Assessing transgenerational trauma transmission: development and psychometric properties of the Historical Intergenerational Trauma Transmission Questionnaire (HITT-Q)

ABSTRACT

Objective/Background: Despite increasing attention on transgenerational trauma, currently no comprehensive model and measure exists to be applied on various populations. This study represents the first step in the validation of such a model and a related scale. The Historical Intergenerational Trauma Questionnaire (HITT-Q) assesses family and offspring self-reported vulnerability and resilience, as well as offspring historical moral injury and current levels of insidious trauma.

Method: We developed the HITT-Q based on the cross-population model (HITT model; [Starrs, C. & Békés, V. (2024). Historical and transgenerational trauma: A conceptual framework. Traumatology. In Press]) which incorporates key findings in existing population specific studies. For initial validation of the model and its measurement, Holocaust survivors’ offspring (N = 1104) completed the HITT-Q, measures of current mental health symptoms (PTSD, C-PTSD, anxiety, and depression), and a resilience scale.

Results: In line with the HITT model, confirmatory factor analyses supported a 12-factor solution with the following factors under theorized dimensions: I. Family Vulnerability: (1) Dysregulated and Trauma-related Communication; (2)Trauma-influenced Parenting, (3) Fear; (4) Distress; II. (5) Family Resilience, III. Offspring Vulnerability: (6) Escape; (7) Heightened Responsibility; (8) Trauma-related distress; IV. Offspring Resilience: (9) Coping; (10) Belonging; (11) Values; V. (12) Historical Moral injury. The 12-factor model showed acceptable to good internal validity, and comparison with an existing measure of transgenerational Holocaust trauma indicated good concurrent validity. Finally, the HITT-Q demonstrated predictive validity for mental health symptoms and current resilience.

Conclusions: The current study represents the first step in validating the HITT-Q as a comprehensive measure of historical intergenerational vulnerability and resilience. Our findings provide strong support for the underlying model, and suggest that the HITT-Q represents a valuable scale for both research and historical trauma-informed care.

HIGHLIGHTS

• The papers provides support for the underlying model of historical and transgenerational trauma.

• Findings showed that the Historical Intergenerational Trauma Questionnaire (HITT-Q) has 12 factors, and that it has good psychometric qualities, including internal, concurrent, and predictive validity.

• The Historical Intergenerational Trauma Questionnaire (HITT-Q) represent a valuable scale for both research and historical trauma-informed care.

Objetivo/Antecedentes: A pesar de la creciente atención al trauma transgeneracional, actualmente no existe ningún modelo ni medida integral que se pueda aplicar en diversas poblaciones. Este estudio representa el primer paso en la validación de dicho modelo y una escala relacionada. El Cuestionario de Trauma Intergeneracional Histórico (HITT-Q en su sigla en inglés) evalúa la vulnerabilidad y resiliencia auto informadas por la familia y los hijos, así como el daño moral histórico de los hijos y los niveles actuales de trauma insidioso.

Método: Desarrollamos el HITT-Q basado en el modelo interpoblacional (modelo HITT en su sigla en inglés; Starrs & Békés, 2024) que incorpora hallazgos clave en estudios poblacionales específicos existentes. Para la validación inicial del modelo y su medición, los descendientes de los sobrevivientes del Holocausto (N = 1104) completaron el HITT-Q, medidas de síntomas de salud mental actuales (TEPT, TEPT-C, ansiedad y depresión) y una escala de resiliencia.

Resultados: De acuerdo con el modelo HITT, los análisis factoriales confirmatorios respaldaron una solución de 12 factores con los siguientes factores bajo dimensiones teorizadas: I. Vulnerabilidad familiar: (1) Comunicación desregulada y relacionada con el trauma; (2) Crianza influenciada por el trauma, (3) Miedo; (4) Estrés; II. (5) Resiliencia Familiar, III. Vulnerabilidad de la descendencia: (6) Escape; (7) Responsabilidad Aumentada; (8) Malestar relacionado con el trauma; IV. Resiliencia de la descendencia: (9) Afrontamiento; (10) Pertenencia; (11) Valores; V. (12) Daño moral histórico. El modelo de 12 factores mostró una validez interna de aceptable a buena, y la comparación con una medida existente del trauma transgeneracional del Holocausto indicó una buena validez concurrente. Finalmente, el HITT-Q demostró validez predictiva para los síntomas de salud mental y la resiliencia actual.

Conclusiones: El presente estudio representa el primer paso para validar el HITT-Q como una medida integral de la vulnerabilidad y resiliencia intergeneracional histórica. Nuestros hallazgos brindan un fuerte apoyo al modelo subyacente y sugieren que el HITT-Q representa una escala valiosa tanto para la investigación como para la atención basada en el trauma histórico.

KEYWORDS:

• Transgenerational trauma
• historical trauma
• holocaust
• resilience
• vulnerability

PALABRAS CLAVE:

• Trauma transgeneracional
• trauma histórico
• Holocausto
• resiliencia
• vulnerabilidad

The transgenerational impact of trauma was first systematically studied in the children of Holocaust survivors (HS) (Danieli, 1998), and subsequently in a wider range of diverse populations and collective events. For example, transgenerational impacts have been reported for the Holodomor famine in Ukraine (Bezo & Maggi, 2015), the Cambodian genocide by the Khmer Rouge Regime (Burchert et al., 2017), and the Tutsi genocide in Rwanda (Perroud et al., 2014); additionally intergenerational impacts have been described for various terrorist attacks (e.g. Garfin et al., 2015) and in those seeking asylum (van Ee et al., 2012). Over the last two decades, there has been a growing understanding that in contrast to one person single event trauma, collective multi-event trauma (e.g. genocide, war) includes complex socio-cultural-historical processes (displacement, exploitation, colonization, forced assimilation, cultural/physical genocide), with ramifications stretching across several generations (Danieli, 1998; Degruy-Leary, 2017; Duran et al., 1998). This ongoing transgenerational traumatization is typically labelled Historical Trauma (HT), and it has recently received increasing research attention, especially in Indigenous Peoples living in the Americas.

1. Measurement issues

Despite the importance of assessing historical trauma processes, currently there are no valid and reliable self-report measures that can be applied across various populations. The few existing measures are population specific, including the Danieli Inventory of Multigenerational Legacies of Trauma Scale for Holocaust survivors (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015), the Historical Loss Scale and the Historical Loss Associated Symptoms Scale for Indigenous Americans (Whitbeck et al., 2004), and the African American Historical Trauma Questionnaire for African Americans (Williams-Washington & Mills, 2018). The empirical literature has taken a population specific approach, in part due to the complexity of the concept, as well as a well-founded desire to be culturally responsive. However, there is a need, especially for clinicians and researchers who work with diverse populations, for a more general HT measure that can be utilized across cultural groups. A second important lacuna of current HT assessments is a singular focus on identifying risk factors in the transgenerational trauma transmission process. However, in order to prevent or diminish the negative impact of trauma in the next generation, it is equally important to consider potential associated resilience factors (e.g. positive coping skills), that facilitate a transition from a state of survival, a reactive state, to one of survivance, a proactive state (Wilbur & Gone, 2023), both in families and offspring. Resilience reflects a process over time (Bonanno et al., 2004), which can be best viewed from a dynamic, multisystem perspective where diverse factors at play, leading to variation in responses in groups and individuals exposed to mass-level trauma (Masten, 2011, 2021). The systems involved in resilience constantly interact and change over time, thus, resilience can be defined as ‘the dynamic capacity of a complex adaptive system to respond successfully to challenges that threaten the function, survival, or development of the system’ (Masten et al., 2021, p. 155). The criteria for successful adaptation varies according to the context and can be manifested in a range of variables and processes including intra-individual and social factors (Masten et al., 2021). Accordingly, the definition and operationalization of resilience needs to be contextualized and adapted to different fields and studied populations (Aburn et al., 2020). Therefore, a more comprehensive model of HT must consider both vulnerability and resilience factors. Thirdly, historical traumas typically carry a sense of being betrayed and not protected by political forces and institutions, which has lacked research attention so far. There is ample evidence for this sense of moral betrayal in various traumatized populations (e.g. Kidwell & Kerig, 2021; Litz et al., 2009; Nickerson et al., 2018), however, to our knowledge it has not been specifically included in current models of HT. Finally, HT exists within a current context that often includes ongoing discrimination and oppression (e.g. racism against migrants), as such, although this may not be a direct part of the main HT process, it is, none the less, a crucial factor to contextualizing historical trauma, and as such we have included it in the HITT-Q scale.

Thus, there is a need for a comprehensive model and measure which covers multiple aspects of historical intergenerational trauma that can be applied in individuals from various populations, and which provides clinicians and researchers with a reliable and valid measurement tool.

2. The Historical Intergenerational Trauma Transmission Model

The Historical Intergenerational Trauma Transmission model offers a comprehensive cross-population model (HITT model, Starrs & Békés, 2024), which includes not only vulnerability but also resilience factors at the family and offspring levels. The main model includes twelve components, organized into five higher-order dimensions, based on mechanisms of trauma transmission supported in existing population specific studies, see Table 1 and Figure 1. Furthermore, current Insidious Trauma (oppression and discrimination) is modelled as an independent contextual contributor to offspring mental health symptoms.

Figure 1. The HITT model depicts the complex process of historical intergenerational trauma transmission. Survivors of historical traumatic event(s) are posited to display patterns of trauma-impacted behaviours that constitute vulnerability, as well as other behaviours that increase resilience in their descendants. In turn, offspring respond with both adaptive and maladaptive behaviours that impact their mental health and functioning. This process of historical trauma transmission happens in the context of historical moral injury and is impacted insidious trauma experiences.

Table 1. Dimensions of the Historical and Intergenerational Trauma Transmission Model (HITT Model).

Family Vulnerability	Offspring Vulnerability	Family Resilience	Offspring Resilience	Historical Moral Injury
Dysregulated Communication	Escape	Developmental Sensitivity	Belonging	Betrayed Personal Morals
Trauma-Related Communication	Heightened Responsibility	Emotional Safety	Moral Values	Diminished Trust
Trauma-Impacted Parenting	Trauma-related Distress		Positive Coping
Fear & Distress

Notes: Given that Insidious Trauma is conceptualized as a current factor and not a historical trauma transmission mechanism per se, it is not considered a main dimension in the model.

2.1. Aims

The primary aim of the present study was to develop a self-report measure of historical intergenerational trauma, which allows for the targeted assessment of concerns specific to historical intergenerational trauma, beyond the scope of general mental health measures. The measure builds on the previously developed HITT model, and which can be applied in a variety of populations. As a first step, in the current paper, we sought to validate the HITT-Q’s psychometric properties, specifically, its factor structure, reliability, and validity in a sample of Holocaust survivor offspring.

Thus, we aimed to (1) establish how the factor structure of the HITT model fits the data through confirmatory factor analyses; (2) test the internal consistency and concurrent validity of the instrument by comparing the HITT-Q scores to the Danieli Inventory of Multigenerational Legacies of Trauma (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015), (3) test the HITT-Q’s predictive validity, by assessing whether its factors predict mental health and resilience outcomes, and (4) test whether Insidious Trauma predicts mental health symptoms.

We expected (1) to find a parsimonious model that fits our data, and, (2) to find good to high internal consistency for the HITT-Q factors, good concurrent validity with positive associations between the HITT-Q’s family vulnerability factors and the Danieli Inventory’s three parenting styles, and the HITT-Q’s offspring vulnerability factors with the Danieli Inventory’s adaptational impact on offspring subscale. Moreover, (3) we also expected to demonstrate the HITT-Q’s predictive validity with HITT-Q family vulnerability factors, offspring vulnerability factors, and Historical Moral Injury factor predicting mental health (PTSD, C-PTSD, anxiety, and depression) symptoms, and HITT-Q family resilience factors and offspring resilience factors predicting perceived coping ability in the present as measured by the Connor-Davidson Resilience Scale (CD-RISC; Campbell-Sills & Stein, 2007) as indicative of positive current outcome. Finally, (4) we expected Insidious Trauma to predict higher levels of mental health symptoms.

3. Methods

3.1. Procedure

Participants were recruited via social media, professional networks, and email lists in Hungary between 21 March and 29 April 2021. Interested participants were directed to an online platform hosted by Qualtrics, with additional information about the study. Individuals were eligible to participate if they had at least one Holocaust survivor parent who they grew up with (G2), or if they had a parent who had a Holocaust survivor parent, that their parent grew up (G3). We added the criterion that they or their parent must have been raised by a HS, in order to be able to observe the specific impact of parenting style. After providing consent, participants completed demographic data and standardized measures in an online survey that took approximately 30 min. The study was approved by the Yeshiva University’s Institutional Review Board.

3.2. Participants

Altogether 1,104 participants started the survey and completed at least one of the subscales from the HITT-Q. From this, 724 were G2 participants, and 380 were G3 participants. In line with existing population measures, for this validation study, we focused on the sample of direct descendants (G2 participants). Participants were 62 years old on average (SD:11.0). Most were female (70.9%), had a college diploma (43.0%), and were in a relationship (62.6%). For detailed demographics information see Table 2. Regarding missing data, all participants completed the HITT-Q measure, however 42 participants did not complete any of the symptom scales or the Danieli inventory. Independent samples t-test showed that these 42 participants had lower HITT-Q scores on Factor 9 (Offspring Coping t(722) = 2.42, p = .016) and Factor 11 (Offspring Values t(722) = 2.38, p = .017), but comparable scores on all the other factors.

Table 2. Participant descriptive characteristics by generation.

	N (%)
Mean Age (SD)	61.9 (11.0)
Gender
Female	513 (70.9)
Male	211 (29.1)
Education
Less than high school	1 (.1)
High school graduate	84 (11.6)
Technical/trade school	30 (4.1)
Some college or professional training	201 (27.8)
College diploma	311 (43.0)
Postgraduate degree	97 (13.4)
Work
Works full or part time	359 (49.5)
Does not work (retired, disability, unemployed)	364 (50.3)
Student	1 (.1)
Socioeconomic Status
Difficult, often struggle to afford basic necessities	54 (7.5)
Enough for necessities, but no extras	283 (39.1)
Relatively comfortable, enough for necessities & some extras	306 (42.3)
Comfortable, can afford luxuries	81 (11.2)
Relationship Status
In a relationship (married, partner, other)	453 (62.6)
Not in a current relationship	271 (37.4)
Survivors in the Family
Mother or Father	286 (39.5)
Both parents	397 (54.8)
Grandparents	522 (72.1)

On average, participants reported 3.62 (SD: 2.00, range: 1–12) traumatic historical events and 4.02 (SD: 2.74, range: 1–15) historical losses. See Table 3 for details.

Table 3. Holocaust-related experiences by generation.

	N (%)
Historical Events
Experienced torture	93 (12.8)
Fled to save their lives	322 (44.5)
Taken to a labor camp	405 (55.9)
Deported	420 (58.0)
Hid from the enemy	288 (39.8)
Experienced violence (e.g., beaten, starved) by an authority	93 (12.8)
Forced to leave their home	489 (67.5)
Forced to attend mainstream/residential school	47 (6.5)
Forcibly relocated	25 (3.5)
Punished for speaking their traditional language	1 (.1)
Forced to change their physical appearance (e.g., head shaved)	59 (8.1)
Forced into unwanted medical procedures (e.g., sterilization)	18 (2.5)
Threatened to be killed	115 (15.9)
Forcibly removed from their parents/caregivers	148 (20.4)
Denied the right to practice their religion/traditional practices	63 (8.7)
I don't know what happened to my family/people	34 (4.7)
Historical Losses
Land and/or possessions	341 (47.1)
Traditional/original language	24 (3.3)
Traditional religious/spiritual ways	168 (23.2)
Family ties	354 (48.9)
Family/community members	593 (81.9)
Traditional holidays or celebrations	171 (23.6)
Traditional dress and/or hairstyles	28 (3.9)
Respect for ourselves and each other	73 (10.1)
Traditional social structure (e.g., matrilineal structure)	47 (6.5)
Sense of our own identity	228 (31.5)
Sense of purpose as a community	97 (13.4)
Traditional values	133 (18.4)
Sense of community identity	158 (21.8)
Sense of community belonging	220 (30.4)
Sense of continuity throughout generations	277 (38.3)

Note: Participants endorsed all that apply.

3.3. Measures

3.3.1 Historical Intergenerational Trauma Questionnaire (HITT-Q)

The HITT-Q was developed based on the HITT model (Starrs & Békés, 2024). The HITT-Q is a self-report measure that assesses the impact of historical intergenerational trauma on family and offspring vulnerability and resilience; its items were partly adapted from population-specific measures (Carter et al., 2013; Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015; Torres-Harding et al., 2012; Whitbeck et al., 2004; Williams-Washington & Mills, 2018), and partly developed by the authors based on key findings in the literature. Preliminary questions identify traumatic events (e.g. torture, labour camp, deportation), and historical losses including various physical, emotional, and cultural losses (e.g. losing family members, traditions, language) based on common experiences reported by survivors of collective traumatic events; participants select all relevant items (yes/no). These initial questions assess the exposure to HT, providing qualitative information for clinicians using the scale in the context of trauma-informed care. The HITT-Q assesses mechanisms of trauma transmission utilizing 65-items organized into 12 components across five higher-order dimensions: (1) Family Vulnerability: dysregulated communication (e.g. ‘Open communication seemed not to exist in our home’), trauma communication, (e.g. ‘My family never discussed the traumatic events’), trauma-impacted parenting (e.g. ‘Weakness was not tolerated in our home’), fear and distress (e.g. ‘My family viewed the world as a dangerous place’, ‘My family worried about everything’); (2) Offspring Vulnerability: escape (e.g. ‘I sometimes use alcohol, substances, sex or food to soothe myself’), heightened responsibility (e.g. ‘I feel that I have to make up for the suffering of my family and/or people’), trauma-related distress (e.g. ‘When I think about what happened to my family/my people, I feel helpless’; (3) Family Resilience: developmental sensitivity (e.g. ‘My family were sensitive about what I needed to know about what happened to our family/people’), emotional safety (e.g. ‘Growing up, it felt safe to share my emotions with my family’); (4) Offspring Resilience: belonging (e.g. ‘I have a strong sense of belonging to my community/people/family’), moral values (e.g. ‘I have strong positive values that guide me in life’), positive coping (e.g. ‘I was able to create my own story/narrative about my family's/my people's experiences’); (5) Historical Moral Injury: diminished trust as a result of trauma (e.g. ‘My people/family were betrayed by someone they once trusted’), betrayed personal morals (e.g. ‘A group of people in power did something bad to my family/my people’). There are also two additional subscales. The first assesses current Insidious Trauma in offspring, i.e. experiences of discrimination and oppression in a broader societal context, (e.g. ‘I am treated like a second-class citizen because of my ethnicity/religion’). The second is a brief PTSD symptom screener (e.g. ‘I experience reoccurring, unwanted, distressing thoughts/dreams related to what happened to my family/my people’) which we included specifically for clinicians utilizing the scale in the context of trauma-informed mental health care. Items are scored on a 5-point Likert scale (0 – Definitely not true, 5 – Very true) and offspring trauma-related distress is rated on a frequency scale (0 – Never/Almost never; 4 – Every time). For the current validation study we utilized the factor scores from the CFA analyses, except for the Insidious Trauma score which was a mean of the component items. The full measure is included in the Supplement.

3.3.2. Parental style and Offspring Adaptational impact

The Danieli Inventory of Multigenerational Legacies of Trauma short version (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015) was used to assess offspring perceptions of their survivor parent’s parental styles and related impact on themselves. Parental style is assessed by 45 statements, representing Victim, Numb, and Fighter styles. The Victim and Numb styles have been found to have detrimental impact on offspring mental health, whereas the Fighter style has been found to be neutral (Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015). Respondents answer the set of questions twice, once for each parent, and the higher of the two scores is retained to produce an final parental score for each style. Offspring Adaptational impact is the perceived impact on offspring of the parental trauma and is measured by 36 items (e.g. ‘Sometime I feel I had to fill a void that was left by murdered family members’, ‘I feel responsible for my parents’ happiness’). Offspring Adaptational impact has been found to be related to elevated mental health symptoms (Békés & Starrs, 2024). Items are scored on a 5-point Likert scale (1 – strongly disagree; 5 – strongly agree). The original, 118-item version has shown moderate to strong internal consistency for parental style and adaptational impact (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015). In the present study, the Cronbach’s alpha was .92 for maternal Victim, .91 for paternal Victim, .73 for maternal Numb, .75 for paternal Numb, .74 for maternal Fighter, and .74 for paternal Fighter, and .87 for offspring adaptational impact.

3.4. Posttraumatic and complex posttraumatic stress symptoms

We assessed PTSD and complex PTSD (C-PTSD) using the International Trauma Questionnaire (ITQ; Cloitre et al., 2018), a 12-item scale based on the International Classification of Diseases (ICD-11) criteria. The ITQ has been found to discriminate effectively between the two diagnoses, and to work equally well with clinical and community samples (Cloitre et al., 2018). In the current study, we followed the protocol described in Danieli and colleagues (2017), and first asked about the occurrence of traumatic life events, then only those who responded yes or maybe, completed the remainder of the questionnaire. In the present study, this included 72.2% of participants. In the current sample, the Cronbach’s alpha was .79 for the PTSD and .83 for the C-PTSD scale.

3.5. Depression and anxiety

The Patient Health Questionnaire was used to assess symptoms of depression and anxiety (PHQ; Spitzer et al., 1999). The depression scale (PHQ-9; Kroenke et al., 2001) measures DSM-IV based depression symptoms and the anxiety scale (GAD-7; Spitzer et al., 2006) measures DSM-IV based generalized anxiety symptoms. The PHQ scales have been used extensively and their reliability and validity are well-documented in the literature (Spitzer et al., 1999). The Cronbach’s alpha was .84 for the PHQ-9 and .90 for GAD-7 in the present study.

3.6. Resilience

The Connor-Davidson Resilience Scale (CD-RISC; Campbell-Sills & Stein, 2007) was used to measure current levels of perceived coping ability. This 10-item scale assesses respondent’s perceptions of their ability to cope in the face of adversity, e.g. adaptation to change and response to obstacles. Items are rated on a 5-point Likert scale (0 – not at all true; 4 – true nearly all of the time). The CD-RISC has been shown to have good internal reliability and construct validity (Campbell-Sills & Stein, 2007). For the current sample, the Cronbach’s alpha was .89.

3.7. Data analysis

Participants who completed at least one subscale of the HITT-Q were included in the current study. Factor analyses were completed using Mplus software version 7.4 which implements Full Information Maximum Likelihood (FIML) to manage missing data. In addition, we selected robust maximum-likelihood (MLR) estimation to manage nonnormal data and the latent factors were allowed to covary. All other analyses were completed in SPSS v.28.

To test the construct validity of the HITT-Q, confirmatory factor analyses (CFAs) were computed. In addition to a baseline single factor model, we completed three models with varying factor structures for comparison; first, a three-factor higher-order model, second, a five-factor structure representing the five higher-order dimensions of the HITT model, and finally, a 12-factor model reflecting the primary components. To assess goodness of fit, we examined multiple indices, including the χ² statistic, the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA), the Standardized Square Root Mean Residual (SRMR), the Akaïke Information Criterion (AIC; Akaike, 1987) and the Bayesian Information Criterion (BIC, Schwarz, 1978).¹ Modification indices (MIs) were considered to optimize model fit, when theoretically appropriate. Mplus input and output files are available at https://osf.io/n5jbs/.

Next, we then examined the reliability and validity of the selected model. To examine the internal consistency, we computed the interitem correlation coefficients (ICC) for each of the factors. For concurrent validity, we completed Pearson correlations to examine the associations between the HITT-Q factors and those on the Danieli Inventory (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015). Finally, for predictive validity, we undertook multiple regression analyses to assess the relationship between the identified HITT-Q factors and mental health symptoms (PTSD, C-PTSD, anxious, and depressive symptoms) and perceived coping ability (CD-RISC). Given that the Insidious trauma subscale, is not a main component of the HITT-Q, but rather assesses a current stressor, we conducted separate linear regression analyses to test whether Insidious trauma was associated with mental health symptoms.

4. Results

4.1. Confirmatory factor analyses

The single factor baseline model did not show good fit, suggesting that the structure of the data was multifactorial. We therefore opted to examine three multifactor models, based on the underlying theoretical model (HITT model; Starrs & Békés, 2024). For model goodness of fit statistics see Table 4.

Table 4. CFA model fit summary.

Measure	χ²	df	CFI	TLI	RMSEA	SRMR	AIC	BIC
1 Factor	10969.66*	1539	.37	.35	.09	.11	127083.66	127867.66
3 Factors	8037.958*	1536	.57	.55	.08	.10	123823.71	124621.47
5 Factors	6484.13*	1529	.67	.66	.07	.09	122160.38	122990.23
12a Factors	5334.69*	1473	.74	.72	.06	.08	120983.31	122069.91
12b Factors	5032.10*	1365	.75	.73	.06	.08	116896.06	117950.57
12c Factors	3849.91*	1365	.83	.82	.05	.07	115633.81	116688.32
12d Factors	2709.58*	1302	.90	.89	.04	.05	112674.38	113760.97

Notes: χ²: chi-square goodness of fit statistic; df: degrees of freedom; CFI: Comparative Fit Index; TLI: Tucker Lewis Index; RMSEA: Root Mean Square Error of Approximation; SRMR: Standardized Square Root Mean Residual; AIC: Akaike Information Criterion ; 12_a Factors: original 12 factor model with no respecification; 12_b Factors: 12 factor model after item deletion; 12_c Factors: 12 factor model after item moves; 12_d: final 12 factor model after all respecification.

*p < .001.

The first model comprised a three-factor structure examining Vulnerability, Resilience, and Historical Moral Injury. This model was rejected as none of the individual fit indicators reached threshold for acceptability. The second model examined a 5-factor structure based on the higher-order dimensions of the HITT model, including Family Vulnerability, Family Resilience, Offspring Vulnerability, Offspring Resilience, and Historical Moral Injury. This model showed improved goodness of fit over the three-factor model, however, CFI and TLI, did not meet acceptability threshold, and was therefore rejected. Lastly, we tested a 12-factor model reflecting the twelve components of the HITT model. This model showed notably improved fit compared to the 5-factor model, however, CFI and TLA remained below threshold. Examination of the factor loadings and the modification indices (MIs) suggested several changes to the factors and the model that would significantly improve the model fit. First, there were three items with loadings below .30 (items 12, 13 and 28). These items were deleted (see Table 4 model 12_b). Second, there were some items that the MIs suggested moving to alternative factors. We studied the content of these items for theoretical consistency and implemented six valence coherent moves (i.e. vulnerability items on vulnerability factors, resilience items on resilience factors) (see Table 4 model 12_c). Finally, the MIs also indicated that the model fit would benefit from correlating some of the residuals with strong content overlap (items 4–16, 4–17, 6–7, 16–17, 19–20, 29–30, 46–47, 59–60, 63–64). Overall, this resulted in a final 12-factor respecified model (Table 4 model 12_d) that demonstrated acceptable goodness of fit and was theoretically meaningful. See Table 5 for the standardized loadings and the inter-factor correlations on this final 12-factor model.

Table 5. Standardized loadings and inter-factor correlations for the final 12-factor model.

Family Vulnerability Factors
1. DTR Comm		2. TI Parenting		3. Fear		4. Distress
Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)
1.	.72 (.03)	6.	.70 (.03)	11.	.60 (.04)	16.	.58 (.04)
2.	.80 (.02)	7.	.77 (.02)	14.	.39 (.04)	17.	.67 (.03)
3.	.79 (.02)	8.	.66 (.03)	15.	.44 (.04)	18.	.77 (.03)
5.	.39 (.04)	9.	.40 (.04)	20.	.44 (.03)	4.	.52 (.04)
		10.	.61 (.03)	19.	.70 (.04)
Family Resilience Factors		Offspring Vulnerability Factors
5. Family Resilience		6. Escape		7. Heigh Resp.		8. TR Distress
Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)
48.	.46 (.04)	21.	.57 (.04)	25.	.57 (.04)	29.	.35 (.03)
49.	.47 (.04)	22.	.51 (.04)	26.	.38 (.05)	30.	.42 (.03)
50.	.65 (.03)	23.	.62 (.04)	27.	.42 (.05)	31.	.39 (.04)
52.	.77 (.02)	24.	.58 (.04)			32.	.81 (.02)
						33.	.82 (.02)
						34.	.86 (.02)
						35.	.75 (.02)
Offspring Resilience Factors
9. Coping		10. Belonging		11. Values		12. HMI
Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)	Item	Est. (SE)
44.	.64 (.04)	40.	.74 (.02)	45.	.71 (.04)	59.	.72 (.03)
51.	.80 (.04)	43.	.78 (.02)	46.	.47 (.04)	60.	.78 (.03)
	38.	.47 (.04)	47.	.60 (.04)	61.	.89 (.02)
	39.	.59 (.03)	62.	.90 (.02)
	41.	.60 (.03)	63.	.44 (.04)
	42.	.76 (.02)	64.	.41 (.04)
			65.	.62 (.03)
Correlations	1	2	3	4	5	6	7	8	9	10	11	12
1. DTR Comm	–
2. TI Parenting	.72**	–
3. Fear	.58**	.81**	–
4. Distress	.78**	.96**	.78**	–
5. Family Res.	−.85**	−.63**	−.45**	−.70**	–
6. Escape	.48**	.62**	.56**	.62**	−.45**	–
7. Heigh. Resp.	.42**	.51**	.73**	.52**	−.29**	.57**	–
8. TR Distress	.38**	.41**	.66**	.46**	−.34**	.62**	.58**	–
9. Coping	−.44**	−.26**	−.42**	−.37**	.68**	−.33**	−.31**	−.36**	–
10. Belonging	−.15*	−.06	.12*	−.03	.38**	−.05	.42**	.01	.23**	–
11. Values	−.48**	.40**	−.21*	−.37**	.76**	−.41**	.11	−.19*	.51**	.61**	–
12. HMI	.09	.18**	.16*	.18**	−.06	.09	.28**	.16**	−.03	.17**	.04	–

Notes: All factor loadings were significant (p < .001); items 12, 13 & 28 were deleted due to loadings < .30; items 36 & 37 (PTSD screener) and items 53–58 (Insidious Trauma) were not included in the CFA. DTR Comm.: Dysregulated and Trauma-related Communication; TI Parenting: Trauma-influenced Parenting; Family Res.: Family Resilience; High. Resp.: Heightened Responsibility; TR Distress: Trauma-related Distress; HMI: Historical Moral Injury.

**p < .001, *p < .05.

Regarding the correlations between the factors, the majority were significant and in the expected direction. One exception to this was the offspring resilience factor Belonging which showed an unexpected and mixed pattern (for example, positively related to the other offspring resilience factors but also positively related to the family factor of Fear). Additionally, the Values factor was positively but nonsignificantly related to Heightened Responsibility. The Historical Moral Injury factor was not significantly correlated with several of the other factors, although these were in the expected direction, and finally HMI were positively related with Belonging and Values, although only Belonging was significant. Of note, the current study was completed in a single culturally homogenous sample, who experienced the same historical trauma event, as such at this early stage, it is unclear if this is specific to this sample.

4.2. Internal consistency of the HITT-Q

Findings, mostly, demonstrate acceptable to good levels of internal consistency (.70 ≤ α < .90; George & Mallery, 2003), with Cronbach’s alphas for the 12-factors ranging from .65 to .89. The only exception was the 3-item Heightened Responsibility (.45), suggesting that adding additional items in the future may be beneficial. The ICC for an overall total score is not reported, as a single score is not theoretically meaningful as demonstrated by the one-factor model fit indices above (Table 6).

Table 6. Internal consistency & descriptive statistics for the 12-Factor Model.

			Cronbach’s	Number
Factor	Mean	SD	Alpha	items
1. DTR Comm	1.36	.99	.75	4
2. TI Parenting	.99	.90	.78	5
3. Fear	1.35	.87	.70	4
4. Distress	.96	1.05	.84	4
5. Family Resilience	2.29	1.06	.70	4
6. Escape	1.13	.86	.65	4
7. Heigh. Resp.	1.81	.93	.45	3
8. TR Distress	1.70	.87	.83	7
9. Coping	1.85	1.27	.68	2
10. Belonging	2.74	.89	.82	6
11. Values	2.59	1.13	.73	3
12. HMI	2.65	.98	.87	7

Notes: DTR Comm.: Dysregulated & Trauma-related Communication; TI Parenting: Trauma-influenced Parenting; Heightened Resp.: Heightened Responsibility; TR Distress: Trauma-related Distress; HMI: Historical Moral Injury.

4.3. Concurrent validity of the HITT-Q

To examine concurrent validity, we computed the correlations between the HITT-Q and the Danieli Inventory’s relevant subscales. See Table 7 for details.

Table 7. Pearson Correlations between the 12-factors of the HITT-Q and the Danieli Inventory.

	Parenting Style
HITT-Q Factor	Victim	Numb	Fighter	Offspring Impact
1. DTR Comm	.56**	.60**	.19**	.44**
2. TI Parenting	.74**	.56**	.20**	.58**
3. Fear	.72**	.52**	.21**	.71**
4. Distress	.74**	.58**	.22**	.59**
5. Family Resilience	−.45**	−.53**	−.12**	−.37**
6. Escape	.53**	.44**	.18**	.68**
7. Heigh. Resp.	.57**	.43**	.24**	.73**
8. TR Distress	.44**	.33**	.17**	.69**
9. Coping	−.27**	−.32**	−.05	−.34**
10. Belonging	.06	.01	.11*	.14**
11. Values	−.22**	−.33**	.02	−.16**
12. HMI	.15**	.09*	.07	.19**

Notes: DTR Comm.: Dysregulated and Trauma-related Communication; TI Parenting: Trauma-influenced Parenting; Heigh. Resp.: Heightened Responsibility; TR Distress: Trauma-related Distress; HMI: Historical Moral Injury Offspring Impact: Offspring Adaptational Impact.

**p < .001, *p < .05.

In line with our expectations, the family and offspring vulnerability scales were significantly and positively correlated with all three trauma-impacted parenting styles, as well as with Offspring adaptational impact (OAI). Family Resilience was negatively and significantly correlated with the three parenting styles, and with OAI. Offspring Coping and Values were significantly and negatively correlated with the Victim and the Numb parenting styles, and with OIA. Belonging was unrelated to Victim and Numb and significantly positively related to OIA, as well as to the Fighter parental style, whereas Coping and Values were unrelated to Fighter style. Finally, HMI was significantly and positively related to Victim, Numb and OIA, but not the Fighter style.

4.4. Predictive validity of the HITT-Q

Table 8 shows the multiple regression analyses for the relationship between the HITT-Q vulnerability factors and mental health symptoms (PTSD, C-PTSD, anxiety, and depression symptoms) and the HITT-Q resilience factors and current perceived coping ability (CD-RISC).

Table 8. Multiple regression analyses for the 12-Factors & Mental Health.

	β	SE	T	p	95%CI	F	df	p	Adj. R^2
Vulnerability Factors
PTSD
DTR Comm	−.31	.60	−.53	.600	[−1.48, .86]	7.51	8, 353	<.001	.13
TI Parenting	−.69	2.74	−.25	.801	[−6.08, 4.70]
Fear	−.94	1.84	−.51	.610	[−4.56, 2.68]
Distress	2.94	2.84	1.03	.302	[−2.65, 8.54]
Escape	−1.15	.95	−1.21	.226	[−3.01, .71]
Heigh. Resp.	1.20	1.02	1.19	.236	[−.80, 3.23]
TR Distress	4.46	1.98	.30	.025	[.57, 8.35]
HMI	.00	.32	.00	.998	[−.62, .62]
C-PTSD
DTR Comm	.13	.48	.27	.788	[−.81, 1.07]	33.97	8, 353	<.001	.42
TI Parenting	−.64	2.19	−.29	.769	[−4.96, 3.67]
Fear	−1.06	1.47	.72	.471	[−1.83, 3.96]
Distress	.57	2.28	.25	.802	[−3.90, 5.04]
Escape	2.71	.76	3.59	<.001	[1.23, 4.20]
Heigh. Resp.	−1.70	.82	−2.08	.039	[−3.31, −.09]
TR Distress	5.49	1.58	3.47	<.001	[2.37, 8.60]
HMI	.57	.25	2.27	.024	[.08, 1.07]
Depression
DTR Comm	−.50	.46	−1.10	.271	[−1.40, .40]	37.62	8, 503	<.001	.36
TI Parenting	−2.22	2.03	−1.10	.274	[−6.21, 1.76]
Fear	−.77	1.38	−.56	.576	[−3.48, 1.94]
Distress	4.43	2.16	2.05	.041	[.19, 8.67]
Escape	2.24	.71	3.19	.002	[.86, 3.63]
Heigh. Resp.	−1.51	.76	−1.98	.048	[−3.00, −.01]
TR Distress	6.69	1.45	4.61	<.001	[3.84, 9.53]
HMI	.06	.24	.24	.811	[−.41, .52]
Anxiety
DTR Comm	−1.27	.43	−2.96	.003	[−2.11, −.43]	22.42	8, 504	<.001	.25
TI Parenting	.21	1.90	.11	.912	[−3.52, 3.94]
Fear	−1.54	1.29	−1.20	.232	[−4.07, .99]
Distress	2.96	2.02	1.46	.144	[−1.01, 6.92]
Escape	.29	.66	.44	.658	[−1.00, 1.58]
Heigh. Resp.	.24	.71	.33	.740	[−1.16, 1.64]
TR Distress	5.92	1.36	4.37	<.001	[3.26, 8.58]
HMI	−.21	.22	−.95	.343	[−.65, .23]
Resilience Factors
CD-RISC
Family Resil.	−3.40	1.12	−3.03	.003	[−5.60, −1.20]	22.00	4, 510	<.001	.14
Coping	2.20	.56	3.93	<.001	[1.10, 3.30]
Belonging	.60	.59	1.01	.313	[−.56, 1.75]
Values	3.00	.86	3.48	<.001	[1.30, 4.69]

Notes: SE: Standard Error; CI: Confidence Intervals; df: Degrees of Freedom; Adj. R²: Adjusted R²; DTR Comm.:  Dysregulated and Trauma-related Communication; TI Parenting: Trauma-influenced Parenting; Heigh. Resp.: Heightened Responsibility; TR Distress: Trauma-related Distress; HMI: Historical Moral Injury; CD-RISC: Connor-Davidson Resilience Scale; Family Resil.: Family Resilience; Offspring Impact: Offspring Adaptational Impact.

Each of the multiple regression models were significant, predicting between .13 and .42 of the variance in mental health outcomes. Within each model though, different individual factors were driving the effects. For PTSD symptoms, the only significant factor was Trauma-Related Distress. CPTSD was significantly predicted by Escape, Trauma-related Distress and HMI, and by Heightened Responsibility although this was a negative effect. Depressive symptoms were predicted by family Distress and offspring Escape and Trauma-related Distress. Again, the effect for Heightened Responsibility was significant but negative. Anxiety symptoms were predicted by parental Dysregulated and Trauma-related Communication, as well as offspring Trauma-Related Distress. Family Resilience was significantly related to CD-RISC, although the relationship was negative. Offspring Coping and Values significantly predicted increased CD-RISC, however offspring Belonging did not.

Finally, given that the HITT model posits that Insidious trauma contributes to current mental health, we examined this hypothesis. Current Insidious Trauma was associated with C-PTSC (F(1, 360) = 7.65, p = .006), depressive (F(1, 510) = 5.80, p = .016), and anxious symptoms (F(1, 511) = 10.21, p < .001), but not current PTSD (F(1, 360) = 3.34, p = .068).

5. Discussion

With the growing interest in the transgenerational impact of historical trauma, the need for a cross-population model and measure for use in both research and clinical settings has become especially urgent. The current study aimed to fill this important gap in the field, by developing such a measure based on our previously posited HITT model (Starrs & Békés, 2024). This comprehensive model considers five main higher-order dimensions, including Family and Offspring Vulnerability and Resilience, as well as Offspring levels of Historical Moral Injury. In addition, in line with current research on the impact of racism on mental health (see Paradies et al., 2015), the model proposes Insidious Trauma as an important contextual variable that affects offspring mental health. Finally, the HITT-Q also includes a PTSD screener, composed of 2 items, one that assesses intrusive reexperiencing and a second that measures avoidance. These items were specifically included to provide clinicians using the HITT-Q with a preliminary indication of the presence of posttraumatic symptoms.

Confirmatory factor analyses indicated that a respecified 12-factor model represents both a plausible and theoretically meaningful model, with acceptable goodness of fit, and sufficient reliability and internal validity. The 12 factors, organized under the higher conceptual dimensions are the following: I. Family Vulnerability: (1) Dysregulated and Trauma-related Communication; (2)Trauma-influenced Parenting, (3) Fear; (4) Distress; II. (5) Family Resilience, III. Offspring Vulnerability: (6) Escape; (7) Heightened Responsibility; (8) Trauma-related distress; IV. Offspring Resilience: (9) Coping; (10) Belonging; (11) Values; V. (12) Historical Moral injury.

The 12 subscales demonstrated acceptable to good levels of internal consistency (expect Heightened Responsibility). In addition, the HITT-Q’s concurrent validity was demonstrated through correlations with the Danieli Inventory of Multigenerational Legacies of Trauma (Danieli, Norris, Lindert, Paisner, Engdahl, et al., 2015; Danieli, Norris, Lindert, Paisner, Kronenberg, et al., 2015) in the expected ways. The HITT-Q also showed good predictive validity with the factors under Family and Offspring Vulnerability dimensions predicting higher reported levels of PTSD, C-PTSD, anxiety, and depression, and the factors under Family and Offspring Resilience dimensions predicting higher perceived ability cope with adversity, as measured by the CD-RISC scale.

5.1. Clinical implications

In line with our expectations, we found that the parental vulnerability factors in the HITT model were associated with increased mental health symptoms, thus suggesting that addressing these vulnerability factors may have the potential to lead to reduced mental health problems later in life. Specific vulnerability factors include a general atmosphere of fear and distress, too much or too little trauma-related communication, generally dysregulated communication patterns, such as emotional avoidance, and trauma-impacted parenting behaviours including the use of shame and guilt to control child behaviour. These parental vulnerabilities can be identified and addressed through parental interventions; there are several existing programmes to address similar problems in the context of preventing parental transgenerational trauma transmission. These interventions most commonly address parental emotional regulation, reflective functioning, and attachment, (Murphy et al., 2015; Steele et al., 2019), which have been shown to be impacted by childhood trauma and are conceptualized as important psychological underpinnings of later parenting behaviour, as are the variables we included in our Family Vulnerability Factor. Inversely, our findings also highlighted the positive impact of family resilience including a heightened sensitivity to the developmental maturity of the child in relation to communication about parental trauma, as well as providing a sense of emotional safety within the family. These parenting skills are also closely connected to parents’ ability to regulate their own emotions, to mentalize their own and their child’s emotional-cognitive states, and to developing a secure attachment relation with their offspring (e.g. Rosenblum et al., 2017). These types of parenting interventions are generally offered to mothers who have experienced childhood interpersonal trauma (Steele et al., 2019), parents with a diagnosis of PTSD (Meijer et al., 2023), or veterans of the armed forces (Creech et al., 2022). Our findings suggest, that it would be beneficial to extend and adapt these programmes to parents who have experienced historical intergenerational traumas, such as indigenous peoples, refugees, or survivors of other social traumas.

In addition, we identified offspring vulnerability (escape heightened responsibility, and trauma-related distress), and resilience factors (sense of belonging to the community, having strong moral values, and positive coping skills), which impacted mental health and resilience in adulthood. Although early traumatic experiences cannot be undone, these offspring factors could also be addressed by specific interventions. For example, school and community programmes that facilitate a sense of belonging and positive identity in the context of being part of a minority group, the development of guiding moral principles, and trauma-specific coping skills such as creating a coherent trauma narrative, could help to counteract its impact.

In addition, our findings show that historical moral injury may be an important contributor to mental health symptoms. Historical moral injury specifically denotes feelings of injustice and betrayal, as such our findings suggest that this element of HT must be addressed, both at the individual level and at the collective level, for example, with such actions as reparations and the adoption of laws and policies that ensure that such horrific events as the Holocaust, residential schools and slavery, can never happen again (United Nations High Commission Human Rights, 2019).

Naturally, in order to address both parental and offspring factors, health and mental health care providers need to acknowledge historical intergenerational traumatization. Often, traumatic events that happened several decades ago and possibly even in the previous generation, are not considered relevant in clinical practice; however, our findings are in line with the growing literature that support the ongoing transgenerational impact of such events. In addition, there is accumulating evidence of continued impact of HT in the children of the offspring of survivors (Greenfeld et al., 2022; Payne & Berle, 2021; Sagi-Schwartz et al., 2008), as well as emerging evidence that the transgenerational family vulnerability and resilience processes identified in survivors, are displayed in their offspring’s subsequent parental behaviour (Békés & Starrs, 2024). Thus, it is crucial that clinicians recognize that the transgenerational impact of historical trauma may affect several generations, even when the events occurred far in the past. As such, it is not sufficient to provide only trauma-informed care, but also transgenerational trauma-informed care.

5.2. Limitations and future research

Our study, despite its strengths, also has several limitations. First, although we had a large sample, the present study only included one population, Holocaust survivors’ offspring. This excluded us from being able to test the cross-population validity of the HITT-Q, and future studies including other traumatized groups are needed. Furthermore, the further studies with different populations must specifically consider the respecification of the model, especially the deletion of three items, which may resonate more highly with other traumatized groups, as well as the inter-item correlations which may be specific to this sample. Our research group has recently collected data in an Indigenous sample in the US and in offspring of survivors of the 1984 Sikh genocide in India, and cross-population findings will be published shortly. Second, our data was collected when the COVID-19 pandemic was ongoing, and thus participants might have experienced an increased level of distress and thus heightened level of mental health symptoms which might have reflected the pandemic-related stress rather than participants’ general levels of distress. Finally, our sample focused on second generation offspring however, future studies should examine the validity of the current model in a third generation sample.

6. Conclusion

In sum, our findings suggest that the HITT-Q is a valid and useful tool for assessing vulnerability and resilience factors related to historical intergenerational trauma in clinical and research settings. Using this measure could enable the identification of transgenerational trauma transmission mechanisms, as well as areas of strengths and vulnerability at the family and offspring levels, in order to offer transgenerational-informed care, that can begin to break the cycle of transgenerational trauma.

Disclosure statement

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

Additional information

Funding

This study was supported by the Conference on Jewish Material Claims Against Germany (Claims Conference).

Notes

1 Hu and Bentler (1999) advise a RMSEA <.06 although they suggest that a criterion of .08 is acceptable. Equally, they propose an ideal CFI/TLI >.96 and, minimally, .90. For the SRMR, Asparouhov and Muthén suggest a value less than .05 is good and less than .08 is acceptable. The chi-square should ideally have a p-value >.05 (Bollen, 1989), however the chi-square is highly sensitive to sample size, leading to significance in larger samples, such as in the current study (Russell, 2002). Thus the model chi-square test results are reported here for descriptive purposes but were not used in evaluating model fit.