Childhood trauma and psychosis - what is the evidence?

In the last decade, a substantial number of population-based studies have suggested that childhood trauma is a risk factor for psychosis. In several studies, the effects held after adjusting for a wide range of potentially confounding variables, including genetic liability for psychosis. Less is known about the mechanisms underlying the association between childhood trauma and psychosis. Possible pathways include relationships between negative perceptions of the self, negative affect, and psychotic symptoms, as well as biological mechanisms such as dysregulated cortisol and increased sensitivity to stress. Psychotic patients with a history of childhood trauma tend to present with a variety of additional problems, including post-traumatic stress disorder, greater substance abuse, higher levels of depression and anxiety, and more frequent suicide attempts. Initial studies suggest that trauma-specific treatments are as beneficial for these patients as for other diagnostic groups.

During the past, few decades, a large body of research has furthered our understanding of the relationships between early adversity and psychological difficulties later in life. At the heart of this research has been the role of childhood trauma, especially sexual and physical abuse, and, increasingly over the years, emotional abuse and neglect.[1-4] Although a variety of other forms of childhood adversity, such as parental loss, separation, and discord, and bullying, contribute to later psychopathology,[5-6] childhood trauma appears to have particularly powerful and long-lasting effects. After controlling for other psychosocial risk factors, childhood trauma has been associated with the development of most psychiatric disorders, including mood and anxiety disorders, eating disorders, personality disorders, dissociative disorders, and substance dependence.[7-11] Until recently, however, researchers have focusscd predominantly on the relationship of childhood trauma to nonpsychotic disorders. The possible reasons for this have been discussed elsewhere.[12] They include a lack of confidence and belief in the utility of intervention in psychotic patients and some uncertainty as to whether patients' reports can be trusted. However, the reliability of psychotic patients' abuse reports has repeatedly been established,[13] and preliminary studies have shown trauma-related interventions to be effective in this group (see below).

Population-based studies

In the last decade, a substantial number of populationbased studies suggested that childhood trauma is also an important risk factor for psychosis (Table I). In almost all of these studies, a history of abuse was related to psychotic symptoms and/or the diagnosis of a psychotic disorder either during adolescence[14-17] or adulthood.[18-24] In a prospective study, Arsène ault et al[17] surveyed mothers of 2232 twin children at 5, 7, 10, and 12 years of age concerning exposure to physical maltreatment and accidents and assessed the twins themselves at age 12 to determine experience of psychotic symptoms as part of the Environmental Risk Longitudinal Twin Study (E-Risk). Children who had experienced intentional physical harm (maltreatment) were more likely to report psychotic symptoms at age 12 than those exposed to unintentional physical harm (accidents). These effects held after adjusting for a wide range of potentially confounding variables including genetic liability for psychosis. An even greater risk for psychotic symptoms was found amongst, children who experienced both physical abuse from an adult, and bullying by peers, indicating a cumulative effect, of trauma on psychosis outcomes in early adolescence. In another study, Bcbbington et al[23] used data from a health survey of 7353 adults to examine whether unwanted sexual experiences were associated with probable psychotic disorder. Again, psychosis was related to traumatic events in a doseresponse fashion, with nonconsensual sexual intercourse evidencing a stronger association than non contact sexual abuse. The association between sexual abuse and psychosis was also only present amongst, women, which replicates previous findings in a clinical population,[25] and suggests further exploration of gender effects in this area is required. Therefore, more robust evidence is mounting to support the role of childhood trauma in the etiology of psychosis. Only one population-based study[26] could not confirm the link between childhood trauma and psychosis. In this prospective study, no increase in schizophrenia was found among adults who had histories of sexual abuse in childhood according to official records. However, a recent study by this group found a relationship between documented abuse and psychosis[21] using a longer follow-up period.

Possible pathways from childhood abuse to psychosis

Less is known about, the mechanisms underlying the association between childhood trauma and psychosis. A few studies have indicated that, childhood trauma (particularly childhood sexual abuse) may result in even higher rates of psychosis or psychotic symptoms when it, occurs together with cannabis use.[27,28] Cross-sectional studies have demonstrated that negative perceptions of the self, anxiety, and depression partially mediated associations between trauma (not always limited to childhood) and psychotic symptoms.[22,29] They suggest strong relationships between negative personal evaluations and low self-esteem, negative affect, and the characteristics of positive symptoms. Lardinois et al[30] found a significant, interaction between daily life stress and childhood trauma on both negative affect, and intensity of symptoms in patients with psychosis, suggesting that, a history of childhood trauma is associated with increased sensitivity to stress. Biological mechanisms such as reduced cortical thickness[31] and dysregulated Cortisol[32] following exposure to childhood trauma have also been recently investigated which may well facilitate the development of psychosis. Moreover, gene-environment interactions are likely to play a role in the relationship between childhood trauma and psychosis. In a recent study, Alemany et al[33] found that the relationship between childhood abuse and psychosis was moderated by the BDNF-Va166Met polymorphism. In a sample of 533 students, Met carriers reported more positive psychotic-like experiences when exposed to childhood abuse than did individuals carrying the Val/Val genotype. These preliminary studies provide direction for future exploration, ideally in longitudinal datascts, of the mechanisms that may form the pathway between childhood trauma and psychosis.

Additional psychopathology

One of the most prevalent consequences of childhood abuse is post-traumatic stress disorder (PTSD). While about, 3% to 5% of individuals in the general population fulfil a current diagnosis of PTSD (eg, rcf 34), the prevalence of the disorder in samples of patients with schizophrenia is 17% to 46% (eg, refs 35,36). Rates of current, PTSD in individuals with bipolar disorder range from 11 % to 24% (eg, refs 37,38). Psychotic patients with a history of childhood trauma and/or PTSD have a more severe clinical profile compared with those without these experiences. They report more current or lifetime substance abuse,[39,40] higher levels of current depression and anxiety,[41,42] and more dissociative symptoms.[43,44] Childhood sexual abuse has specifically been linked to hallucinations and delusions[20,45] and the content of these positive symptoms may be related to patients' traumatic experiences.[46] Psychotic patients with a history of childhood trauma tend to present with a variety of additional problems, similar to that of other populations with childhood trauma. Victims of abuse report increased levels of suicidal ideation and more frequent suicide attempts.[40] They have also been reported to be less able to sustain intimacy, and to be more prone to emotional instability.[47] Finally, a. history of childhood abuse is associated with worse overall social functioning,[48,49] lower remission rates,[50] and poorer compliance with treatment.[40,51]

Promising treatments for patients with childhood trauma

Initial studies suggest that trauma-specific treatments are as beneficial for patients with psychosis as for other diagnostic groups. Psychotic patients with early and complex trauma can benefit, from present-focused treatments with an emphasis on psychoeducation, stabilization, and the development of safe coping skills. Trappier and Newville,[52] for instance, treated 24 patients with chronic schizophrenia and complex PTSD using the first phase of skills training in affect, and interpersonal regulation (STAIR).[53] The first phase of this cognitive-behavioral therapy (CBT) program is focussing on skills training in affect, and interpersonal regulation. A control group of patients received supportive psychotherapy sessions. After 12 weeks of treatment, the patients in the STAIR group showed significant reductions in Impact, of Events Scale scores and positive psychotic symptoms, while no improvement in these was observed in the control group. Furthermore, several case studies and open trials have reported that exposure-based interventions can also be used safely and effectively in patients with psychosis. Frueh et al[54] treated 20 patients with PTSD and either schizophrenia or schizoaffective disorder via an 11-week CBT intervention that, consisted of 14 sessions of psychoeducation, anxiety management, and social skills training, as well as 8 sessions of exposure therapy, provided at community mental health centers. Treatment completers showed significant. PTSD symptom improvement, maintained at 3-month followup. Moreover, significant improvements existed with regard to other targeted domains (eg, anger, general mental health). A further approach to treating PTSD in patients with psychosis was developed by Mueser et al.[55] The 12- to 16-session program combines psychoeducation and breathing retraining with cognitive restructuring to address thoughts and beliefs related to trauma experiences and their consequences. In a recent randomized controlled trial the program was compared with treatment as usual in 108 patients with severe mental illness (39% bipolar disorder, schizophrenia or schizoaffective disorder). At 6-month follow-up, CBT clients had improved significantly more in PTSD symptoms, perceived health, negative trauma-related beliefs, and case manager working alliance.

Conclusions

The evidence for an association between childhood trauma and psychosis is steadily accumulating, and exploration of potential mechanistic pathways has begun. Emerging findings from longitudinal studies and demonstration of a dose-response relationship in others suggest a role of childhood trauma in the development of psychosis. The relative influence of other variables in this relationship, however, warrants further investigation. Independent from the question of causality, childhood trauma and PTSD are frequent in patients with psychosis and severely affect, course and outcome. More research is therefore needed to further develop and evaluate appropriate treatments for psychotic patients suffering from the consequences of childhood trauma. Nevertheless, the existing trials suggest that patients with psychotic disorders can benefit from both presentfocused and trauma-focused treatments, despite severe symptoms, suicidal thinking, and vulnerability to hospitalizations.

Table I.

Population-based studies investigating the association between childhood abuse and psychosis. Adj, adjusted for confounders. OR, odds ratio; RR, relative risk. Adapted from ref 56: Morgan C, Fisher H. Environment and schizophrenia: environmental factors in schizophrenia: childhood trauma-a critical review. Schizophr Bull. 2007;33:3-10. Copyright © Oxford University Press, 2007

Authors	Study design	Sample	Measure of association with psychosis
Bebbington et al[22] (UK)	Cross-sectional survey	8580 adults aged 16-74	Sexual abuse vs none: Adj OR* 2.9 (1.3-6.4)
			*Adusted for interrelationship between other adverse
			events and depression
Janssen et al [23] (Netherlands)	Prospective cohort	4045 adults aged 18-64	Abuse vs no abuse:
			BPRS any psychosis Adj OR* 2.5 (1.1-5.7)
			BPRS pathology level Adj OR* 9.3 (2.0-43.6)
			Need-based disorde Adj OR* 7.3 (1.1-49.0)
			* Adjusted for a range of variables, including any other
			psychiatric dignosis and psychosis in first-degree relatives
Spataro et al [26](Australia)	Prospective cohort	3 141 357 adults	Relatives risk of schizophrenic discorder in controls vs
		(mean age 27)	schizophrenic discorder in cases with documented sexual
			abuse:RR 1.2 (0.7-2.1)
Whitfield et al[24] (USA)	Cross-sectional survey	17 337 subjects	Risks of ever having had a hallucination:
		(mean age 57)	Emotional abuse: Adj OR* 2.3 (1.8-3.0)
			Psysical abuse: Adj OR* 1.7 (1.4-2.1)
			Sexual abuse: Adj OR* 1.7 (1.4-2.1)
			*Adjusted for age, sex, ethnicity, and educational
			attainment
Lataster et al[14] (Netherlands)	Cross-sectional survey	1290 adolescents aged 12-16	Nonclinical psychotic symptoms:
			Adj OR* 4.5 (1.5-13.3)
			*Adjusted for age, gender and socioeconomic status
Spauwen et al[15] (Germany)	Prospective cohort	2524 subjects aged 14-24	Narow psychosis
			Any trauma: Adj OR* 1.9 (1.2-3.1)
			Sexual abuse: Adj OR* 1.6 (0.5-5.1)
			Physical threat: Adj OR* 2.1 (1.2-3.9)
			Rape: Adj OR* 2.3 (0.6-9.2)
			*Adjusted for gender, socieconomic status, urbanicity,
			cannabis use, baseline psychiatric discorders, and psychosis
			proneness
Shevlin et al[19] (United States)	Cross-sectional survey	5887 adults aged 15-54	Nonaffective psychosis
			Physical abuse: Adj OR* 2.7 (1.1-6.5)
			Sexual abuse: OR not reported but non-significant
			*Adjusted for depression
Kelleher et al[16] (ireland)	Cross-sectional survey	211 adolescents aged 12-15	Psychotic symptoms:
			Psysical abuse: Adj OR* 6.0 (1.3-28.0)
			Sexual abuse: Adj OR* 4.2 (0.3-50.5)
			*Adjusted for gender and socioeconomic status
Cutajar et al[21] (Australia)	Prospective cohort	5436 adults aged 14-57	Relative risk of shizophrenic discorder in controls vs
			schizophrenic disorder in cases with documented sexual
			abuse: OR 2.6 (1.6-4.4)
Shevlin et al[20] (United States)	Cross-sectional survey	2353 adults (mean age 44)	Lifetime experience of visual hallucinations:
			Physical abuse: Adj OR* 3.2 (1.5-7.1)
			Rape: Adj OR* 3.4 (1.7-6.8)
			Lifetime experience of auditory hallucinations:
			Physical abuse: Adj OR* 4.56 (2.0-10.6)
			Rape: Adj OR* 3.0 (1.4-6.3)
			*Adjusted for gender, age, urbanicity, marital status,
			educational attainment, employment, and substance
			dependence
Arseneault et al[27] (UK)	Prospective cohort	2232 children aged 12	Psychotic symptoms:
			Physical abuse: Adj RR* 2.5 (1.5-4.2)
			*Adjusted for gender, socioeconomic deprivation and IQ
Bebbington et al [23] (UK)	Cross-sectional survey	7353 adults aged 16 or over	Probable psychosis:
			Any sexual abuse: Adj OR* 3.2 (1.3-7.6)
			*Adjusted for age, social class, ethnicity, educational
			attainement, household income, and family structure