a whole myriad of psychiatric disorders. Yet, findings that neurobiological changes can also be observed in individuals with a history of maltreatment, but no psychiatric disorder is perplexing. It may be that there are certain alterations that early life stress brings about, and also that there are compensatory mechanisms in resilient individuals, which enables them to balance out such neurodevelopmental insults [68, 91,92,93,94,96]. To make things more complex, the different pathways from childhood trauma to psychosis discussed in this chapter may act at different and complementary conceptual levels of examination (i.e., neurobiological, psychological), which, when taken together, might converge into an integrated model of psychosis [155, 180, 181].