PTSD, complex PTSD and the dissociative subtype of PTSD in patients with eating disorders

Abstract

Introduction

Posttraumatic stress disorder (PTSD) and dissociation are associated with eating disorders (EDs) and serve as markers of higher severity and comorbidity. However, the role of complex PTSD (CPTSD) and the dissociative subtype of PTSD (DPTSD) in EDs remains relatively unexplored.

Methods

Participants were 635 patients (60% female) admitted to a higher level of care (residential, partial hospital or intensive outpatient) for treatment of a DSM-5 defined ED, substance use disorder (SUD), PTSD and/or mood disorder. In addition to a semi-structured interview, patients completed self-report assessments within 72 h of admission, including the Eating Disorder Examination Questionnaire; Life Events Checklist for DSM-5; PTSD Checklist for DSM-5; International Trauma Questionnaire; Dissociative Subtype of PTSD Scale; Alcohol Use Disorder Identification Test; Drug Abuse Screening Test-10; Patient Health Questionnaire-9; Spielberger State Trait Anxiety Scale short form, and the World Health Organization Quality of Life Abbreviated Scale. Patients with (n = 231) and without EDs (n = 404) were compared using multivariant analyses of variance and age, admission BMI, gender, race, sexual orientation and level of care as covariates.

Results

In this highly comorbid group, patients with EDs had significantly higher rates of self-reported types of traumatic events (8 v. 7), provisional PTSD (68% v. 48%), DPTSD (50% v. 27%), and CPTSD (33% v. 19%) compared to those without EDs (p ≤ .01). Patients with DPTSD and/or CPTSD had significantly higher symptom scores for EDs, major depression, substance use, state and trait anxiety, and worse quality of life. In addition, there was significant overlap between the diagnoses of CPTSD and DPTSD with 25% of those with EDs meeting criteria for both disorders.

Conclusions

DPTSD and CPTSD are common in patients with EDs and indicate more complex psychiatric comorbidity associated with high doses of trauma. These results emphasize the importance of thorough assessment procedures and the need for integrated treatment approaches that directly address the full spectrum of trauma-related symptoms.

Plain language summary

A study examined the prevalence of complex PTSD (CPTSD) and the dissociative subtype of PTSD (DPTSD) in patients with eating disorders (EDs) compared to those without EDs. Participants with EDs reported more traumatic events and exhibited significantly higher rates of both DPTSD (50% v. 27%) and CPTSD (33% v. 19%) compared to those without EDs. The presence of DPTSD or CPTSD was associated with more severe ED symptoms, higher rates of depression, substance misuse, anxiety, and a poorer quality of life, indicating the importance of comprehensive trauma assessment and integrated treatment approaches for this population. In addition, 25% of those with EDs met criteria for both DPTSD and CPTSD, indicating substantial overlap between these conditions.

Introduction

It is well established that posttraumatic stress disorder (PTSD) and its symptoms are associated with eating disorders (EDs), especially binge-purge type EDs [8, 9, 36, 70]. Furthermore, PTSD symptomatology has been shown to be a reliable marker of higher severity and comorbidity, especially in patients being treated in higher levels of care [15, 18, 19, 21, 59, 72, 77]. Patients with an ED and comorbid PTSD (ED + PTSD) typically have higher scores on measures of ED symptoms, including the Eating Disorder Examination Questionnaire (EDE-Q) and the Eating Disorders Inventory-2 (EDI-2), as well as symptoms of mood and anxiety disorders, headache, and worse quality of life [13–16, 18, 20], Patients with ED + PTSD have also been shown to have greater recurrence of experiential avoidance, depression and ED symptoms after residential treatment [78]. However, much less is known about complex PTSD (CPTSD) and the newly defined DSM-5 diagnosis of dissociative subtype of PTSD (DPTSD), in relationship to EDs.

The diagnosis of complex PTSD (CPTSD) was first coined by Judith Herman, but it was only recently codified into the 11th version of the International Classification of Disease (ICD-11) by the World Health Organization (WHO) in 2022 [44, 90]. CPTSD reflects traumatic effects in individuals with not only typical PTSD symptoms of reexperiencing, avoidance and hyperarousal, but also symptoms that reflect disturbances in self-organization (DSO) characterized by affective dysregulation, negative self-concept and disturbances in relationships. These personality features suggest a more debilitating disorder than PTSD [61]. CPTSD is estimated to occur in approximately 25–50% of individuals with PTSD and is indicative of more complex trauma histories typically characterized by multiple traumatic events beginning in childhood and continuing into adulthood (complex trauma) [91]. Individuals with early onset EDs beginning in childhood have been reported to have a higher trauma dose compared to those with later ED onsets [13, 20, 22, 60]. Multiple traumatic events and/or types in a person’s life have been associated in non-ED populations with greater severity and symptom complexity [30, 37, 79, 81]. To date, there has been only one report in the literature regarding ICD-11 defined CPTSD and DSM-5 defined EDs. In this study, Day and colleagues reported that CPTSD was commonly found in patients seeking treatment for an ED and that it was associated with even higher severity than typical PTSD [32, 33]. This is in concordance with studies showing that higher doses of childhood maltreatment (including sexual, physical and emotional abuse and physical and emotional neglect) are associated with high ED severity and comorbidity [60]. Other investigators have reported on the unique challenges of treating comorbid EDs and complex PTSD, including a focus on the treatment alliance with an emphasis on emotional-relational processes [31, 64, 73]. Although the diagnosis of CPTSD has been controversial [56, 68, 69, 91], many clinical investigators emphasize its clinical utility [2, 43, 63]. In a major meta-analysis, investigators found that treatment outcome may be moderated by the developmental time of onset of psychological trauma, with childhood trauma being associated with smaller effects of psychological therapies on CPTSD symptoms [52].

In parallel to the evolution of CPTSD in the psychiatric literature, DPTSD was accepted into the DSM-5 in 2013 to describe a subset of PTSD characterized by the additional dissociative symptoms of depersonalization and/or derealization [1, 88]. In a recent systematic review and meta-analysis, DPTSD was found to occur in as many as 48% of patients with DSM-5 defined PTSD, although the authors noted a high degree of heterogeneity across studies [88]. Other studies have noted that DPTSD is associated with greater severity and increased comorbidity when compared to individuals with PTSD without depersonalization and/or derealization [39, 45, 53]. Although DPTSD has been reported in association with substance use disorders (SUDs), to our knowledge there are no known studies of DPTSD in ED samples despite high degrees of association with severe prior traumas, PTSD, dissociative disorders, and dissociation, including depersonalization and depersonalization [12, 15, 18, 19, 21, 34, 41, 58, 66]. Further delineating the effects of trauma in ED patients is likely to increase our understanding of the nature of these symptoms and how to best treat them. In addition, there are no known studies examining the similarities and differences between DPTSD and CPTSD in ED patients, although CPTSD has been associated with high degrees of dissociation [38, 42, 47, 49].

Given the above, we endeavored to examine the interrelationships between DSM-5 PTSD, DSM-5 DPTSD and ICD-11 CPTSD in a mixed group of patients admitting to higher levels of care (residential, partial hospital and intensive outpatient) for the treatment of ED, SUD, mood disorder, and/or other trauma related psychiatric comorbidity, such as PTSD. In comparison to a similar non-ED group of patients also seeking treatment, we hypothesized that ED patients would have1) higher rates of ICD-11 defined CPTSD, and 2) higher rates of DSM-5 defined PTSD and DPTSD. Third, we also hypothesized that among the ED patients alone, the CPTSD and DPTSD subgroups would have higher severity and comorbidity in comparison to those ED patients without these conditions. Finally, we also expected to find significant overlap between CPTSD and DPTSD in ED patients.

Methods

Setting

The study was conducted at SunCloud Health (SCH) programs, a multi-site, multi-level comprehensive treatment program for patients with EDs, addiction related disorders, PTSD and/or mood disorders that includes residential, partial hospital, and intensive outpatient program levels of care. Patients’ conditions are addressed in an integrated, whole-person approach. The transdiagnostic program philosophy emphasizes treating multiple, simultaneous diagnoses, especially where trauma plays a central role [8, 74, 75].

Ethics

This research was approved by the Brany Institutional Review Board (File # 23-12-396-1474), and all participants gave written informed consent for the use of their anonymized assessment results.

Participants

Participants included all adult patients admitted to one of SCH’s programs who agreed to take part in the study. Of 780 patients who were invited to participate, 635 (81.4%) gave written informed consent for the use of their assessment results. The demographic characteristics of the sample are summarized in Table 1, including mean age and admission BMI, sex at birth, gender, sexual orientation, race/ethnicity, and level of care. Specific ED and SUD diagnoses are also listed in Table 1.

Table 1.

Demographic and baseline clinical characteristics of patients with (n = 231) and without (n = 404) an eating disorder (ED) diagnosis using general linear model multivariate analyses with age, admit BMI, gender, race, and sexual orientation as covariates or chi-square (* = p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001)

Variable	ED (mean + SD)	No ED (mean + SD)	t-score	Hedges’ g
Age (years)	30.1 ± 10.4	34.6 ± 12.4	4.918***	11.701
Admit BMI (kg/m2)	29.4 ± 9.9	26.8 ± 6.1	− 3.687**	7.743
	ED (n)	No ED (n)	Chi-square	Cramer’s V
Sex @ Birth			65.499***	0.321
Female	186	193
Male	45	211
Gender identity			57.097***	0.300
Woman	159	175
Man	45	199
Transman	5	5
Transwoman	2	6
Non-binary	18	17
Gender fluid	2	2
Sexual orientation			13.936*	0.148
Heterosexual	110	240
Queer	56	56
Gay or lesbian	11	25
Bisexual	22	36
Asexual	2	3
Pansexual	10	6
Prefer not to say	24	34
Race			8.323	0.114
White	179	294
Black	12	26
Hispanic/Latin	12	38
Middle Eastern	1	2
Asian	11	9
Multi-racial	14	28
Native Hawaiian/	0	1
Pacific Islander	0	1
Prefer not to say	2	6
Level of care			5.285	0.091
Residential	155		236
PHP	68		144
IOP	8		24
Eating disorder diagnoses
AN-R	27
AN-BP	12
AN-unspecified	1
BN	15
BED	54
OSFED	116
UFED	3
ARFID	7
Substance use disorder diagnoses
Alcohol	93		164
Tobacco/nicotine	5		9
Cannabis	96		152
Hallucinogens	6		3
Inhalants	4		0
Opioids	5		10
Hypnotics	2		3
Sedatives	13		10
Stimulants	29		34
Other	16		10
Number of substance use disorder diagnoses
One diagnosis	96		169
Two diagnoses	40		68
Three diagnoses	20		26
4 diagnoses	7		2
5 diagnoses	1		1

AN-BP: anorexia nervosa; binge-purge type; AN-R: anorexia nervosa, restricting type; ARFID: avoidant restrictive food intake disorder; BMI: body mass index; BED: binge eating disorder; BN: bulimia nervosa; IOP: intensive outpatient program; OSFED: other specified feeding and eating disorder; PHP: partial hospital program; SD: standard deviation; SUD: substance use disorder; UFED: unspecified feeding and eating disorder

Of the 635 patients admitted who consented to participate, 231 (36.4%) met DSM-5 criteria for an ED and 404 did not. Demographic characteristics of the study sample by ED diagnosis are shown in Table 1. Patients with EDs were significantly younger, more likely to be female, and had a higher mean BMI than those without an ED. Of those with EDs, their specific DSM-5 diagnoses are also listed in Table 1. Over two-thirds of the patients in the study also met DSM-5 criteria for a SUD. The percentage of subjects with a SUD diagnosis did not significantly vary between those with (71%) and without an ED (66%, chi-square = 5.269, NS).

Measures

The diagnosis of current psychiatric disorders of interest, i.e., EDs, SUDs, mood disorders, PTSD, and other comorbid disorders, were made by admitting psychiatrists using semi‐structured interviews based on DSM‐5 criteria (American Psychiatric Association, [1]). After consent was obtained, the participants’ relevant medical records and self-report questionnaire data were collected onsite within 72 h of admission. The following data points were compiled from the medical record for analysis: age; level of care admitted; specific program admitted; eating disorder diagnosis; SUD diagnosis; PTSD diagnosis; mood disorder diagnosis; other comorbid diagnoses; admission weight; admission height; sex assigned at birth; gender identity; race; and sexual orientation. In addition, the following self-report screening instruments were utilized, all of which have been found to have good reliability and validity:

• The Eating Disorder Examination Questionnaire (EDE-Q) is a 28-item self-report instrument that assesses ED symptomatology during the prior 28 days [62]. The EDEQ has a mean global scale for an overall assessment of ED symptoms, which consists of 4 separate subscales in the domains of restraint, eating concern, weight concern, and shape concern.

• The Life Events Checklist for DSM-5 (LEC-5) is a self-report instrument that assesses for 17 possible PTSD criterion A traumatic events [87]. Patients who endorse a life-threatening event or sexual assault that happened to them and/or was witnessed, together with patient responses on the PTSD Checklist (see below), will qualify for a provisional DSM-5 diagnosis of PTSD.

• The PTSD Checklist for DSM-5 (PCL-5) is a 20-item self-report measure that assesses for DSM-5 cluster B, C, D, and E symptoms of PTSD over the previous month [4]. A total cutoff score of ≥ 33 plus meeting DSM-5 criteria B-E has been used as a reliable indicator of provisional PTSD in other studies which assesses for DSM-5 symptoms of PTSD over the past month [14, 19, 21].

• The International Trauma Questionnaire (ITQ) is an 18-item self-report instrument that assesses for ICD-11-defined symptoms of both PTSD and complex PTSD [29, 48]. Numerous studies have shown support for the factorial and discriminant validity of the ICD-11 diagnoses of both PTSD and complex PTSD [67].

• The Dissociative Subtype of PTSD Scale (DSPS) is a 15-item self-report instrument that assesses for symptoms of DSM-5 PTSD dissociative subtype. Factor analyses support three subscales reflecting derealization/depersonalization, loss of awareness, and psychogenic amnesia [40, 92]. The derealization/depersonalization scale is used to diagnose DPTSD when PTSD criteria are also met, while the other two subscales are indicative of clinically significant dissociative symptoms.

• The Alcohol Use Disorder Identification Test (AUDIT) is a 10-item self-report screen that assesses for symptoms of alcohol use disorder (AUD) [76]. Cut-off scores of 7 and 8 show good sensitivity and specificity, respectively [57].

• The Drug Abuse Screening Test (DAST-10) is a 10-item self-report screening tool that assesses for symptoms of a SUD other than alcohol and tobacco. A cut-off score of 2 shows good sensitivity and specificity, which assesses for symptoms of a SUD other than alcohol [57, 93].

• The World Health Organization Quality of Life abbreviated scale (WHOQOL-BREF) is a 26-item self-report measure of an individual’s quality of life in 4 separate domains: physical health, psychological health, social relations, and environment [89]. Confirmatory factor analysis has shown that the WHOQOL-BREF has good to excellent psychometric properties of reliability and validity [80].

Data analysis

Patients with (n = 231) and without EDs (n = 404) were compared using multivariate analyses of covariance with age, admission BMI, gender, race, and sexual orientation as covariates. Missing data varied between 2–10% of the total sample size depending on the assessment instrument. The Little’s MCAR (missingness completely at random) test was significant (p ≤ 0.001), which indicated that our data were not missing completely at random. We could not identify any participant characteristics associated with missingness. All missing data except diagnoses were imputed using a fully conditional specification method in SPSS 28 [50]. A total of 12 imputed data sets were generated, and the results were pooled using Rubin’s rule (Rubin, 1987). All statistical analyses were performed using either chi-square analyses or general linear model multivariate analyses in SPSS 28 [50]. Effect sizes were calculated as Cramer’s V (for chi-square analyses) and partial Eta squared (ηp²) (for general linear model multivariate analyses). Cramer’s V ranges from 0 to 1 and indicates the association between categorical variables. Threshold values of small, medium, or large effects vary depending on the number of variables measured, but generally < 0.10 to 0.20 indicates a negligible or weak effect size, 0.20 to 0.40 indicates a moderate effect size, and > 0.40 indicates a strong effect size. Partial Eta squared ranges from 0 to 1 with small, medium, or large effects measured as 0.01, 0.06, and 0.14.

In addition to comparing ED v. non-ED patients, and in order to explore the clinical significance of DPTSD and CPTSD, we also performed two other sets of comparisons on the ED group alone: (1) those ED patients with and without a diagnosis of DPTSD, based on a diagnosis of provisional PTSD (as described above using the LEC-5 and the PCL-5) and having a DSPS derealization-depersonalization subscale score ≥ 2; and (2) those ED patients with and without a diagnosis of CPTSD, based on published ITQ criteria. Given the exploratory nature of this study, we did not perform post-hoc corrections for the number of statistical procedures performed in this study, although in an attempt to avoid type I errors, an alpha = 0.01 was used to confer significance. Effect sizes are also provided, which are thought to be a better method than p-values alone of conferring clinical significance [82].

Results

The means (± SD) of the psychometric measures and clinical diagnoses of those with and without EDs are shown in Table 2. Statistics shown are those after covarying for all covariates. The patients with EDs had significantly higher scores on all of the following measures: EDE-Q Global scores, LEC-5 total scores (happened to and witnessed combined); PCL-5 total scores, all DSPS subscale scores, the ITQ- PTSD and DSO dimensional scores, the PHQ-9 total scores, the STAIS-5 state and trait scores, and the WHOQOL-BREF subscale scores. In addition, the ED group had significantly higher rates of provisional DSM-5 PTSD (based on the LEC-5 and the PCL-5) (68% v. 48%), DPTSD (based on the DSPS) (50% v. 27%), provisional ICD-11 PTSD (37% v. 25%) and CPTSD (based on the ITQ) (33% v. 19%) in comparison to the non-ED group (see Table 2). The ED group also had significantly higher rates of major depressive disorder (MDD), state and trait anxiety and more alcohol (AUDIT) and substance use (DAST-10) than the non-ED group.

Table 2.

Clinical characteristics of patients by eating disorder (ED) diagnosis using general linear model multivariate analyses with age, admit BMI, gender, race, and sexual orientation as covariates (* = p < 0.05, ** p < 0.01, ***p < 0.001; Partial Eta squared = ηp²)

Variable	ED (n = 203	No ED (n = 353)	F-score	ηp2
EDE-Q global score	3.4 ± 1.5	1.0 ± 1.1	83.761***	478
LEC-5 total (H + W)	8.0 ± 7.0	7.0 ± 6.0	4.863***	0.050
PCL-5 total	44.0 ± 18.7	35.0 ± 19.2	13.506***	0.129
DSPS
Derealization-depersonal	3.1 ± 2.6	1.9 ± 2.2	26.012***	0.221
Loss of awareness	2.8 ± 1.8	2.1 ± 1.9	13.923***	0.132
Psychogenic Amnesia	1.2 ± 0.9	0.9 ± 0.9	14.410***	0.136
ITQ-PTSD dimension	11.1 ± 6.6	8.2 ± 6.5	13.214***	0.126
ITQ-DSO dimension	15.4 ± 5.6	12.5 ± 6.4	11.187***	0.109
PHQ-9	15.9 ± 6.6	12.9 ± 6.6	4.884***	0.083
STAI-S	2.6 ± 0.8	2.3 ± 0.8	3.700***	0.039
STAI-T	2.9 ± 0.8	2.6 ± 0.9	4.638***	0.048
AUDIT	9.4 ±10.2	7.9 ± 9.9	3.404**	0.036
DAST-10	3.2 ± 3.3	2.5 ± 2.8	10.932***	0.107
WHOQOL-BREF
Physical health	3.0 ± 0.7	3.2 ± 0.8	8.101***	0.081
Psychological health	2.1 ± 0.7	2.6 ± 0.8	13.160***	0.126
Environmental	3.5 ± 0.8	3.6 ± 0.7	1.232	0.013
Social relations	2.8 ± 0.9	2.9 ± 0.9	2.203*	0.024
PTSD diagnosis (DSM-5)	68 ± 0.47	0.48 ± 0.50	9.060***	0.090
DPTSD diagnosis	0.50 ± 0.50	0.27 ± 0.44	15.506***	0.145
PTSD diagnosis (ICD-11)	0.37 + 0.48	0.25 + 0.43	4.429***	0.053
CPTSD diagnosis	0.33 ± 0.48	0.19 ± 0.44	4.250***	0.051
MDD diagnosis	0.80 + 0.40	0.67 + 0.47	4.546***	0.048
Any SUD diagnosis	0.71 + 0.46	0.66 + 0.48	4.533***	0.055

AUDIT: Alcohol Use Disorder Identification Test; CPTSD: complex posttraumatic stress disorder; DAST-10: Drug Abuse Screening Test; Depersonal: depersonalization; DPSD: Dissociative Subtype of PTSD Scale; DSM-5: Diagnostic and Statistical Manual, 5th Edition; DSPTSD: dissociative subtype of PTSD; DSO: disturbances in self-organization; EDE-Q: Eating Disorder Examination Questionnaire; ICD-11, International Classification of Disease, 11th Edition; ITQ: International Trauma Questionnaire; LEC-5: Life Events Checklist for DSM-5; MDD: major depressive disorder; mYFAS 2.0: Modified Yale Food Addiction Scale 2.0; PCL-5: PTSD Checklist for DSM-5; PHQ-9: Patient Health Questionnaire; SUD: substance use disorder; WHOQOL-BREF: World Health Organization Quality of Life Abbreviated Scale

Two sets of secondary analyses were conducted in the ED group alone to compare 1) those with and without CPTSD (Table 3), and 2) those with and without DPTSD (Table 4). In both cases, ED patients with CPTSD and/or DPTSD had statistically significant higher scores on the total number of lifetime trauma types, symptom measures of PTSD (both PCL-5 and ITQ), depersonalization-derealization (DSPS), loss of awareness (DSPS), psychogenic amnesia (DSPS), major depression (PHQ-9), state anxiety (STAIS-5), trait anxiety (STAIT-5), EDE-Q global scores, substance use (DAST-10), as well as worse quality of life (WHOQOL-BREF) in the psychological and physical health domains. In addition, ED patients with DPTSD (but not those with CPTSD) had significantly higher scores on the AUDIT. Despite the differences in p-values, all these differences had either medium (ηp² ≥ 0.06 < 0.14) to large effect sizes (ηp² ≥ 0.14), respectively.

Table 3.

Clinical characteristics of eating disorder patients alone by CPTSD diagnosis using general linear model multivariate analyses with age, admit BMI, gender, race, and sexual orientation as covariates (* = p < 0.05, ** p < 0.01, ***p < 0.001; Partial Eta squared = ηp²)

Variable	CPTSD (n = 66)	No CPTSD (n = 137)	F-score	ηp2
EDE-Q global score	3.8 ± 1.5	3.2 ± 1.5	3.105**	0.087
LEC-5 total (H + W)	10.0 ± 7.3	7.0 ± 6.4	4.336***	0.117
PCL-5 total	57.6 ± 10.2	36.8 ± 18.4	16.489***	0.335
DSPS
Derealization-depersonal	4.0 ± 2.4	2.7 ± 2.5	10.499***	0.243
Loss of awareness	3.4 ± 1.8	2.4 ± 1.7	6.863***	0.174
Psychogenic amnesia	1.6 ± 0.7	1.0 ± 0.9	5.708***	0.149
ITQ-PTSD dimension	16.8 ± 3.6	8.3 ± 5.9	22.223***	0.405
ITQ-DSO dimension	19.2 ± 2.8	13.5 ± 5.9	10.382***	0.241
PHQ-9	18.2 ± 5.6	14.6 ± 6.8	3.333**	0.092
STAI-S	2.8 ± 0.7	2.4 ± 0.8	2.899*	0.082
STAI-T	3.1 ± 0.7	2.7 ± 0.9	3.235**	0.090
AUDIT	9.3 ± 9.5	8.1 ± 9.6	0.899	0.027
DAST-10	2.7 ± 3.4	4.3 ± 3.1	5.330***	0.140
WHOQOL-BREF
Physical health	2.7 ± 0.6	3.2 ± 0.8	5.388***	0.141
Psychological health	1.8 ± 0.5	2.3 ± 0.7	5.054***	0.134
Environmental	3.1 ± 0.7	3.6 ± 0.7	3.711**	0.102
Social relations	2.5 ± 0.8	2.9 ± 0.95	2.135	0.061
DPTSD diagnosis	0.74 ± 0.44	0.37 ± 0.49	8.762***	0.212
MDD diagnosis	0.89 + 0.30	0.75 + 0.43	1.72	0.050
Any SUD diagnosis	0.78 + 0.42	0.68 + 0.47	1.521	0.049

AUDIT: Alcohol Use Disorder Identification Test; CPTSD: complex posttraumatic stress disorder; DAST-10: Drug Abuse Screening Test; Depersonal: depersonalization; DPSD: Dissociative Subtype of PTSD Scale; DSPTSD: dissociative subtype of PTSD; DSO: disturbances in self-organization; EDE-Q: Eating Disorder Examination Questionnaire; ITQ: International Trauma Questionnaire; LEC-5: Life Events Checklist for DSM-5; MDD: major depressive disorder; mYFAS 2.0: Modified Yale Food Addiction Scale 2.0; PCL-5: PTSD Checklist for DSM-5; PHQ-9: Patient Health Questionnaire; SUD: substance use disorder; WHOQOL-BREF: World Health Organization Quality of Life Abbreviated Scale

Table 4.

Clinical characteristics of eating disorder patients alone by DPTSD diagnosis using general linear model multivariate analyses with age, admit BMI, gender, race, and sexual orientation as covariates (* = p < 0.05, ** p < 0.01, ***p < 0.001; Partial Eta squared = ηp²)

Variable	DPTSD (n = 100)	No DPTSD (n = 102)	F-score	ηp2
EDE-Q global score	3.7 ± 1.4	3.0 ± 1.5	3.358**	0.093
LEC-5 total (H + W)	9.9 ± 7.9	6.3 ± 5.4	4.181***	0.113
PCL-5 total	54.9 ± 10.3	32.9 ± 18.6	19.284***	0.371
DSPS
Derealization-Depersonal	4.8 + 1.7	1.4 + 2.1	37.792***	0.536
Loss of Awareness	3.8 + 1.5	1.7 + 1.5	19.046***	0.368
Psychogenic Amnesia	1.6 + 0.7	0.9 + 0.9	7.359***	0.184
ITQ-PTSD Dimension	14.7 + 4.8	7.7 + 6.3	14.121***	0.302
ITQ-DSO Dimension	17.7 + 4.1	13.2 + 6.1	6.757**	0.171
PHQ-9	18.5 ± 5.5	13.3 ± 6.5	7.059***	0.178
STAI-S	2.8 ± 0.7	2.3 ± 0.8	5.223***	0.138
STAI-T	3.1 ± 0.8	2.6 ± 0.9	3.708**	0.102
AUDIT	10.0 ± 10.4	7.0 ± 8.6	2.718**	0.106
DAST-10	3.8 ± 3.5	2.6 ± 2.9	4.507***	0.121
WHOQOL-BREF
Physical health	2.7 ± 0.6	3.3 ± 0.8	8.588***	0.208
Psychological health	1.9 ± 0.5	2.4 ± 0.1	5.075***	0.134
Environmental	3.3 ± 0.7	3.7 ± 0.7	3.151**	0.088
Social relations	2.7 ± 0.9	2.8 ± 1.0	1.385	0.041
CPTSD diagnosis	0.51 ± 0.50	0.24 ± 0.43	3.246**	0.124
MDD diagnosis	0.90 ± 0.30	0.71 + 0.46	2.672*	0.076
Any SUD diagnosis	0.74 + 0.44	0.71 + 0.46	1.273	0.044

AUDIT: Alcohol Use Disorder Identification Test; CPTSD: complex posttraumatic stress disorder; DAST-10: Drug Abuse Screening Test; Depersonal: depersonalization; DPSD: Dissociative Subtype of PTSD Scale; DSPTSD: dissociative subtype of PTSD; DSO: disturbances in self-organization; EDE-Q: Eating Disorder Examination Questionnaire; ITQ: International Trauma Questionnaire; LEC-5: Life Events Checklist for DSM-5; MDD: major depressive disorder; mYFAS 2.0: Modified Yale Food Addiction Scale 2.0; PCL-5: PTSD Checklist for DSM-5; PHQ-9: Patient Health Questionnaire; SUD: substance use disorder; WHOQOL-BREF: World Health Organization Quality of Life Abbreviated Scale

Lastly, 67% of ED patients with CPTSD also met criteria for DPTSD (Table 3) (with large effect size), and conversely, 51% of ED patients with DPTSD also met criteria for CPTSD (with large effect size). Looking at the ED group as a whole, 25% of the ED patients alone met criteria for both CPTSD and DPTSD, while 39% met criteria for neither one (p ≤ 0.001, chi-square = 16.711, Cramer’s V = 0.287). In the non-ED group, 15% met both sets of criteria, while 61% met neither one (p ≤ 0.001, chi-square = 54.368, Cramer’s V = 0.391).

Discussion

Our first two main hypotheses were supported by the results, i.e., that rates of both CPTSD and DPTSD were significantly higher in patients with EDs compared to those without EDs. Patients with EDs had significantly higher rates of provisional DSM-5 PTSD (68%) and DPTSD (50%), as well as ICD-11 PTSD (37%) and CPTSD (33%) compared to those without EDs. Notably, the effect sizes of these findings were large for DPTSD, medium for DSM-5 PTSD, and small to medium for ICD-11 PTSD and CPTSD.

Our results are similar to the findings by Day and colleagues, who reported that 28.4% of a group of 212 treatment seeking ED patients met ITQ criteria for CPTSD [32, 33]. In their study, those seeking treatment in residential and partial hospital programs had even higher rates of CPTSD. In line with their findings, we also found higher severity and greater impairment in multiple domains in our ED patients with CPTSD compared to those without CPTSD. This included significantly greater traumatization, severity and complexity of symptoms, as well as functional impairment as evidenced by higher levels of ED symptoms, major depression, state and trait anxiety, dissociation, PTSD, substance abuse, and worse quality of life ratings, particularly in physical and psychological health domains. As noted in the Day et al. report, prior studies in the literature reporting high rates of PTSD and associated comorbidity have likely encapsulated ICD-11 CPTSD cases, although it was not measured.

Our data are also notable for demonstrating that those with CPTSD had significantly greater dissociative symptoms as well as DPTSD. This included endorsement of loss of awareness and psychogenic amnesia, which has been reported in ED patients previously [12, 15, 18]. Other investigators have noted the high degree of dissociative symptoms in individuals with CPTSD and that it may be an essential component [38, 42, 47, 49, 65].

Our second main hypothesis was also confirmed in that ED patients endorsed significantly higher rates of DPTSD compared to the non-ED group. To our knowledge, this is the first study to measure rates of DPTSD in patients with EDs in contrast to those without. Fifty percent of patients with EDs met the DSM-5 based DSPS criteria for DPTSD, thereby establishing not only very high rates of PTSD in this group but also the additional symptoms of derealization and/or depersonalization. In addition, the 68% rate of provisional PTSD based on DSM-5 criteria in our study is significantly higher than the rates of approximately 50% previously reported from other residential programs [19, 21, 72, 77]. This may be because of the greater proportion of patients with concurrent SUDs, which tends to be a highly traumatized group, and the lower percentage of patients with anorexia nervosa, restricting type, which tends to be a less traumatized group [6, 7, 9, 11, 17, 60].

Our findings also confirm that ED patients with DPTSD are more symptomatic, more complex and have higher severity of a variety of comorbid symptoms. Like ED patients with CPTSD, ED patients with DPTSD had significantly greater traumatization, severity and complexity of symptoms, as well as functional impairment. This was evidenced by greater symptoms of EDs, major depression, state and trait anxiety, dissociation, PTSD, substance misuse, and worse quality of life ratings. As noted in regard to CPTSD, prior studies reporting high rates of PTSD and associated comorbidity in treatment seeking ED patients have likely encapsulated DPTSD cases, although it was not measured. As a result, dissociative symptoms, and perhaps overt dissociative disorders, are likely to have been missed and untreated. The presence of dissociative symptoms has been noted to significantly influence treatment approaches and PTSD outcomes [26, 28, 52]. For example, when using cognitive processing therapy (CPT), a promising treatment for ED patients with PTSD [10, 14, 83–85], individuals with significant dissociative symptoms have better outcomes when a trauma narrative is included [68, 69]. Other treatment studies of complex trauma and PTSD using the Skills Training in Affective and Interpersonal Regulation (STAIR) approach show efficacy when incorporating the use of narrative therapy [26–28, 55].

Our study is the only study to date to examine the overlap between CPTSD and DPTSD using standardized measures in a clinical population of EDs, SUDs, PTSD, and mood disorders. Given that we found that 25% of ED patients met criteria for both disorders, it is evident that each of these conditions significantly overlaps with the other, and that further research is needed to delineate and explore the implications of our findings. Nevertheless, it is safe to say that both CPTSD and DPTSD each are likely to make treatment much more challenging, i.e., they may increase treatment refractoriness and negatively impact treatment engagement. Studies indicate that individuals with CPTSD may have higher dropout rates and manifest lower subjective improvement following standard trauma-focused psychotherapies compared to those with PTSD, particularly in those with histories of childhood maltreatment [5, 46, 71]. Given this, clinicians increasingly advocate for modular or phase-based approaches rather than standard trauma-focused protocols for CPTSD, given its complex clinical presentation [35, 51]. Similar findings have been reported in patients with PTSD and dissociation, although addressing dissociation early in treatment can be beneficial [54, 86].

Our study has several notable limitations, which include a limited sample size, a non-representative sample of treatment-seeking individuals, and an uneven gender distribution. In addition, most of our ED patients had OSFED (primarily atypical anorexia nervosa) or BED and only a smaller percentage had AN or BN. In addition, there were high rates of SUDs in our ED sample, which could have skewed the results in favor of greater psychopathology. DPTSD has been reported to be very common in populations with SUDs [53]. As such, similar findings may not be the same in other ED treatment settings, and future studies are needed in other ED treatment programs to confirm these results.

In summary, we found high rates of both CPTSD and DPTSD in a group of ED patients seeking treatment in a higher level of care in comparison to a non-ED group with similar rates of SUDs. Both CPTSD and DPTSD were associated with markedly greater severity, complexity and impairment that pose major treatment challenges to patients, clinicians and programs alike. Given how common PTSD, CPTSD and DPTSD occur in conjunction with EDs, especially in higher levels of care, there is a dire need for more treatment focused research as well as the development of practice guidelines for these co-occurring conditions. Principles and strategies toward the development of such guidelines that have been borrowed from several relevant evidence-based approaches have been recently proposed [10]. Meanwhile, integrated, trauma-informed treatment plans can be successfully employed to relieve suffering and symptoms [3, 8, 14, 15, 15, 18, 18, 23–26, 55, 64, 83, 84].

Author contributions

TDB and KD conceptualized and designed the study. TDB wrote the initial draft of the manuscript. CN, NB and KD performed data acquisition. TDB and CN performed statistical analyses of the data and prepared the tables. All authors participated in data checking and the interpretation of results. All authors reviewed the manuscript. All authors approved the submitted version of the manuscript.

Funding

Not applicable.

Data availability

Data are proprietary and not shared.

Declarations

Permission to reproduce material from other sources

Not applicable.

Ethics approval and consent to participate

This research was approved by the Brany Institutional Review Board (File # 23-12-396-1474).

Consent for publication

All patients gave written informed consent for the use of their anonymized data.

Competing interests

TDB is a paid consultant to SunCloud Health. CN, NB and KD are all full-time employees of SunCloud Health.