Response

Dear Editors,

The article authored by Köhler-Forsberg et al^{Reference Köhler-Forsberg, Ge, Hauksdóttir, Thordardottir, Ásbjörnsdóttir and Rúnarsdóttir1} and published in the British Journal of Psychiatry (2024) 224, 6–12 provides a commendable contribution to the analysis of the literature concerning adverse childhood experiences (ACEs) and their impact on the psychological functioning of women diagnosed with schizophrenia (SCZ) or bipolar disorder. However, certain aspects of this study warrant careful consideration.

First, the authors focus on SCZ and bipolar disorder as severe mental illness (SMI). Nonetheless, there exist other SMI such as major depressive disorder (MDD), eating disorders, borderline personality disorder (BPD) and substance use disorder (SUD).^{Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi2,Reference Martínez-Martínez, Richart-Martínez and Ramos-Pichardo3} Integrating these mental disorders into the study would undeniably broaden its scope and enhance its comprehensiveness.

Second, in the Background section, depression is portrayed as an additional comorbid symptom within people affected by SCZ and bipolar disorder. It is recommended to avoid underscoring depression since it constitutes an integral component of bipolar disorder; unipolar depression is also recognised as a distinct mental disorder with significant associated costs, and should be considered thus in future research.^{Reference Martínez-Martínez, Richart-Martínez and Ramos-Pichardo3}

Third, concerning the instrument used for collecting ACEs, the Adverse Childhood Experiences International Questionnaire (ACE-IQ) has the merit of encompassing various types of ACEs. However, the majority of scales (e.g. ACE-IQ and the Childhood Trauma Questionnaire) that measure ACEs lack crucial details, such as the frequency (whether it was a single episode or chronic abuse) at which the ACEs occurred, the age of the children when they experienced the ACEs and the relationship between the abuser and the victim. The instruments are retrospective evaluation, and there is a possible recall bias. Although not being an integral part of the instrument, this information is essential to evaluate the impact of violence on the mental health of the individuals.^{Reference Liebschutz, Buchanan-Howland, Chen, Frank, Richardson and Heeren4}

Fourth, Table 1 delineates numerous variables. It is fundamental to mention the absence of evaluation regarding alcohol and other substance use, despite their well-documented and substantial influence on the manifestation and progression of bipolar disorder and SCZ.^{Reference Stokes, Kalk and Young5,Reference Smith and Hucker6} Furthermore, there are clinical data that remain undescribed, such as whether the person had been admitted as an inpatient and the history of family SMI, if they received psychotherapy and pharmacotherapy, the duration of treatment, whether they were undergoing treatment at the time of data collection and if there were any instances of suicidal behaviour. Additionally, aspects such as mania, psychotic symptoms and the onset of symptoms are not explicitly addressed. Given the significant influence of substance use on the development and progression of SCZ and bipolar disorder, it is essential to account for these factors to broaden a comprehensive understanding of the person's mental health.

Fifth, the article overlooks specifying the health status of the individuals in the household, whether they were receiving appropriate treatment and the potential relationship between substance abuse and mental disorder. These details are essential for achieving a comprehensive understanding of such intricate associations. It is notable that while the instrument explores mental illness and substance abuse in household members (well-established adverse childhood experience), it neglects to inquire about the person's own substance abuse and other comorbidities, as was discussed in the previous paragraph. Additionally, some SUDs in adolescence (e.g. cannabis) are associated with higher risk of later psychosis. Comorbidities with clinical disease (e.g. diabetes mellitus) or mental disorder (e.g. SUD) are possibly associated with poor quality of life, higher disability and worse treatment response. Childhood trauma is associated with higher risk of mental disorders and worse of course, but the differences between groups regarding treatment response remains uncertain.^{Reference Wrobel, Köhler-Forsberg, Sylvia, Russell, Dean and Cotton7}

Sixth, considering that bipolar disorder entails episodes of mania and depression, the use of the Patient Health Questionnaire-9 (PHQ-9) scale is questionable because it evaluates only depressive symptomatology, based on self-report. Scales such as the Mood Disorder Questionnaire (MDQ) and the scale developed by Palmier-Claus et al (2012) assess bipolar disorder and psychotic disorders respectively, and they could be incorporated in the study. It remains unclear why the authors opted for an instrument that measures solely the depressive aspect, thereby limiting the scope of the paper.^{Reference Wang, Woo, Ahn, Ahn, Kim and Bahk8,Reference Palmier-Claus, Ainsworth, Machin, Barrowclough, Dunn and Barkus9}

Seventh, in Table 1, the PHQ-9 scale is presented as the mean (s.d.). In the no-severe mental disorder group, the mean score was 7.38 (5.92), while in the severe mental disorder group, it was 14.7 (7.09). Notably, a score of ≥10 represents a well-established validated cut-off for a probable diagnosis of depression. The large standard deviation in the no-severe mental disorder group suggests that many subjects in this group would surpass the cut-off score of ≥10, indicating that a substantial number of individuals in this group meet the criteria for depression.

In conclusion, the article makes valuable contributions to the literature on ACEs and SMI, and it is important for future research to address certain areas for improvement. This can be achieved by including a broader range of severe mental disorders, clarifying the role of depression in bipolar disorder, enhancing ACE measurement instruments, incorporating variables related to substance use, assessing health and substance use within the household and using appropriate assessment scales. Such measures will enrich the analysis of the complex interactions between ACEs and SMI, facilitating a more balanced and well-founded interpretation of the results. Consequently, this will promote the development of more effective and personalised treatment strategies for individuals.