Psychological Trauma Predicts Obesity in Welsh Secure Mental Health Inpatients

Abstract

People in secure psychiatric services experience high levels of exposure to early psychological trauma, are often diagnosed with severe mental illness (SMI) and have increased risk for obesity. Developmental trauma, SMI and obesity are associated with poor physical health outcomes and early death. This study aimed to assess the predictive power of Adverse Childhood Experiences (ACEs), self-harm and psychiatric diagnosis for secure inpatient weight gain and obesity. Data for 248 Welsh patients accommodated in low, medium, and high secure hospitals throughout England and Wales was analyzed. Patient admission bodyweight (in kilograms), bodyweight at the time of audit, and patient BMI at the time of audit was collected. Sex, psychiatric diagnosis, length of current admission, number of ACEs, and frequency of self-harm were also examined. Patients gained significant amounts of weight between admission until the time of the audit (median period = 1 – 2 years) and showed high levels of obesity. Self-harm frequency significantly predicted weight difference. Number of ACEs and diagnosis of schizophrenia significantly predicted patient BMI at the time of audit. The study findings indicate that psychological trauma and the nature of mental illness are important factors driving weight gain and excess morbidity in this vulnerable group.

Introduction

People with Serious Mental Illness (SMI), have a markedly greater prevalence and severity of obesity compared to the general population (Afzal et al., 2021). Obesity is generally measured through Body Mass Index (BMI; calculated by weight in kilograms divided by height in meters, squared (kg/m²)). A BMI greater than 29.9 kg/m² is considered to reflect obesity. High BMI values are associated with more mental and physical health co-morbidities (National Institute of Diabetes & Digestive & Kidney Diseases, 2018). Given the high prevalence of obesity within SMI populations, it is unsurprising that obesity-related co-morbidities are also prevalent (Public Health England, 2018). It has been argued that these physical health co-morbidities significantly contribute to the 14.5-year reduced life expectancy of people diagnosed with SMI compared to the general population (Hjorthøj et al., 2017). This is particularly concerning for individuals treated in secure inpatient services as 70% of these die prematurely, primarily due to obesity related illnesses such as respiratory and cardiovascular disease (Rees & Thomson, 2020).

Secure and locked inpatient mental health services in the England and Wales treat people detained involuntarily under the Mental Health Act ( 1983; Mental Health Act Revised, 2007) or, less often, the Mental Capacity Act ( 2005) who have high rates of SMI diagnosis and obesity related comorbidities (Mills & Davies, 2022). A review by Moss et al. (2021) found that the prevalence of obesity in secure inpatient populations ranged between 30—80%. It has been noted that excessive weight gain can be seen to follow from admission for secure inpatient mental health treatment (Davies et al., 2023a; 2023b; Every-Palmer et al., 2018; Hilton et al., 2015). Davies et al. (2023b), for example, found that patients gained an average of 4 kg in weight over the first 12 wk of secure inpatient treatment. The risk of weight gain also continues past the early stages of inpatient treatment: Hilton et al. (2015) found, in a sample of 122 secure inpatients, that 52% were either overweight or obese on admission, and that this proportion increased to 68% after six months of treatment, average weight gain during admission was 8.7 kg.

Rapid weight gain and high obesity rates in secure service inpatients exist despite Government policy and initiatives aimed at tackling this issue (Public Health England, 2021). A limitation of policy and standards to date is that they do not fully address the multifaceted, interactive, and complex nature of obesity in people with SMI. Policies have failed to consider the impact of ultra processed foods which are energy dense, contain high levels of fat, sugar, and salt, and can increase appetite when consumed (Hall et al., 2019). These foods are also highly accessible within secure inpatient settings (Walker et al., 2023). Obesity in secure inpatient populations could be attributed to the side effects of antipsychotic medications. For example, powerful atypical antipsychotics such as clozapine and olanzapine are associated with significant increases in weight upon their use (Leucht et al., 2023), and these are prescribed frequently within secure inpatient services (Farrell & Brink, 2020). However, a recent study in Welsh secure inpatients found that high weight-gain risk antipsychotics did not significantly predict weight gain over the initial stages of inpatient treatment (Davies et al., 2023b).

There is also literature that suggests factors present in inpatient settings such as limited exercise and the food provided in such services have a significant impact on patients and their lifestyle in these settings (Day & Johnson, 2017; Hilton et al., 2022). Furthermore, the secure inpatient environment has been referred to as obesogenic, due to the quality and quantities of food that are made available to patients, as well as the restrictive nature of the environment, which makes it difficult for patients to engage in physical activity and stay motivated to maintain a healthy lifestyle (Davies et al., 2023a). However, there is a gap in the literature regarding the impact that psychological factors have on poor weight management in this population.

Obesity research in the general population has provided an insight into the deleterious effects that exposure to traumatic experiences in early life can have on mental and physical health outcomes throughout the life course (Hughes et al., 2017). A seminal study conducted by Felitti et al. (1998) established that Adverse Childhood Experiences (ACEs) are stressful and potentially traumatic events or situations that occur during childhood and/or adolescence and are implicated in poor physical and mental health. More recently, the Welsh ACE study (Bellis et al., 2016) explored the prevalence of ACEs and their associations with a variety of adverse social and health outcomes in later life in a sample of 2,000 adults from the Welsh population. These ACEs included exposure to verbal abuse, physical abuse, sexual abuse, and growing up in a household that included parental separation, domestic violence, mental illness, alcohol abuse, drug use and incarceration. Bellis et al. (2016) found a dose dependent relationship between number of ACEs and the prevalence of a poor diet, whilst also controlling for socio-demographic factors, with four or more ACEs associated with a 2.2-fold increased risk of having a poor diet. This may contribute to explaining why there are also strong associations between ACEs and adult obesity (Schlauch et al., 2022). Whilst Bellis et al. (2016) did control for socio-economic factors, the authors do not state exactly which socio-demographic factors were controlled for, and so it could be argued that the relationship is mediated by socioeconomic deprivation, given that this underpins risk of experiencing ACEs and risk of poor physical health (Huang et al., 2020).

Early psychological trauma is also strongly associated with the risk of developing SMI and personality disorders in adulthood (Herzog & Schmahl, 2018; Matheson et al., 2012; Michel et al., 2022; Varese et al., 2012). The strong link between ACEs and mental ill health is supported by an observed doubling in ACEs in secure inpatients relative to the general population (Mills & Davies, 2022). In their review of Welsh patients in secure care, Mills and Davies (2022) found that 28% of patients had experienced four or more ACEs, compared to 14% highlighted in Bellis et al.’s (2016) national study. The prevalence rates of ACEs for women in secure care are even more concerning with exposure to four or more ACEs that is four times higher than the Welsh population.

Apart from impacting physical and mental health, the traumatic childhood experience can also impact inpatient psychiatric treatment outcomes. Exposure to early psychological trauma can increase the likelihood of challenging institutional behaviors (Banks, 2021), such as self-harm (Holden et al., 2021). Self-harm has been conceptualized as a feature of emotional dysregulation, a problem that often has its roots in experiences of trauma and adversity (Kerig, 2020). Emotional dysregulation is also postulated to be a mechanism at play when individuals overeat (Micanti et al., 2016). Some trauma informed models of mental distress conceptualize both self-harm and over-eating as strategies that service users can resort to for managing their emotional needs and self-soothing in the absence of more helpful ways of doing so (Johnstone & Boyle, 2018). Therefore, it is plausible that weight gain and obesity in secure services are mediated by similar psychological mechanisms as those that underpin institutional self-harm (Buckholdt et al., 2014).

Present study

To date there have been no published research that has explored the predictive power of indices of trauma on secure inpatient weight gain and obesity. A better understanding of the role of trauma in obesity amongst secure mental health inpatients would highlight the importance of integrating trauma informed approaches to health promotion and healthy eating within secure inpatient settings. The aim of the current study was to build on previous literature that outlines predictors of weight gain and obesity in secure inpatient services. Specifically, the current study aimed to explore if sex, primary psychiatric diagnosis, number of ACEs, frequency of self-harm, and frequency of abuse victimization are reliable predictors of weight gain and obesity in a population of Welsh secure inpatients.

Method

Participants

The current analysis was commissioned by the Quality Assurance Improvement Service (QAIS) department of the Welsh National Collaborative Commissioning Unit as part of a broader routine audit of patient care using data derived from patients’ healthcare records. The sample were Welsh nationals accommodated across England and Wales in high, medium, or low secure services. The three high secure services in England are maintained within the NHS and serve Wales. Medium and low secure beds for Wales are commissioned from both the NHS and the private sector and are located throughout Wales and England. Two hundred and forty-eight service users’ health care records were included in this study. Service users had a mean age of 39.8 years [range 18-79 years], 81.5% were male (N = 202) and 18.5% were female (N = 46). The distribution of service users across secure service settings is presented in Table 1.

Table 1.

Patients in NHS and private hospitals, by security level.

	Low	Medium	High	Total
NHS	51	48	19	118
Private	89	41	0	130
Total	140	89	19	248

Ethics approval

Ethics approval was obtained from Cardiff Metropolitan University School of Sport and Health Sciences Ethics Panel (application number PGT-4347). The study was also approved as a service audit by all relevant Welsh Health Board Research and Development departments. The audit complied with General Data Protection Regulation (GDPR) requirements for the use of confidential patient data without consent.

Design

The current analysis was a retrospective cohort-regression design and there were several predictor variables. Sex was a dichotomous variable with levels male or female. Primary psychiatric diagnosis was categorized as 1) schizophrenia, which included other psychosis derived disorders as outlined by the International Classification of Diseases (ICD-10; World Health Organization [WHO], 2016), and 2) personality disorder. These variables were dichotomous with levels presence or no presence. Number of ACEs were recorded as a continuous variable. Self-harm frequency, and victim of physical, psychological and/or emotional abuse frequency were ordinal variables with levels no history since admission, once or twice since admission, monthly, weekly, or daily. Given that patients will have varying length of time spent in the placement at the time the audit was conducted, this was controlled for by including the length of their current admission at the time of audit as a predictor variable. This was an ordinal variable with five levels: < 1 year, 1 - 2 years, 2 - 4 years, 4 - 7 years, and 7 years or more.

There were two separate outcome variables in this study: 1) weight difference, in kilograms (kg), between admission to their most recent recorded weight and 2) most recent recorded BMI, calculated by patient weight in kilograms divided by height in meters squared (kg/m²). Both outcome variables were continuous measures.

Materials

Information extracted from patient healthcare records were used in this analysis. The healthcare records included physical and digital copies of care and treatment plans, multidisciplinary clinical reports, contemporaneous clinical notes, and medical charts. With regards to data collected, patients’ most recent primary psychiatric diagnosis collected from medical records were categorized in line with the ICD-10. Number of ACEs were reported by staff members in each hospital prior to audit and were also extracted from medical records. Self-harm frequency, and victim of physical, psychological and/or emotional abuse frequency were extracted from the National Collaborative Commissioning Units Commissioning Care Assurance and Performance System, used to record patient incidents. The relevant data were extracted, de-identified, stored in an encrypted Microsoft Excel document, and analyzed using the Statistical Package for Social Sciences (SPSS version 29; IBM, 2022).

Procedure

All data for the analysis was collected onsite by the QAIS audit team on behalf of the NHS commissioners of specialist care for Welsh secure inpatients, for the purposes of quality and performance monitoring of service providers, and clinical outcomes against care and treatment plans for patients. Four auditors collected the data from physical and digital patient healthcare records held by the individual service providers between November 2019 and November 2020.

Method of analysis

A repeated measures t-test was conducted to evaluate differences in weight in kilograms on admission and weight in kilograms at the time of audit. A backward regression analysis was conducted with sex, length of admission, number of ACEs, primary diagnosis of schizophrenia, primary diagnosis of a personality disorder, self-harm frequency and victim of physical, psychological, and/or emotional abuse frequency as predictor variables. Variables were included in the regression model either because they were direct markers of trauma (i.e., number of ACEs and victim of physical, psychological, and/or emotional abuse frequency), or because certain characteristics show a greater propensity to having experienced psychological trauma (i.e., sex and diagnosis). The longer a patient has been receiving treatment in secure hospital has been shown to be associated with greater levels of weight gain (Hilton et al., 2015) and so length of admission was included to control for this.

Weight change in kilograms from admission to the most recent recording by the time of the audit was the outcome variable. A second backward regression analysis was conducted with the same predictor variables but with most recent BMI at time of audit as the outcome variable.

Results

Table 2 displays descriptive information for the sample used in this analysis, by sex. Most patients in the sample had a primary diagnosis of schizophrenia (67.7%). However, by sex, women were less likely to have a primary diagnosis of schizophrenia (39.1%) compared to men (74.3%). The opposite was true for a primary personality disorder diagnosis: 25% of the total patient sample had a diagnosis of personality disorder, comprising 54.3% of women compared to 18.3% of men. Similar sex differences were observed for self-harm: 52.1% of the women had between one and two incidents of self-harm since admission compared to 12.5% of men. The majority of the sample in this study did not have any recorded experiences of inpatient abuse in their current placement (89.1%) and this was the case for both men and women. Women had a greater median number of recorded ACEs (mdn = 4), compared to men (mdn = 3). More men were in their current placement for longer than four years (17.4%) compared to women (13.0%). The median length of stay for both men and women were 1 -2 years.

Table 2.

Descriptive statistics for sample.

Predictor Variables	Male N (%)	Female N (%)	Overall N (%)
Primary diagnosis schizophrenia
Yes	150 (74.3)	18 (39.1)	168 (67.7)
No	52 (25.7)	28 (60.9)	80 (32.3)
Primary diagnosis personality disorder
Yes	37 (18.3)	25 (54.3)	62 (25.0)
No	165 (81.7)	21 (45.7)	186 (75.0)
Frequency of self-harm since admission
No history	177 (87.6)	22 (47.8)	199 (80.2)
Once or Twice	7 (3.5)	7 (15.2)	14 (5.6)
Monthly	8 (4.0)	7 (15.2)	15 (6.0)
Weekly	5 (2.5)	10 (21.7)	15 (6.0)
Daily	5 (2.5)	0	5 (2.0)
Frequency of victim of physical, psychological, and/or emotional harm from others since admission
No history	183 (90.6)	38 (82.6)	221 (89.1)
Once or Twice	14 (6.9)	7 (15.2)	21 (8.5)
Monthly	1 (0.5)	0	1 (0.4)
Weekly	3 (1.5)	1 (2.2)	4 (1.6)
Daily	1 (0.5)	0	1 (0.4)
Number of ACEs
Median	3.0	4.0	3.0
Range	10.0	9.0	10.0
Length of Admission
< 1 year	71 (35.1)	21 (45.7)	92 (37.1)
1–2 years	59 (29.2)	11 (23.9)	70 (28.2)
2–4 years	37 (18.3)	8 (17.4)	45 (18.1)
4–7 years	26 (12.9)	6 (13.0)	32 (12.9)
7+ years	9 (4.5)	0	9 (3.6)

Weight difference (in kilograms) over the audit period

Table 3 outlines weight data for the sample in this analysis. Overall weight gain was captured for 165 patients. Mean weight gain was 7.3 kg (SD = 13.8, range = −26.7 − 61.0). Body Mass Index was captured for 214 patients due to missing data. The mean BMI was 33.0 (SD = 7.5, range 16.8 − 58.4), which is within the obesity range.

Table 3.

Weight data for sample.

Weight data
Body weight (kg)
Admission (N = 177)	M = 93.9 SD = 22.6 Range = 53.0 − 188.1
Time of audit (N = 225)	M = 101.8 SD = 25.7 Range = 53.4 − 195.0
Difference (N = 165)	M = 7.3 SD = 13.8 Range = −26.7 − 61.0
Body Mass Index (kg/m²) at time of audit (N = 214)	M = 33.0 SD = 7.5 Range = 16.8 − 58.4

A repeated samples t-test was used to evaluate differences in mean weight in kilograms on admission and at the time of audit. There was a statistically significant increase in weight (7.3 kg, 95% CI [5.256, 10.180]), t (171) = 6.189, p < .001, d = 0.47). Figure 1 (below) illustrates the difference in weight (in kilograms) on admission and at the time of audit.

Figure 1.

Weight difference in kilograms between admission and time of audit.

Predictors of weight change and BMI

Backward multiple regression was run to predict weight difference (in kilograms) between admission and at the time of audit. The predictor variables were sex, length of admission, number of ACEs, primary diagnosis of schizophrenia, primary diagnosis of a personality disorder, self-harm frequency and victim of physical, psychological, and/or emotional abuse frequency. The original model was not significant (see Table 4). The regression was run again with the variable with the largest probability of F being removed. This operation was repeated until the only significant predictors remained, with the final model having the greatest level of significance. A summary of the final model can be found in Table 5. The resulting model only included self-harm frequency and was a significant predictor of 7.3% of the variance in changed body weight (F (1, 121) = 10.585, p < .001, adj. R² = .073). Greater self-harm frequency was a significant predictor of weight change (p <.001).

Table 4.

Full regression model for weight difference between admission and time of audit.

		95% CI for B
Weight Difference (kg)	B	LL	UL	t	SE B	β
Constant	22.136*	5.216	39.057	2.584	8.565
Sex	−9.085	−15.352	−2.818	−2.864	3.172	−0.261
Length of admission	−0.386	−1.743	.971	−0.562	.687	−0.044
Number of ACEs	.110	−0.986	1.206	.199	.555	.017
Primary diagnosis of schizophrenia	.973	−4.162	6.108	.374	2.600	.033
Primary diagnosis of personality disorder	−3.702	−9.383	1.979	−1.287	2.876	−0.115
Self-harm frequency	4.270**	−1.014	9.555	1.596	2.675	.145
Victim of physical, psychological, and emotional abuse frequency	.683	−1.739	3.104	.557	1.226	.045

Note. Model = “Backward” method in SPSS statistics; B = unstandardized regression coefficient; CI = confidence interval; LL = lower limit; UL = upper limit; t = t value; SE B = standard error of the coefficient; β = standardized coefficient. *p < .05. **p < .001.

Table 5.

Final multiple regression model for weight difference between admission and time of audit.

		95% CI for B
Weight Difference (kg)	B	LL	UL	t	SE B	β
Constant	4.894*	2.306	7.483	3.744	1.307
Self-harm frequency	4.075*	1.595	6.555	3.253	1.253	.284

Backward multiple regression was also run to predict patient BMI at the time of audit. The predictor variables were sex, length of admission, number of ACEs, primary diagnosis of schizophrenia, primary diagnosis of a personality disorder, self-harm frequency and victim of physical, psychological, and/or emotional abuse frequency. The original model was not significant (see Table 6). The process of establishing the strongest model in the first regression was repeated here and can be found in Table 7. The resulting model only included number of ACEs and primary diagnosis of schizophrenia and was a significant predictor of 5.6% of the variance in BMI (F (2, 159) = 5.741, p = .014, adj. R² = .056). A greater number of ACEs was a significant predictor of BMI (p = .040), as well as a primary diagnosis of schizophrenia (p = .049).

Table 6.

Full regression model for BMI at time of audit.

		95% CI for B
BMI (kg/m²)	B	LL	UL	t	SE B	β
Constant	31.308*	22.897	39.720	7.339	4.266
Sex	−1.317	−4.317	1.683	−0.866	1.521	−0.067
Length of admission	.090	−0.545	.724	.279	.322	.019
Number of ACEs	.464	−0.074	1.003	1.700	.273	.125
Primary diagnosis of schizophrenia	1.739	−0.741	4.219	1.383	1.258	.108
Primary diagnosis of personality disorder	−0.765	−3.519	1.990	−0.548	1.397	−0.044
Self-harm frequency	1.696	−0.688	4.080	1.403	1.209	.107
Victim of physical, psychological, and emotional abuse frequency	.173	−1.039	1.384	.281	.614	.020

Table 7.

Final multiple regression model for BMI at time of audit.

		95% CI for B
BMI (kg/m²)	B	LL	UL	t	SE B	β
Constant	33.053**	30.159	35.948	22.554	1.465
Number of ACEs	.730*	.066	1.394	2.173	.336	.166
Diagnosis of schizophrenia	−3.203*	−5.655	−0.752	−0.198	1.241	−0.198

p < .05. **p < .001.

Discussion

The current analysis formed part of an audit of quality, performance and clinical outcomes in the care and treatment of Welsh patients detained in secure mental health services. The current aim was to explore the role of trauma on weight gain and obesity in a sample of Welsh patients treated in low, medium, and high secure services across England and Wales. A total of 248 participants were included in this study, with the sample being predominantly males from low secure services with a primary diagnosis of schizophrenia. The sample gained on average 7.7 kg (SD = 13.8) of weight from their admission until the audit period, where the median length of admission was 1—2 years. There was a statistically significant increase in weight, a finding that is consistent with previous studies that have explored secure inpatient weight gain from admission (Davies et al., 2023b; Every-Palmer et al., 2018; Hilton et al., 2015). For example, Hilton et al. (2015) found that secure inpatients gained 8.7 kg of weight over an average of 6 months of treatment. The current analysis also found that patient’s mean BMI was within the obesity range (33.0 kg/m²). This accords with the high prevalence rates of obesity found in Moss et al’s. (2021) review.

This analysis also found that self-harm frequency was a significant predictor of weight gain from admission to the time of audit, and number of ACEs and a primary diagnosis of schizophrenia significantly predicted patient BMI. These findings suggest that similar psychological mechanisms may mediate both self-harm and overeating behavior and are supported by research that shows individuals who engage in self-harm are more likely to display dysregulated eating behavior (Peebles et al., 2011). These psychological mechanisms could include emotional dysregulation, psychological inflexibility, and experiential avoidance. These concepts can be invoked to explain how difficulties managing environmental changes and distressing inner experiences (such as difficult thoughts, images, and emotions), and being unable to deal with problems in a flexible way, may lead to maladaptive escape/avoidance behaviors (Cherry et al., 2021), such as consuming highly palatable food to excess. This is compounded by the obesogenic nature of the secure inpatient setting, with easy access to energy dense foods (Day & Johnson, 2017) and restrictions that service to impact patient motivation to engage in healthier alternatives (Davies et al., 2023a). It has been suggested that individuals who experience psychological trauma engage in experiential avoidance to distract from uncomfortable thoughts and feelings (Bishop et al., 2017; Lee & Bong, 2018; Lewis & Naugle, 2017; Roche et al., 2019; Russell et al., 2020). This is also the case for people with SMI (Castilho et al., 2017; Espejo et al., 2017; Spinhoven et al., 2014; Wheaton & Pinto, 2017).

Experiential avoidance is maintained through negative reinforcement, as the relief of discomfort arising from avoidance increases the likelihood that the behavior will continue (Schaumberg et al., 2016). It can take many forms, including self-harm (Holden et al., 2021) binge-eating and sedentary behavior (Afari et al., 2019; Lillis et al., 2011; 2011). Experiential avoidance may therefore lead to problematic adult behaviors such as the overconsumption of high calorie foods, difficulty in maintaining dietary restrictions (Afari et al., 2019; Schaumberg et al., 2016; Wilkinson et al., 2010), emotional eating (Litwin et al., 2017) and self-harm (Brereton & McGlinchey, 2020).

However, caution should be considered when interpreting the results in this manner, given the current analysis did not measure these psychological mechanisms nor did it capture eating behavior, or the overall dietary environment of the hospitals concerned. To make a more definitive link in this regard, future research should look to explore levels of experiential avoidance and dysregulated eating behavior in the context of the dietary environments of secure inpatient services.

Findings in the current study could alternatively be explained by the weight inducing side-effects of antipsychotic medications (Leucht et al., 2023) that are frequently used within secure services (Farrell & Brink, 2020). Furthermore, patients admitted into secure services frequently have a history of substance misuse (Eagle et al., 2019). Given the biological similarities between use of certain illicit substances and high reward foods (Raghu & Bhat, 2022), it could be argued that weight gain following admission into inpatient settings could be the result of high calorie food replacing addiction to illicit substances, where access to the illicit substance is significantly reduced and, contrariwise, access to high calorie foods within inpatient settings is significantly increased (Day & Johnson, 2017). However, a recent study conducted with Welsh secure inpatients found that neither a history of substance misuse nor high weight gain risk antipsychotic medications were predictive of secure inpatient weight gain during the initial stages of inpatient treatment (Davies et al., 2023b).

Limitations, implications and future research

The current study has several limitations to consider. In particular, the sample in the current study primarily includes Welsh males with a primary diagnosis of schizophrenia. As such, the current findings may not be generalizable to patient outside of the Welsh context or female inpatients. Future research should look to include a less ethno- and gender-centric sample to increase the generalizability of results. This is of particular importance for female inpatients, given the high prevalence of early psychological trauma and the implications this has for obesity. The fact that the current study was retrospective by design also presented a limitation. Data in the current study was collected from patient medical records which may have been inconsistently recorded, impacting the reliability of the study data. For example, only patient weight from admission and at the time of audit, and BMI at the time of audit were collected. Patient height was not collected and so BMI could not be calculated from their earliest weight. Therefore, we could not establish a BMI difference. Future research should look to collect prospective data to the associations between weight gain, obesity, and psychological trauma.

The current study also did not include a history of substance misuse as a predictor variable. Whilst previous research in Welsh secure inpatient settings shows no association between substance misuse history and patient weight gain (Davies et al., 2023b), given the strong association between experiences of psychological trauma in childhood and substance misuse in adulthood (Grummitt et al., 2022) it is possible that a history of substance misuse may mediate the relationship between ACEs and patient weight gain. As such, more focused work should be conducted to explore the relationship between substance misuse history and secure inpatient weight gain.

Despite these limitations, this is the first recorded study that has attempted to explore the association between psychological trauma and weight gain and obesity amongst secure service inpatients. As such, the current findings have important implications for practice. In particular, the findings suggest associations between self-harm and weight gain. This highlights the benefits of promoting holistic care and treatment planning that addresses the mental health and physical health of patients together. Naylor et al. (2016) describe how, despite evidence of the adverse impact that poor physical health can have on long term mental health problems, holistic care and treatment planning remains elusive and should be a target for further research and service development. However, it is important to acknowledge that, given the analytical methods employed, the current findings present only correlations between trauma and obesity. As such we do not recommend inferring causality in this regard.

This notion is further supported by the current findings that early psychological trauma is a significant predictor of obesity. There has been a growing movement into trauma-informed approaches to care within mental health settings with an emphasis on the positive impacts this can have for the patient experience and mental health treatment outcomes (Maguire & Taylor, 2019). We would advocate that a similarly trauma informed approach should be considered for patients’ physical health and wellbeing.

Footnotes

Acknowledgements

We would like to thank the Quality Assurance Improvement Service of the National Collaborative Commissioning Unit for their support with conducting this research. We would also like to extend our sincere gratitude for the patient’s whose data was used in the current study.

Conflict of interest

No potential conflict of interest was reported by the author(s).

References

Afari

, Herbert, M. S., Godfrey, K. M., Cuneo, J. G., Salamat, J. S., Mostoufi, S., Gasperi, M., Ober, K., Backhaus, A., Rutledge, T., & Wetherell, J. L. (2019). Acceptance and commitment therapy as an adjunct to the MOVE! programme: a randomized controlled trial. Obesity Science and Practice, 5(5), 397–407. 10.1002/osp4.356

Afzal

Siddiqi

Ahmad

Afsheen

Aslam

Ali

Ayesha

Bryant

Holt

Khalid

Ishaq

Koly

K. N.

Rajan

Saba

Tirbhowan

, & Zavala

G. A.

(2021). Prevalence of overweight and obesity in people with severe mental illness: Systematic review and meta-analysis. Frontiers in Endocrinology, 12, 769309. 10.3389/fendo.2021.769309

Banks

(2021). Childhood maltreatment and the risk of institutional violence in forensic settings: A systematic review & the relationship between adverse childhood experiences, violence, empathy and psychosis within forensic settings [Doctoral dissertation, University of Edinburgh]. Edinburgh Research Archive.

Bellis

(2016). Adverse childhood experiences and their impact on health-harming behaviours in the Welsh adult population . 10.13140/RG.2.1.4719.1122

Bellis

, Ashton, K., Hughes, K. E., Ford, K., Bishop, J., & Paranjothy, S. (2016). Welsh Adverse Childhood Experiences (ACE) Study; Adverse Childhood Experiences and their impact on health-harming behaviours in the Welsh adult population (ISBN 978-1-910768-23-5). Public Health Wales NHS Trust. 10.13140/RG.2.1.4719.1122

Bishop

L. S.

Ameral

V. E.

, & Palm Reed

K. M.

(2017). The impact of experiential avoidance and event centrality in trauma-related rumination and posttraumatic stress. Behavior Modification, 42(6), 815–837. 10.1177/0145445517747287

Brereton

, & McGlinchey

(2020). Self harm, emotion regulation and experiential avoidance: A systematic review. Archives of Suicide Research: Official Journal of the International Academy for Suicide Research, 24(sup1), 1–24. 10.1080/13811118.2018.1563575

Buckholdt

K. E.

Parra

G. R.

Anestis

M. D.

Lavender

J. M.

Jobe-Shields

L. E.

Tull

M. T.

, & Gratz

K. L.

(2014). Emotion regulation difficulties and maladaptive behaviors: Examination of deliberate self-harm, disordered eating, and substance misuse in two samples. Cognitive Therapy and Research, 39(2), 140–152. 10.1007/s10608-014-9655-3

Castilho

Martins

M. J.

Pinto

A. M.

Viegas

Carvalho

, & Madeira

(2017). Understanding the effect of attachment styles in paranoid ideation: The mediator role of experiential avoidance. Journal of Contextual Behavioral Science, 6(1), 42–46. 10.1016/j.jcbs.2016.11.007

10.

Cherry

K. M.

Hoeven

E. V.

Patterson

T. S.

, & Lumley

M. N.

(2021). Defining and measuring “psychological flexibility”: A narrative scoping review of diverse flexibility and rigidity constructs and perspectives. Clinical Psychology Review, 84, 101973. 10.1016/j.cpr.2021.101973

11.

Davies

J. L.

Bagshaw

Watt

Hewlett

, & Seage

(2023a). Staff perspectives on obesity within a Welsh secure psychiatric inpatient setting. The Journal of Mental Health Training, Education and Practice, 18(1), 44–52. 10.1108/JMHTEP-06-2022-0050

12.

Davies

J. L.

Watt

Bagshaw

Hill

, & Seage

(2023b). Weight gain is not associated with antipsychotic medication, sociodemographic factors, or diagnosis in a Welsh secure mental health unit. International Journal of Forensic Mental Health, 1–10. 10.1080/14999013.2023.2218287

13.

Day

, & Johnson

(2017). Working together to address obesity in adult mental health secure units: A systematic review of the evidence and a summary of the implications for practice (Cm 2016614). Public Health England.

14.

Eagle

, & Sinclair

(2019). Integrated substance use rehabilitation in a secure forensic facility. Journal of Forensic Practice, 21(1), 50–60. 10.1108/JFP-09-2018-0037

15.

Espejo

E. P.

Gorlick

, & Castriotta

(2017). Changes in threat-related cognitions and experiential avoidance in group-based transdiagnostic CBT for anxiety disorders. Journal of Anxiety Disorders, 46, 65–71. 10.1016/j.janxdis.2016.06.006

16.

Every-Palmer

Huthwaite

M. A.

Elmslie

J. L.

Grant

, & Romans

S. E.

(2018). Long-term psychiatric inpatients’ perspectives on weight gain, body satisfaction, diet and physical activity: A mixed methods study. BMC Psychiatry, 18(1), 300. 10.1186/s12888-018-1878-5

17.

Farrell

, & Brink

(2020). The prevalence and factors associated with antipsychotic polypharmacy in a forensic psychiatric sample. Frontiers in Psychiatry, 11, 263. 10.3389/fpsyt.2020.00263

18.

Felitti

Anda

Nordenberg

Williamson

Spitz

Edwards

Koss

, & Marks

(1998). Psychological adjustment in children with episodic migraine: A population-based study. American Journal of Preventive Medicine, 14(4), 245–258. 10.3922/j.psns.2014.1.06

19.

Grummitt

Barrett

Kelly

, & Newton

(2022). An umbrella review of the links between adverse childhood experiences and substance misuse: What, why, and where do we go from here? Substance Abuse and Rehabilitation, 13, 83–100. 10.2147/SAR.S341818

20.

Hall

K. D.

Ayuketah

Brychta

Cai

Cassimatis

Chen

K. Y.

Chung

S. T.

Costa

Courville

Darcey

Fletcher

L. A.

Forde

C. G.

Gharib

A. M.

Guo

Howard

Joseph

P. V.

McGehee

Ouwerkerk

Raisinger

, … Zhou

(2019). Ultra-processed diets cause excess calorie intake and weight gain: An inpatient randomized controlled trial of AD Libitum Food Intake. Cell Metabolism, 30(1), 67–77.e3. 10.1016/j.cmet.2019.05.008

21.

Herzog

J. I.

, & Schmahl

(2018). Adverse childhood experiences and the consequences on neurobiological, psychosocial, and somatic conditions across the lifespan. Frontiers in Psychiatry, 9, 420. 10.3389/fpsyt.2018.00420

22.

Hilton

N. Z.

Ham

Hill

Emmanuel

, & Thege

B. K.

(2022). Predictors of weight gain and metabolic indexes among men admitted to forensic psychiatric hospital. International Journal of Forensic Mental Health, 21(2), 164–174. 10.1080/14999013.2021.1952356

23.

Hilton

N. Z.

Ham

Lang

, & Harris

G. T.

(2015). Weight gain and its correlates among forensic inpatients. Canadian Journal of Psychiatry. Revue Canadienne de Psychiatrie, 60(5), 232–238. 10.1177/070674371506000505

24.

Hjorthøj

Stürup

A. E.

McGrath

, & Nordentoft

(2017). SA57. Life expectancy and years of potential life lost in schizophrenia: A systematic review and meta-analysis. Schizophrenia Bulletin, 43(suppl_1), S133–S134. 10.1093/schbul/sbx023.056

25.

Holden

Stables

Brown

, & Fotiadou

(2021). Adverse childhood experiences and adult self-harm in a female forensic population. BJPsych Bulletin, 46(3), 148–152. 10.1192/bjb.2021.34

26.

Huang

Liu

, & Masum

(2020). Adverse childhood experiences and physical and mental health of adults: Assessing the mediating role of Cumulative Life Course Poverty. American Journal of Health Promotion: AJHP, 35(5), 637–647. 10.1177/0890117120982407

27.

Hughes

Bellis

M. A.

Hardcastle

K. A.

Sethi

Butchart

Mikton

Jones

, & Dunne

M. P.

(2017). The effect of multiple adverse childhood experiences on health: A systematic review and meta-analysis. The Lancet Public Health, 2(8), e356–e366. 10.1016/s2468-2667(17)30118-4

28.

Johnstone

, & Boyle

(2018). The power threat meaning framework: Towards the identification of patterns in emotional distress, unusual experiences and troubled or troubling behaviour, as an alternative to functional psychiatric diagnosis. British Psychological Society.

29.

Kerig

P. K.

(2020). Emotion dysregulation and childhood trauma. In Beauchaine

T. P.

& Crowell

S. E.

(Eds.), The Oxford handbook of emotion dysregulation (pp. 265–283). Oxford University Press.

30.

Lee

E. S.

, & Bong

E. J.

(2018). Impact of parents’ problem drinking on suicidal ideation of their university student children: The multiple mediating effects of childhood trauma, experiential avoidance and depression. Journal of Korean Academy of Nursing, 48(5), 565–577. 10.4040/jkan.2018.48.5.565

31.

Leucht

Schneider‐Thoma

Burschinski

Peter

Wang

Dong

Huhn

Nikolakopoulou

Salanti

, & Davis

J. M.

(2023). Long‐term efficacy of antipsychotic drugs in initially acutely ill adults with schizophrenia: Systematic Review and network meta‐analysis. World Psychiatry: Official Journal of the World Psychiatric Association (WPA), 22(2), 315–324. 10.1002/wps.21089

32.

Lewis

, & Naugle

(2017). Measuring experiential avoidance: Evidence toward multidimensional predictors of trauma sequelae. Behavioral Sciences (Basel, Switzerland), 7(1), 9. 10.3390/bs7010009

33.

Lillis

Hayes

S. C.

, & Levin

M. E.

(2011). Binge eating and weight control: The role of experiential avoidance. Behavior Modification, 35(3), 252–264. 10.1177/0145445510397178

34.

Lillis

Levin

M. E.

, & Hayes

S. C.

(2011). Exploring the relationship between body mass 234 index and health-related quality of life: A pilot study of the impact of weight self-stigma and experiential avoidance. Journal of Health Psychology, 16(5), 722–727. 10.1177/1359105310388321

35.

Litwin

Goldbacher

E. M.

Cardaciotto

L. A.

, & Gambrel

L. E.

(2017). Negative emotions and emotional eating: The mediating role of experiential avoidance. Eating and Weight Disorders: EWD, 22(1), 97–104. 10.1007/s40519-016-0301-9

36.

Maguire

, & Taylor

(2019). A systematic review on implementing education and training on trauma-informed care to nurses in forensic mental health settings. Journal of Forensic Nursing, 15(4), 242–249. 10.1097/jfn.0000000000000262

37.

Matheson

S. L.

Shepherd

A. M.

Pinchbeck

R. M.

Laurens

K. R.

, & Carr

V. J.

(2012). Childhood adversity in schizophrenia: A systematic meta-analysis. Psychological Medicine, 43(2), 225–238. 10.1017/s0033291712000785

38.

Mental Capacity Act . (2005).

39.

Mental Health Act Revised . (2007).

40.

Mental Health Act . (1983).

41.

Micanti

Iasevoli

Cucciniello

Costabile

Loiarro

Pecoraro

Pasanisi

Rossetti

G. L.

, & Galletta

(2016). The relationship between emotional regulation and eating behaviour: A multidimensional analysis of Obesity Psychopathology. Eating and Weight Disorders: EWD, 22(1), 105–115. 10.1007/s40519-016-0275-7

42.

Michel

Kindler

, & Kaess

(2022). Traumatic experiences in childhood and the development of psychosis spectrum disorders. European Child & Adolescent Psychiatry, 31(2), 211–213. 10.1007/s00787-021-01938-9

43.

Mills

, & Davies

J. L.

(2022). Making days count: National review of patients cared for in secure mental health hospitals. The National Collaborative Commissioning Unit. https://nccu.nhs.wales/qais/national-reviews/making-days-count/mdc-documents/making-days-count1/

44.

Moss

Meurk

Steele

M. L.

, & Heffernan

(2021). The physical health and activity of patients under forensic psychiatric care: A scoping review. International Journal of Forensic Mental Health, 21(2), 194–209. 10.1080/14999013.2021.1943570

45.

National Institute of Diabetes and Digestive and Kidney Diseases . (2018). Health risks of overweight and obesity . https://www.niddk.nih.gov/health-information/weight-management/adult-overweight-obesity/health-risks

46.

Naylor

Das

Ross

Honeyman

Thompson

, & Gilburt

(2016). Bringing together mental and physical health: A new frontier for integrated care . The King’s Fund Report, March, 120.

47.

Peebles

, Wilson, J., & Lock, J. (2011). Self-injury in adolescents with eating disorders: correlates and provider bias. Journal of Adolescent Health, 48(3), 310–313. 10.1016/j.jadohealth.2010.06.07

48.

Public Health England . (2018). Severe mental illness (SMI) and physical health inequalities: Briefing . 27^th September 2018. https://www.gov.uk/government/publications/severe-mental-illness-smi-physical-health-inequalities/severe-mental-illness-and-physical-health-inequalities-briefing

49.

Public Health England . (2021). Managing a health weight in adult secure services - practice guidance . https://www.england.nhs.uk/publication/managing-a-healthy-weight-in-adult-secure-services-practice-guidance/

50.

Raghu

S. V.

, & Bhat

(2022). Neurobiology of food addiction. In Future foods: Global trends, opportunities, and sustainability challenges (pp. 425–431). Academic Press. 10.1016/B978-0-323-91001-9.00035-9

51.

Rees

, & Thomson

(2020). Exploration of morbidity, suicide and all-cause mortality in a Scottish forensic cohort over 20 years. BJPsych Open, 6(4), e62. 10.1192/bjo.2020.40

52.

Roche

A. I.

Kroska

E. B.

Miller

M. L.

Kroska

S. K.

, & O'Hara

M. W.

(2019). Childhood trauma and problem behavior: Examining the mediating roles of experiential avoidance and mindfulness processes. Journal of American College Health: J of ACH, 67(1), 17–26. 10.1080/07448481.2018.1455689

53.

Russell

L. H.

Bardeen

J. R.

Clayson

K. A.

Dolan

S. L.

, & Fergus

T. A.

(2020). The closed response style and posttraumatic stress: Examining the interaction between experiential avoidance and cognitive fusion among women experiencing sexual trauma. Psychological Trauma: Theory, Research, Practice and Policy, 12(6), 627–634. 10.1037/tra0000580

54.

Schaumberg

Schumacher

Rosenbaum

Kase

Piers

Lowe

Forman

, & Butryn

(2016). The role of negative reinforcement eating expectancies in the relation between experiential avoidance and disinhibition. Eating Behaviors, 21, 129–134. 10.1016/j.eatbeh.2016.01.003

55.

Schlauch

K. A.

Read

R. W.

Neveux

Lipp

Slonim

, & Grzymski

J. J.

(2022). The impact of aces on BMI: An investigation of the genotype-environment effects of BMI. Frontiers in Genetics, 13, 816660. 10.3389/fgene.2022.816660

56.

Spinhoven

Drost

de Rooij

van Hemert

A. M.

, & Penninx

B. W.

(2014). A longitudinal study of experiential avoidance in emotional disorders. Behavior Therapy, 45(6), 840–850. 10.1016/j.beth.2014.07.001

57.

Varese

Smeets

Drukker

Lieverse

Lataster

Viechtbauer

Read

van Os

, & Bentall

R. P.

(2012). Childhood adversities increase the risk of psychosis: A meta-analysis of patient-control, prospective- and cross-sectional cohort studies. Schizophrenia Bulletin, 38(4), 661–671. 10.1093/schbul/sbs050

58.

Walker

Yates

Dening

Völlm

Tomlin

, & Griffiths

(2023). Quality of life, wellbeing, recovery, and progress for older forensic mental health patients: A qualitative investigation based on the perspectives of patients and staff. International Journal of Qualitative Studies on Health and Well-Being, 18(1), 2202978. 10.1080/17482631.2023.2202978

59.

Wheaton

M. G.

, & Pinto

(2017). The role of experiential avoidance in obsessive–compulsive personality disorder traits. Personality Disorders, 8(4), 383–388. 10.1037/per0000200

60.

Wilkinson

L. L.

, Rowe, A. C., Bishop, R. J., & Brunstrom, J. M. (2010). Attachment anxiety, disinhibited eating, and body mass index in adulthood. International Journal of Obesity, 34(9), 1442–1445. 10.1038/ijo.2010.72