Neurodevelopmental aspects of adverse childhood experiences in psychosis: Relevance of the allostatic load concept

Previous studies have shown that neurodevelopmental characteristics of adverse childhood experiences (ACEs), such as their accumulation and minimal age at exposure, might moderate their impact on clinical expression of psychosis. However, it remains unknown as to whether specific neurodevelopmental characteristics of ACEs are associated with biological alterations observed in psychosis. In this study, we tested the hypothesis that younger minimal age at exposure as well as greater accumulation and severity of ACEs are associated with systemic biological dysregulations captured by the allostatic load (AL) index in patients with psychosis. The present study included 65 inpatients with psychotic disorders and 56 healthy controls (HCs). A total of 15 biomarkers were used to measure the AL index. Individuals with psychosis had significantly higher AL index as well as they reported greater accumulation and severity of ACEs compared to HCs. After adjustment for age, sex, the number of education years and the dosage of antipsychotics, greater accumulation of ACEs and younger minimal age at exposure were significantly associated with higher AL index in patients with psychosis. None of neurodevelopmental characteristics of ACEs was associated with the AL index in HCs. Our findings indicate that greater accumulation of ACEs and younger minimal age at exposure are related to biological dysregulations captured by the AL index in patients with psychosis. Future studies investigating the role of ACEs in the pathophysiology of psychosis need to consider their neurodevelopmental characteristics. It is also important to further explore timing of exposure to indicate critical developmental periods related to psychosis risk and better inform potential interventions.


Introduction
To date, several studies have shown that adverse childhood experiences (ACEs), such as emotional abuse or neglect, physical and sexual abuse, contribute to the development of various mental disorders, including psychosis (Varese et al., 2012). Importantly, exposure to ACEs has a negative impact on brain development, due to their effects on maturation of cortical areas and hippocampus (Teicher and Samson, 2013;Whittle et al., 2017). It also limits satisfaction of basic needs such as physical proximity and secure relationships leading to long-term psychological consequences (Cicchetti and Toth, 1995). Previous studies have consistently found that a history of ACEs is not only related to a risk of psychosis but also contributes to greater severity of psychotic symptoms and worse functional outcomes (for review see Misiak et al., 2017).
In recent years, a particular attention has been paid to neurodevelopmental aspects of ACEs. The "Life Cycle Model of Stress" has been proposed, and posits that there are several sensitive periods of brain development which are specifically susceptible to the effects of ACEs (Lupien et al., 2009). For instance, it has been shown that dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis response to acute stress might be the consequence of exposure to ACEs during infancy, while altered circadian regulation of cortisol release might occur in the context of later ACEs (Kuhlman et al., 2015). Another study (Raymond et al., 2021), demonstrated that individuals who were first exposed to ACEs at the age of 3-7 years show increased cortisol awakening response and blunted cortisol reactivity to stress. This age period reflects intensive development of amygdala and complex dendritic connections between the prefrontal cortex and the amygdala. However, the association between timing of exposure and the function of amygdala might be even more complex. Indeed, Zhu et al. (2019) found that amygdala response to threatening visual stimuli might be blunted in adults reporting exposure to physical abuse between ages 3 -6 years but increased in those with a history of emotional bullying between ages 13 -15 years. Some studies have also shown a dose-response association of ACEs with cortical and total brain volume reductions in adults (Begemann et al., 2021;Rosada et al., 2021).
Studies performed in various cohorts of patients with psychosis support the importance of neurodevelopmental aspects of ACEs. Trauelsen et al. (2015) found that a risk of non-affective psychosis increases two and a half times for each additional category of ACEs. Our group revealed that younger age at first exposure and greater accumulation of ACEs are associated with worse performance of cognition in subjects with schizophrenia spectrum disorders (Kasznia et al., 2021). Another study demonstrated that impairment of social cognition is associated with neglect experienced at the age of 11-12 years in patients with psychosis (Schalinski et al., 2018). Similarly, Alameda et al. (2015) found that early exposure to ACEs (before the age of 11 years), but not late exposure to ACEs (between ages 12 and 15 years), might be associated with long-term impairment of social functioning in subjects with early psychosis.
Importantly, it remains largely unknown what are the exact biological alterations mediating the effects of ACEs on various health outcomes. Although several studies have focused on investigating various measures of the HPA axis to understand lasting effects of ACEs, it seems reasonable to address multisystemic dysregulations (Schär et al., 2022). To support this approach, it should be noted that children exposed to high doses of corticosteroids, due to a diagnosis of asthma or arthritis, do not develop mental disorders attributable to ACEs (Teicher et al., 2002). Accumulating evidence indicates that exposure to various stressors, including early-life stress, activates a number of biological processes that might play adaptive roles. These processes have been named as "allostasis" (McEwen, 1998). Their prolonged activation, known as "allostatic load" (AL) might be deleterious, leading to the development of chronic disease outcomes. This theory has been the basis for developing the AL index, which might measure biological dysregulations (cardiovascular, metabolic, neuroendocrine and immune-inflammatory) associated with exposure to stress (Juster et al., 2010). Importantly, ACEs might be linked to higher AL index through disturbed HPA axis activity, subclinical inflammation and impaired glucose homeostasis (Baumeister et al., 2016;Danese and McEwen, 2012;Huang et al., 2015). Higher AL index has been observed in patients with psychosis, and might be related to greater severity of depressive and psychotic symptoms, worse cognitive, lower general functioning and cortical volume reductions (Berger et al., 2018;Chiappelli et al., 2017;Misiak et al., 2018).
Our previous study showed that ACEs, especially sexual abuse, are associated with elevated AL index in patients with psychotic disorders. Furthermore, we found that social diversion (stress coping strategy trough seeking social interactions) might suppress the impact of ACEs on the AL index in patients with psychosis (Piotrowski et al., 2020). However, this study was limited to investigating the association of specific types of ACEs (parental loss, parental antipathy and neglect, physical abuse and sexual abuse) with the AL index in psychosis. As mentioned above, previous studies have demonstrated that timing and accumulation of exposures might be important determinants of biological consequences of ACEs in terms of the HPA axis dysregulations as well as neurostructural and neurofunctional impairments. Following these considerations, the present study aimed to test the hypothesis that younger minimal age at exposure to ACEs as well as greater accumulation and severity of ACEs are associated with elevated AL index in patients with psychosis.

Participants
This study was based on previously reported sample of 65 individuals with psychotic disorders (39 patients with first-episode psychosis and 26 inpatients with schizophrenia) and 56 healthy controls (Piotrowski et al., 2020). In brief, they were recruited at two university hospitals in Wroclaw and Szczecin (Poland). Patients were diagnosed according to the DSM-IV criteria validated using the Operational Criteria for Psychotic Illness (OPCRIT) checklist (McGuffin, 1991). Most patients (95.4 %) were medicated on the day of assessment (chlorpromazine equivalent dosage: 373.0 ± 216.5 mg/day; there were 3 antipsychotic-naïve patients). Healthy controls were enrolled from the local community through advertisements. Both groups of participants were matched for age, sex and the level of parental education. The later one was applied as a proxy measure of socioeconomic status as this indicator might be associated with parental occupational position and income (Aarø et al., 2009).
The study received approval of the Ethics Committee at Wroclaw Medical University (Wroclaw, Poland), and all participants signed written informed consent.

Assessment of ACEs
The Polish version of the Childhood Experience of Care and Abuse Questionnaire (CECA.Q) developed by Bifulco et al. (2005) was used to assess exposure to ACEs (Kasznia et al., 2022). The CECA.Q is a self-report that records exposure to various ACEs under the age of 17 years. For the purpose of this study, we analyzed three aspects of ACEs: minimal age at exposure, accumulation and severity. Minimal age at exposure was defined as the age when any category of ACEs was experienced for the first time. Accumulation was calculated as the number of ACEs (parental loss, mother antipathy, mother neglect, father antipathy, father neglect, physical abuse and sexual abuse) reported under the age of 17 years (range: 0-7). The severity of exposure was calculated for all ACEs together, except for parental loss (this category of ACEs is not rated on the severity score) as described previously (Kasznia et al., 2021).

The AL index
A total of 15 biomarkers were used to calculate the AL index: (1) anthropometric measures: body-mass index (BMI) and waist-to-hip ratio (WHR); (2) cardiovascular markers: systolic and diastolic blood pressure; (3) lipids: total cholesterol, low-and high-density lipoproteins (LDL and HDL) and triglycerides; (4) glucose homeostasis parameters: glucose and insulin; (5) neuroendocrine markers: cortisol and dehydroepiandrosterone sulfate (DHEA-S) and (6) immune-inflammatory markers: high-sensitivity C-reactive protein (hsCRP), fibrinogen and albumin. Laboratory procedures were described in our previous publication (Misiak et al., 2018). The AL index was determined according to sex-specific percentile distributions in healthy controls. The score "1" was assigned to markers with the level above of the 75th percentile (below the 25th percentile for HDL, DHEA-S and albumin). Percentiles were calculated separately for male and female participants. Afterwards, the sum of all markers of above-mentioned categories was divided by the total number of markers in each category to provide balanced representations of distinct biological systems. Finally, the AL index was the sum of scores in each category (Misiak et al., 2018;Seeman et al., 2001).

Data analysis
The Kolmogorov-Smirnov test was applied to test data distribution.
Between-group differences were evaluated using the χ 2 test or the Mann-Whitney U-test, where appropriate. Spearman rank correlation coefficients were used to analyze correlations of continuous variables.
Non-parametric tests were used due to non-normal distribution of characteristics of ACEs (minimal age at exposure, accumulation and severity). Outliers were visualized using box plots. The relationship between characteristics of ACEs (minimal age at exposure, accumulation and severity) and the AL index was assessed using multiple regression analyses after checking necessary assumptions. The characteristics of ACEs (minimal age at exposure, accumulation and severity were transformed to z-scores to limit potential collinearity (Aiken and West, 1991). Next, interaction terms of the group status (psychosis vs. healthy controls) with the characteristics of ACEs were calculated (group × minimal age at exposure, group × accumulation and group × severity). The AL index was included as the dependent variable. Group status (psychosis vs. healthy controls), the characteristics of ACEs and interaction terms were included as independent variables. Age, the number of education years and chlorpromazine equivalent dosage were added as covariates. Age, sex and education were added as covariates due to previously reported association with the AL index (Crimmins et al., 2006;Ding et al., 2019;Kerr et al., 2020). Chlorpromazine equivalent dosage was controlled for in data analysis due to potential metabolic side effects of antipsychotics ( Lieberman, 2004). Results were interpreted as statistically significant if the p-value was lower than 0.05. Statistical analyses were performed using the SPSS software, version 28. Post-hoc power analysis was performed using the G*Power software (Faul et al., 2007).

Results
There were no significant between-group differences in terms of age, sex and the level of parental education (Table 1). Cigarette smoking rates were significantly higher in patients with psychosis compared to healthy controls. The number of education years was significantly lower in patients with psychosis. Individuals with psychosis reported significantly higher accumulation and severity of ACEs compared to healthy controls. Minimal age at exposure to ACEs was similar in both groups of participants. A history of any ACEs was significantly more frequent in patients with psychosis. Finally, the AL index was significantly higher in patients with psychosis compared to healthy controls (2.1 ± 1.1 vs. 1.2 ± 1.0, p < 0.001, power: 99.5 %).
Bivariate correlations between investigated variables are reported in Table 2. Significant correlations of the AL index with characteristics of ACEs were found in subjects with psychosis, but not in healthy controls (Fig. 1). Specifically, there were significant positive correlations of the AL index with accumulation and severity of ACEs among individuals with psychosis. Moreover, a significant negative correlation between minimal age at exposure to ACEs and the AL index was observed. Older age was also significantly associated with higher AL index in subjects with psychosis, but not in healthy controls. Accumulation of ACEs was related to their greater severity in both groups of participants. However, a significant negative correlation between minimal age at exposure and accumulation of ACEs was observed only in subjects with psychosis. The analysis of boxplots for characteristics of ACEs identified four outliers. Therefore, results of multiple regression analyses (Table 3) are reported before and after excluding outliers (n = 4). In the unadjusted analysis, the AL index was significantly associated with the group status (psychosis vs. healthy controls), main effects of accumulation of ACEs and the interaction term (group status × accumulation of ACEs). After adjustment for age, sex, education years and chlorpromazine equivalent dosage (model 2), there were significant associations of the AL index with the group status (psychosis vs. healthy controls), interaction terms (group status × accumulation of ACEs and group status × minimal age at exposure). Main effect of accumulation of ACEs appeared to be not significant after including covariates (Model 2). Both sets of analyses (with and without outliers) provided similar results.

Discussion
Main findings from the present study indicate that neurodevelopmental characteristics of ACEs, including greater accumulation and younger minimal age at exposure, are associated with elevated AL index in psychosis. There were no significant associations of these characteristics of ACEs with the AL index in healthy controls. However, accumulation of ACEs was significantly higher in patients with psychosis compared to healthy controls. This observation might suggest that ACEs might impact the AL index after exceeding certain cut-off.
Neurodevelopmental characteristics of ACEs, including their accumulation and minimal age at exposure might be important moderators of their effects with respect to psychosis risk and/or clinical manifestation. A recent meta-analysis showed that ACEs increase a risk of psychosis especially in subjects exposed to multiple trauma types (Pastore et al., 2022). However, little is known about the effects of timing of exposure with respect to psychosis risk as the majority of previous studies have not recorded this variable (Pastore et al., 2022;Varese et al., 2012). For instance, Fisher et al. (2010) in a case-control cohort of patients with first-episode psychosis and geographically matched controls failed to find the effect of timing of exposure to ACEs with psychosis risk. Another study of a population-based birth cohort demonstrated stronger association of ACEs experienced proximally to the occurrence of psychotic-like experiences (Croft et al., 2019). However, it has been shown that younger age at first exposure to ACEs and greater accumulation of ACEs might be associated with greater cognitive impairment and more severe positive symptoms (Kasznia et al., 2021;Schalinski et al., 2019Schalinski et al., , 2018. Interestingly, younger minimal age at exposure to ACEs has been associated with unfavorable functional outcomes in subjects with first-episode psychosis (Alameda et al., 2015). At this point, it should be noted that previous studies have also found a negative correlation between the AL index and social functioning in subjects at clinical risk of psychosis and patients with psychotic disorders (Berger et al., 2018;Berger et al., 2020;Piotrowski et al., 2019). Moreover, previous studies have demonstrated that elevated AL index is associated with greater severity of psychotic symptoms, working memory impairments and reduced cortical thickness (Berger et al., 2018;Chiappelli et al., 2017;Misiak et al., 2019). Importantly, working memory deficits are highly prevalent in patients with psychosis and might be attributable to dysfunction of the prefrontal cortex characterized by high density of glucocorticoid receptors (Arnsten, 2009 Abbreviations: ACEsadverse childhood experiences, ALallostatic load, MADRSthe Montgomery-Asberg Depression Rating Scale, PANSS-N -the Positive and Negative Syndrome Scalenegative symptoms, PANSS-Pthe Positive and Negative Syndrome Scalepositive symptoms, SOFASthe Social and Occupational Functioning Assessment Scale. Significant differences (p < 0.05) were marked with bold characters.
There are some limitations of the present study that should be considered. This study was performed in a relatively small sample of patients meeting the criteria of various psychotic disorders. In this regard, we were not able to indicate what time period of exposure for ACEs might be most sensitive for developing elevated AL in adulthood. Another point is that we did not test observed associations separately for patients with first-episode psychosis and those with chronic schizophrenia. This would be important due to the following reasons: (1) the AL has been found to increase between 20 and 60 years of age (Crimmins et al., 2003) and might be higher in older patients with schizophrenia (Savransky et al., 2018) and (2) a history of ACEs may not be associated with elevated AL index in older individuals (Scheuer et al., 2018). Moreover, assessment of ACEs was based on self-reports, and thus a recall bias cannot be excluded. Another point is a lack of general consensus about calculating the AL index. It should also be noted that this study does not provide specific insights into the mechanisms linking neurodevelopmental characteristics of ACEs with the AL index. Indeed, there are various processes that might mediate or moderate the association between ACEs and the AL index in adults, including educational attainments, social support, health behaviors, adult stress, post-traumatic stress disorder, coping strategies and aging . Moreover, it remains unknown what is the relevance of observed associations with respect to clinical manifestation of psychosis. To further understand this point, it would be crucial to include the measures of functioning and the level of distress. Finally, a cross-sectional design does not provide grounds to conclude about causality.
In summary, the present findings imply that younger minimal age at exposure to ACEs and their greater accumulation are associated with elevated AL index in patients with psychotic disorders. These observations suggest a necessity to include neurodevelopmental characteristics in studies investigating the association between ACEs and the development of psychosis. Longitudinal studies investigating a variety of psychological processes and potential neural substrates are needed to provide potential insights into mechanisms underlying the association of  Abbreviations: ACEsadverse childhood experiences, ALallostatic load, CPZeqchlorpromazine equivalent dosage, HCshealthy controls. a p < 0.01. b p < 0.05. c p < 0.001.

Fig. 1.
Bivariate correlations between characteristics of adverse childhood experiences (ACEs) and the allostatic load (AL) index. Top plots present correlations in subjects with psychosis and bottom plots refer to healthy controls. Minimal age at exposure was expressed in years. In two participants minimal age at exposure was recorded as zero due to early maternal loss.
ACEs with AL in psychosis. Recognition of these mechanisms might be of importance for developing specific interventions that aim to restore homeostasis and psychological well-being in subjects with psychosis exposed to ACEs.

Funding source
This study was funded from science budget resources granted for the years 2016-2019 (the Iuventus Plus grant awarded by the Ministry of Science and Higher Education, Poland , grant number: IP2015 052474).

Table 3
Results of multiple regression analysis testing the associations between neurodevelopmental characteristics of adverse childhood experiences (minimal age at exposure, their accumulation and severity) and the allostatic load index (dependent variable) in subjects with psychosis (n = 65) and healthy controls (n = 56). Abbreviations: ACEs, adverse childhood experiences; CPZeq, chlorpromazine equivalent dosage; HCs, healthy controls. Significant associations (p < 0.05) were marked in bold characters.