Childhood Trauma Questionnaire-Short Form: Evaluation of Factor Structure and Measurement Invariance

Childhood trauma is known to put individuals at risk for mental and physical challenges later in life. Retrospective assessment of early abuse and neglect is critical for prevention and intervention efforts targeted at reducing the negative impacts of childhood trauma. The Childhood Trauma Questionnaire-Short Form (CTQ-SF) is a widely used measure to assess trauma among adolescents and adults, though there are some inconsistencies with regard to its factor structure and psychometric properties. The purpose of the current study was twofold: (1) to evaluate the hypothesized five-factor structure of CTQ-SF and (2) to test measurement invariance (equivalence) of the instrument across gender and race in a large, nationally representative sample of US adults (N = 863). Confirmatory factor analysis results indicate that the five-factor model fit the data well. The results also show that the five-factor model is generally invariant by gender and racial/ethnic groups and that the form’s subscales positively correlate with depression, anxiety, and abnormal inflammatory biomarker activity. Assessment is critical for the advocacy and treatment of individuals who have experienced abuse and neglect as children and adolescents. Our findings suggest that the CTQ-SF is a valuable tool for assessing childhood trauma and can be used in advocacy and treatment efforts.

. The origin and exacerbation of these conditions among individuals who have experienced childhood traumas may be underappreciated by healthcare providers, either because the afflicted individuals tend to depersonalize and dissociate (and therefore underreport or completely deny their trauma symptoms) or because they instead seek medical treatment for urgent or seemingly unrelated matters (such as acute injuries, alcohol intoxication, and drug overdoses).
Research suggests that universal screening approaches, such as those that underscore the commonality of adverse childhood experiences and diversity of trauma associated with early abuse and neglect (e.g., trauma-informed care; Substance Abuse and Mental Health Services Administration, 2014), have measurable and positive outcomes such as enhancing the patient-provider relationship, mitigating over-pathologizing and stigma issues, enhancing recovery, and promoting resilience.As such, universal screening approaches have the potential to help individuals who have Over the past decade, there has been much interest in understanding the prevalence and impact of adverse childhood experiences and traumatic events that unfold over sensitive periods of development (e.g., abuse, rejection, and abandonment by caregivers; interpersonal violence exposure) -given recent developments suggesting these experiences have broad, cumulative, and lasting effects.A longitudinal study linking childhood trauma to negative outcomes in adulthood concluded that trauma is widespread (i.e., 70% prevalence) and that individuals who report having had multiple traumas in their early years tend to have worse outcomes than those who do not (Copeland et al., 2018).
Traumatic experiences in childhood are linked to pervasive and debilitating psychological disorders and medical illnesses such as PTSD, substance abuse, psychosis, Crohn's disease, and seizures (i.e., Baudin et al., 2017; experienced childhood trauma recover, prosper, and end the cycle of trauma for generations to come. However, much research is still needed to better understand the effectiveness of available assessment tools designed for adults who were exposed to abuse as children.For example, some tools are limited to a specific type of trauma (e.g., childhood neglect or exposure to domestic violence; Cotter et al., 2018;Kobulsky et al., 2020;McGuigan et al., 2018), and thus neglect to assess poly-victimization (i.e., prolonged exposure to multiple types of trauma; Charak et al., 2019;Dierkhising et al., 2019;Musicaro et al., 2019) and its cumulative impact.Moreover, studies evaluating trauma suggest that inconsistent psychometric evidence limits many of the currently available trauma symptom assessment tools (Eklund et al., 2018;McLennan et al., 2020).
There is currently a great need for tools that can retroactively assess childhood trauma.Any assessment of early abuse and neglect, however, must acknowledge that individuals who have gone through trauma, including repeated experiences of abuse and neglect, are likely to experience high levels of fear, depression, anxiety, hostility, aggression, and low frustration tolerance (Glück et al., 2017;Hébert et al., 2018;Storvestre et al., 2020;Zhu et al., 2020).Each of these may, in turn, lead these individuals to deny having had early traumatic experiences.Indeed, there is preliminary evidence suggesting that individuals with childhood trauma histories tend to be suspicious of people they distrust or possibly perceive to have malevolent intentions and, as a result, may refrain from sharing their feelings or disclosing their trauma on self-report measures (Alaggia et al., 2019;Burns et al., 2020;Lahtinen et al., 2022).Attention to the evaluative criteria associated with childhood trauma, may give rise to studies that further explore the associations between trauma and childhood in ways that can enrich the advocacy, assessment, and treatment of individuals who report having traumatic experiences in childhood.Among the most commonly used instruments to assess experiences of early traumas, the Childhood Trauma Questionnaire -Short Form (CTQ-SF; Bernstein et al., 2003) was designed specifically to assess childhood trauma in adults.

Childhood Trauma Questionnaire-Short Form
The CTQ-SF is a 28-item screener designed specifically to assess childhood trauma in adults and has been widely used to assess trauma for clinical and research purposes.CTQ-SF validation studies have been carried out with several groups of people, including individuals in the community and patients hospitalized for psychiatric disturbances (Khosravani et al., 2017;Kim et al., 2017;Luoni et al., 2018;Mizuki & Fujiwara, 2021;Xie et al., 2018).The results of multiple factor analyses (i.e., exploratory and confirmatory) have revealed that the CTQ-SF reliably assesses trauma (Aloba et al., 2020;He et al., 2019;Jiang et al., 2018).Moreover, some studies have provided psychometric evidence for a 3-item Minimization/Denial subscale and an Overprotection/Overcontrol factor (e.g., MacDonald et al., 2016;Şar et al., 2021;Schmidt et al., 2020;Wu et al., 2022).
Several studies have shown that the CTQ-SF is multidimensional, suggesting that its items measure trauma symptoms of multiple underlying constructs (i.e., sexual, physical, emotional abuse, and physical, emotional neglect; Bernstein et al., 2003).However, other studies have not been able to replicate this five-factor structure (Forde et al., 2012).In addition, studies have yielded inconsistent results with regard to the measure's reliability with men versus women (Forde et al., 2012;Rodriguez et al., 2018;Thombs et al., 2007;Wright et al., 2001).A number of studies have also raised the possibility that the CTQ-SF may unfavorably bias racial and ethnic minorities (Forde et al., 2012;Rodriguez et al., 2019).A critical question that emerges from these findings is whether differences in CTQ-SF scores reflect actual differences in trauma occurrence or whether they are due to differences in how individuals conceptualize, interpret, and report their traumas when responding to the CTQ-SF items.Thus, there is a need for further research on whether the CTQ-SF measures trauma symptoms similarly across genders and ethnic groups in the general population.This is compatible with recent reviews of studies reporting on the CTQ-SF that indicate a need for more research on the instrument's reliability, cross-cultural/multigroup validity, and systematic evaluation of measurement error and risk for bias in multigroup comparisons (Georgieva et al., 2021).Measurement invariance is important to assess in the CTQ-SF because it ensures that the questionnaire is measuring the same construct across different groups of people.The CTQ-SF is a widely used self-report measure of childhood trauma, it is commonly used to compare trauma levels across different groups, such as different genders or cultural backgrounds.However, if the measurement properties of the questionnaire are not consistent across these groups, then any differences observed in trauma levels may be due to measurement bias rather than true differences in trauma experiences.For instance, if the items on the CTQ-SF are interpreted differently by different groups, then the questionnaire may not be measuring the same construct in these groups.This could lead to differences in reported trauma levels that are not reflective of true differences in trauma experiences.By assessing measurement invariance, researchers can ensure that the CTQ-SF is measuring the same construct across different groups, allowing for more accurate comparisons of trauma levels.

The Current Study
The current study's objective is to provide psychometric data for the CTQ-SF to contribute to the quantitative literature examining instruments designed to assess childhood trauma in adults.Based on prior research with both the CTQ and the CTQ-SF, we hypothesized that CTQ-SF scale items would converge to a five-factor model consistent with the multifactorial structure.To verify the measurement equivalence of the CTQ-SF factors for men and women as well as for Black and White Americans, we implemented measurement invariance tests.We also predicted that CTQ-SF scores would positively correlate with physical and mental health outcomes.Specifically, we hypothesized that higher levels of trauma (as indicated by high CTQ SF scores) would correlate with higher levels of depression, anxiety, and proinflammatory cytokines, thereby supporting the measure's convergent validity.

Methods
The current study data were drawn from the 2012-2016 Midlife in the United States (MIDUS) Refresher Biomarker Study (Weinstein et al., 2019).The MIDUS is a national, multidisciplinary, longitudinal survey designed to investigate biological, psychological, and social factors critical to understanding health outcomes among middle-aged Americans.The MIDUS Study protocol included a comprehensive health interview as well as a physical examination, biological specimen collection (i.e., blood, urine, saliva), and brain imaging.Data collection for the MIDUS Study began in 1995 with a random sample of over 7,000 adults, ages 25 to 74, who were re-assessed from 2004 to 2009 (MIDUS 2), 2012-2013 (MIDUS Refresher), and 2013-2014 (MIDUS 3).The CTQ-SF (Bernstein et al., 2003) was included in the MIDUS Refresher Study; therefore, we used the MIDUS Refresher Study database for the current project.

Participants
The final analytic sample for the current study included 863 adults (48% male) ranging in age from 26 to 78 years (M = 52.73;SD = 13.41).About 70% of participants were Caucasian, and 20% were Black.Approximately 52% of participants in the study had a bachelor's degree or higher.

Measures
As discussed above, the Childhood Trauma Questionnaire-Short Form (CTQ-SF; Bernstein et al., 2003) is a 28-item self-report instrument developed to assess childhood trauma, including physical abuse, emotional abuse, sexual abuse, physical neglect, and emotional neglect.In addition, the instrument has three items designed to detect socially desirable responses.Each item is rated on a 5-point Likert scale from Never True to Very Often True to reflect the frequency of abuse experiences.The CTQ-SF is one of the most widely used measures of childhood trauma.Scores can be continuous or differentiated by severity level to identify high-risk patients.Reliability for the CTQ-SF is good; the measure has shown high internal consistency reliability scores.The Emotional Abuse, Physical Abuse, Sexual Abuse, Emotional Neglect, and Physical Neglect subscales have reported coefficients of 0.82-0.87,0.77-0.86,0.92-0.94,0.66-0.90, and 0.60-0.81,respectively (Forde et al., 2012;MacDonald et al., 2015;Rasmussen et al., 2018).
The CTQ also includes a Minimization/Denial (MD) subscale to detect underreporting of childhood trauma.Three reverse-scored statements are rated on a Likert scale, with high minimization/denial present if the participant reports having a perfect family, a perfect childhood, and no desire to change anything about their family.The MD scale has been shown to have good internal consistency reliability, with Cronbach's alpha coefficient ranges of 0.68 to 0.77 (Church et al., 2017;MacDonald et al., 2015MacDonald et al., , 2016)).In the current study, the Cronbach's alpha reliability coefficient for the MD subscale is 0.75.Depression was measured using the Center for Epidemiologic Studies-Depression scale (CES-D; Radloff 1977).The CES-D is a 20-item survey that participants rate using a 4-point Likert scale from Rarely or None of the Time to Most or All of the Time to report on the frequency of their depressive symptoms (e.g., "I felt depressed," "I had crying spells").Total CES-D scores range from 0 to 60, with higher scores indicating greater levels of depression.The CES-D is a well-validated measure for depression, with satisfactory We focused on Black and White racial groups because the sample sizes of the other racial/ethnic groups were small.Equivalence of the CTQ-SF across groups was tested using a series of analyses involving increasingly restrictive levels of measurement invariance (Cheung & Rensvold, 2002).Evidence for configural invariance, which measures the association between scale items and their hypothesized latent constructs, was evaluated.This analysis involved testing the baseline model with identical loading patterns across groups without imposing any constraints.This model served as a reference against which the more restrictive models were compared.Second, evidence for metric invariance (i.e., factor loadings and items measuring latent constructs across groups) was evaluated by constraining factor loadings to be equal across groups.Third, evidence for scalar invariance was evaluated by constraining factor loadings and item intercepts to be equal across groups.
For tests of the presence of measurement invariance, we used chi-square statistics, which measure differences between the original data and the model-implied solution and changes in CFI, RMSEA, and Bayesian Information Criterion (BIC).The chi-square test is sensitive to sample size and model complexity.For the sake of consistency with previous validation studies, however, we also report changes in our chi-square values.An increase of less than 0.010 for CFI and a decrease of 0.015 for RMSEA are indicative of support for the more constrained model (Cheung & Rensvold, 2002).With regard to BIC, the model with the lowest BIC value is preferred, as it indicates a better balance between model complexity and goodness-of-fit.All CFA analyses were performed in R (R Core Team, 2022) using Lavaan, an R statistics package for structural equation modeling (Rosseel, 2012).Because the distribution of the data deviates from normality, we used robust maximum likelihood estimation.In addition to CFA, we computed Cronbach's alphas to assess internal consistency reliability for each subscale of the CTQ-SF.

Descriptive Statistics
Means, standard deviations, and correlations among variables are summarized in Table 1.There were moderate to strong intercorrelations between CTQ-SF subscales.The Abuse and Neglect subscales of the CTQ-SF positively correlated with IL-6 and CRP levels; however, there was no association between CTQ-SF scores and TNF-α levels.Minimization/denial negatively correlated with the other CTQ-SF subscales and was most strongly associated with evidence of reliability and validity (Cosco et al., 2017).In the current study, the scale's internal consistency reliability as measured using Cronbach's alpha is 0.88.
Trait anxiety was measured using the Spielberger Trait Anxiety Inventory (STAI; Spielberger, 1983).This scale consists of 20 items rated on a 4-point Likert scale and can be used in clinical settings to diagnose anxiety and to distinguish anxiety syndromes from depressive ones.Example items include "I wish I could be as happy as others seem to be" and "I worry too much."Higher STAI scores indicate greater trait anxiety.In the current study, the STAI demonstrated excellent reliability (Cronbach's alpha = 0.90).
The MIDUS Refresher Study database consists of several biomarker measures.The current study examines three pro-inflammatory cytokine markers: Interleukin-6 (IL-6), Tumor necrosis factor-α (TNF-α), and Acute Phase C-reactive protein (CRP).Serum IL-6 levels were measured with enzyme-linked immunosorbent assays (ELISA), commonly used to assess antibodies and antigens in biological samples.The current study's assay range is 0.156 to 10 picograms per milliliter (pg/mL), with inter-and intra-assay coefficients of variability (CV) of 3.73% and 15.66%.The Quantikine ELISA test was used to measure TNF-α levels.TNF-α assays ranged from 0.69 to 248 pg/mL, with coefficients of variability of 7% and 3.19%.C-reactive protein (CRP) levels were assessed by a Behring Nephelometer II (BNII) Analyzer.The assay range is 0.164 to 800 µg/ml, with variability ranges of 2.3-4.4% and 4.72-5.16%.

Data Analyses
Confirmatory factor analysis (CFA) was used to evaluate whether the five-factor model proposed in previous research with the CTQ-SF (Bernstein et al., 2003;Devi et al., 2019;He et al., 2019;Sacchi et al., 2018) fit the current data.Model goodness of fit was assessed using several indices, including the model chi-square, the comparative fit index (CFI), the Tucker-Lewis index (TLI), the root mean square error of approximation (RMSEA), and the standardized root-mean-square residual (SRMR).Although the model chi-square test is almost always significant in large samples, and thus results in high false rejection rates, chi-square values were reported along with other indices for comparison purposes.CFI and TLI values greater than or equal to 0.90 are indicative of an adequate fit, and values greater than 0.95 are considered a good fit (Hu & Bentler, 1999).RMSEA and SRMR values less than 0.05 indicate a good fit, and values between 0.05 and 0.08 indicate an adequate fit (Kline, 2013).
After establishing adequate fit, we used multigroup confirmatory factor analysis (MGCFA) to test the final model's invariance across gender and racial/ethnic groups.emotional neglect.Minimization/denial also correlated with depression and anxiety scores.
The second objective of this study was to test the invariance of the final CFA model (Model 2).The results of the measurement invariance tests are presented in Table 3.In general, our findings provide evidence for configural, metric, and scalar invariance of the CTQ-SF across genders and two racial/ethnic groups when the measure is modeled to have a five-factor structure.
Concerning invariance across gender, for the configural model, the fit statistics provided evidence of a good fit to the data (see Table 3), suggesting that the factors are adequately defined and measured for men and women.Subsequent models imposed equality constraints across genders for factor loadings (metric invariance) and intercepts (scalar invariance).Factor loadings, and intercepts, all indicate that the CTQ-SF is invariant across gender.These findings suggest that comparisons of latent means and manifest means using the CTQ-SF are justified.
To test whether the five-factor structure is invariant across race/ethnicity, a series of nested models were examined, as was done for gender.Again, the configural model fit the data reasonably well (see Table 3).The changes in CFI and RMSEA indicate that the CTQ-SF demonstrates equivalent factor loadings and intercepts across Black and White American groups.

Discussion
This study's primary goal was to provide psychometric evidence regarding the Childhood Trauma Questionnaire -Short Form, including the extent to which the instrument reliably assesses for childhood trauma in men and women as well as Black and White Americans.Consistent with our predictions, which were guided by the original measurement  To our knowledge, this is the first study to assess the CTQ-SF's generalizability to diverse individuals with measurement invariance testing in a population survey that includes assessment of inflammatory mediators of disease.We found that CTQ-SF trauma scores positively correlate with proinflammatory cytokines (i.e., IL-6 and CRP) as well as to model proposed by Bernstein et al. (2003), our results support a five-factor solution for the CTQ-SF.In addition, we have shown that CTQ-SF items adequately measure emotional, physical, and sexual abuse, as well as emotional and physical neglect.We also found that the CTQ-SF, which has been somewhat inconsistent in previous studies with regard to racial and gender differences, can be reliably used for men and women and across the two racial groups.Overall, our results suggest that CTQ-SF scores can be attributed hostile and aggressive behaviors) as appropriate responses to fear, anxiety, and powerlessness, including feeling emotionally vulnerable when seeking treatment.The recognition and accurate assessment of trauma is also vital to mitigating the high rates of adverse health consequences attributed to trauma, including those that pose physiological risks for morbidity and mortality.
Findings from the current study should be interpreted in light of the study's limitations.First, the study relied on secondary data collected from the MIDUS Refresher Biomarker Study, which may not necessarily generalize to all individuals who have experienced trauma as children.In addition, the study is correlational.Thus, conclusions regarding the causal influence of trauma on physical and mental health, including the range of symptom severity, cannot be directly confirmed by our findings.Finally, the current investigation focused on the associations between trauma and depression/ anxiety; however, it did not address the interactions and additive effects of depression and anxiety.Several scholars assert that depression and anxiety are correlated and independent (i.e., mediator and moderator variables) pathways to illness that have measurable effects on trauma recovery and health outcomes.

Conclusion
An accurate assessment of childhood trauma is vital for the prevention of its negative impacts.Without recognition and intervention, trauma is likely to continue to cause insidious damage to children, families, and communities.Retrospective assessment of childhood trauma may help to identify and, in turn, mitigate many of the common responses to early trauma and ACEs such as violence and incarceration, addiction, and suicide by strengthening safe, supportive, and trusting relationships, self-efficacy and competence, and faith and optimism.Our study confirms that the CTQ-SF is a valid and reliable measure of childhood trauma that can be used effectively across diverse communities.
The presence of trauma and deprivation in early life has substantial implications for health-related interventions and clinical research.Attention to the evaluative criteria associated with early experiences of trauma may give rise to studies that further explore the associations between trauma and childhood in ways that can enrich the advocacy, assessment, and treatment of individuals who report having ACEs and interpersonal traumas.Deepening our knowledge of childhood trauma concerning psychological constructs related to prevention and recovery would advance our understanding of human emotional experience, including those that give rise to positive adaptations and resilience.Future research should consider the development of interventions seeking reports of anxiety and depression.These significant correlations provide evidence of the CTQ-SF's convergent validity and suggest that individuals with trauma are at risk of confronting mental health problems in addition to their traumas, which likely carry additive health consequences.These findings are consistent with previous research confirming that a broad range of psychological and medical problems can be traced to physiological disturbances associated with adverse childhood experiences.Many studies suggest that individuals who have experienced childhood traumas tend to be vulnerable to chronic illness because of impaired innate immunity, which has been linked to chronic physiological markers of disease (Koball et al., 2019;Pedrotti Moreira et al., 2018;Salihoğlu et al., 2018).Although the CTQSF appears to be associated with changes in IL-6 and CRP, there was no correlation between CTQ-SF scores and TNF-α.This lack of connection between trauma and TNF-α may be due to differences in the underlying mechanisms of TNF-α (as compared to those of IL-6 and CRP).
Along with results from many other studies, our findings suggest that ACEs are particularly damaging because they jeopardize physical and emotional health, which may cause these individuals to experience the world as dangerous and threatening, regardless of their circumstances and possible opportunities for loving, caring, and supportive relationships.Individuals who have experienced ACEs and trauma often work to avoid the appearance of emotional vulnerability (because they distrust others, likely because they have attachment disturbances originating from their traumas).In turn, they may reject help from others, including healthcare providers, teachers, coworkers, and friends.The added influences of depression, anxiety, and physiological distress likely cause these individuals to withdraw, avoid, and emotionally numb themselves (e.g., through alcohol and/or drug abuse, under-or overconsumption of food, self-mutilation, and/or suicide attempts; van Bentum et al., 2022), which further complicate their recovery from trauma.
Without a trauma-informed framework that underscores universal screening practices and mainstream training and education efforts, these individuals will likely feel subjugated, stigmatized, and misunderstood.The accurate assessment of early trauma has the potential to improve the lives of many individuals who suffer in silence, often with profound emotional pain stemming from their disturbing childhood memories.Implementing universal screening in mainstream healthcare practice with measurement tools like the CTQ-SF has the potential to promote awareness of the commonality of trauma, and in turn, advance recovery efforts.Likewise, universal screening may help mitigate mental health stigmas associated with trauma, primarily when providers and trauma-informed advocates aid in reframing any behavioral disturbances that may arise (e.g., p < .05;**p < .01;TNF-α = Tumor Necrosis Factor alpha to real differences in reports of childhood trauma and not extraneous factors or differing conceptualizations across respondents.

Table 1
Means, standard deviations, and correlations among variables

Table 2
Summary of Fit Statistics for Testing the Measurement Invariance of the Five-Factor Model of CTQ-SF by Gender and by Race/Ethnicity CFI = comparative fit index; ΔCFI = change in the comparative fit index; RMSEA = root mean square error of approximation; ΔRMSEA = change in root mean square error of approximation; Δχ 2 (df ) = change in chi-square; df = degrees of freedom; BIC = Bayesian Information Criterion

Table 3
Standardized factor loadings for the 25-item Childhood Trauma Questionnaire-Short Form