Introduction

While for most women giving birth is a happy and positive event, 20 to 48% of women worldwide experienced childbirth as traumatic (Simpson & Catling, 2016). In some cases, negative experiences associated with childbirth may lead to the development of childbirth-related post-traumatic stress symptoms (CB-PTSS) and disorder (CB-PTSD) (Heyne et al., 2022). CB-PTSS/PTSD has been a subject of increasing interest by the scientific community, although there is no consensus regarding its definition (Akik & Batigun, 2020; Ayers et al., 2018; Horesh et al., 2021). The most observed symptomatology in mothers experiencing CB-PTSD includes birth-related symptoms of reliving the traumatic event (e.g., nightmares about childbirth, flashbacks about the moment of childbirth) and persistent avoidance (e.g., avoiding thinking about the birth and things that remind the birth), as well as general symptoms of arousal (e.g., annoyance, impaired concentration, and memory) (Ayers et al., 2018; Stramrood et al., 2010).

Approximately 4.7% of women giving birth are identified with CB-PTSD and 12.3% with CB-PTSS (Heyne et al., 2022). Among at-risk groups (e.g., mothers reporting traumatic delivery or infant complications), these estimates rise to 6.8–21.1% (Dekel et al., 2017; Grekin & O’Hara, 2014; Heyne et al., 2022). CB-PTSS has well-documented pervasive consequences on the mother-infant bonding, child development, as well as on fathers, even when mothers did not full diagnostic criteria for PTSD (Heyne et al., 2022; Stuijfzand et al., 2020; Van Sieleghem et al., 2022). Besides, it is highly comorbid with postpartum depressive symptoms (Andersen et al., 2012; Grekin & O’Hara, 2014).

Due to its prevalence and negative impact, risk factors for CB-PTSS/PTSD have been widely studied and grouped into prenatal vulnerability factors (e.g., depression in pregnancy, history of psychological problems and traumatic experiences), birth-related factors (e.g., obstetrical emergencies such as emergency c-section and vaginal instrumental birth, subjective distress in labor), and postnatal factors (e.g., depression and other comorbid symptoms, infant complications) (Anderson et al., 2012; Ayers et al., 2016; Grekin & O’Hara, 2014). Despite the recent developments in this field, CB-PTSS/PTSD remains poorly recognized and detected in obstetric settings (Moran Vozar et al., 2021), which may be due to the absence of routine screening (such as the one already implemented for postpartum depression in many countries). In line with this, the timely identification of CB-PTSS/PTSD through screening approaches based on brief, validated scales, has been recommended (Akik & Batigun, 2020; Ayers et al., 2018; Heyne et al., 2022; Moran Vozar et al., 2021).

Recent reviews revealed that a diversity of instruments is used to assess PTSS or PTSD in the postpartum period, but most of them assess general PTSD symptoms rather than specific symptoms related to the impact of traumatic childbirth experiences (Henye et al., 2022; Williams et al., 2022). This is an important limitation because there is evidence that CB-PTSS/PTSD differs from other traumatic events like childhood sexual abuse, which might reflect a distinct symptom structure. For instance, CB-PTSS/PTSD cover additional specific symptoms (e.g., intrusive thoughts or nightmares related to childbirth) and not only symptoms of PTSD that are normative experiences in the postpartum period (e.g., sleep interference) (Grekin & O’Hara, 2014). Accordingly, the development and validation of more accurate measures of CB-PTSS/PTSD has been suggested (Ayers et al., 2018; Henye et al., 2022; Horesh et al., 2021; Simpson & Catling, 2016).

The Perinatal PTSD Questionnaire (PPQ; DeMier et al., 1996; Quinnell & Hynan, 1999) represents a good option for addressing this gap and priority. The PPQ is a 14-item self-report questionnaire that was originally developed to assess post-traumatic stress symptoms related to childbirth among at-risk mothers. The PPQ was later modified by Callahan et al. (2006) to convert the original dichotomous response scale into a Likert scale format. This adaptation allowed a wider range of possible scores, without extending the measure, and improving its reliability and predictive utility to identify mothers that may be referred to mental health services. Despite its original three-factor structure meeting the diagnostic PTSD criteria specified in the Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV-TR (APA, 2000), the PPQ-II comprises items that measure each of the four symptom clusters specified in the latest version of DSM (i.e., re-experiencing of the traumatic event, cognitive and behavioral avoidance, negative alterations in mood and cognitions, and hyperarousal; APA, 2022). The PPQ and PPQ-II have been translated and validated across several countries (i.e., France, Spain, Turkey, Africa, Korea, and China) and, despite some divergences regarding its factor structure, research supported its good psychometric properties and cross-cultural utility (Akik & Batigun, 2020; Gondwe et al., 2018; Hernández-Martínez et al., 2021; Park et al., 2016; Pierrehumbert et al., 2004; Zhang et al., 2017).

Since the experience and rates of CB-PTSS/PTSD are likely to vary across countries (Heyne et al., 2022) and the cultural context could potentially impact CB-PTSS/PTSD measurement (Williams et al., 2022), testing the extent to which the PPQ-II fits each cultural context is deemed necessary to avoid misclassification of symptoms. To date, as far as we know, a validated Portuguese version of the PPQ-II is not yet available. Since Portuguese is spoken in 18 countries and is the official language of 10 (WorldData, 2023), a Portuguese version of the PPQ-II would enhance research on CB-PTSS in Portuguese-speaking countries, contributing to cross-cultural comparisons. This study aims to examine the factor structure and psychometric properties of the PPQ-II in European Portuguese mothers.

Method

Participants

A total of 451 Portuguese mothers living in Portugal, who were at 1 to 18 months postpartum (M infant’s age = 7.06 months), were included in this study. Mothers had an overall age of 32.84 years, and mostly had a university/postgraduate degree, were employed, had a monthly household below 1583€, and were either married, cohabitating, or partnered without living together. Most mothers were primiparous and had no prior psychological problems; they reported gestational complications and most children were born post-term through vaginal delivery. Children stayed at the maternity/hospital for an average of 3.70 days and only a few of them present current health problems (Table 1).

Table 1 Sociodemographic and clinical characteristics of the sample (N = 451)
Full size table

Procedure

Data collection occurred between February and March of 2020, in the context of a larger research project focused on risk and protective factors for mothers’ adjustment to motherhood in the postpartum period. The study was approved by the Ethics and Deontology Committee of the School of Psychology and Life Sciences of Lusófona University. A convenience sample was recruited through an online survey advertised on social networks (e.g., Internet blogs and Facebook® groups addressing motherhood topics). Participants were eligible to participate if they: (a) were mothers who have given birth of a single baby in the past 24 months; (b) conceived the baby in the context of a different-sex relationship; (c) were 18 years or older; and (d) were able to read and understand Portuguese. The survey contained a detailed description of the study, including ethical considerations, and participants given their consent to participate by clicking on the consent box before starting to answer the assessment protocol. Participation was voluntary and not remunerated.

A total of 556 mothers completed the survey. Of those, six were excluded because of more than 20% of missings in the study variables and 24 were excluded because they didn’t have a Portuguese nationality and/or were not living in Portugal. Additionally, in line with the original and previous validation studies of the PPQ-II, only mothers with infants between 1 and 18 months were included in this study (N = 451).

Measures

Sociodemographic, obstetric, and infant-related data. Mothers completed a self-report questionnaire specifically developed for this study assessing sociodemographic, obstetric and infant’s data.

Modified perinatal PTSD questionnaire. Childbirth-related posttraumatic stress symptoms were measured using the Modified perinatal PTSD questionnaire (PPQ-II; Callahan et al., 2006). The original PPQ-II is composed by 14-items divided in 3 subscales: intrusiveness, composed by the first 3 items (“Did you have bad dreams of giving birth or of your baby’s hospital stay?”), avoidant behaviors, composed by the following six items (“Were you unable to remember parts of your baby’s hospital stay?”) and hyper-arousal, composed of the last five items (“Were you more irritable or angry with others than usual?”). Participants were instructed to rate the items regarding their experience with their child over the past month, using a 5- point Likert-Like Scale (“0 = never, 1 = rarely, 2 = sometimes, 3 = Frequently, but only this past month, 4 = Frequently and for more than a month”). The total score for the PPQ-II ranges from 0 to 56, with higher scores indicating higher levels of symptoms. The original version of PPQ-II revealed good internal consistency (α = 0.90) and psychometric properties (Callahan et al., 2006). Permission from the authors to translate the PPQ-II to European Portuguese was obtained. Two independent researchers translated the original scale into Portuguese and a first merged version was developed. The translation process occurred without major difficulties or adjustments face to the original version. The translated version was further translated into English and compared to the original one to ensure semantic consistency.

Depressive, anxiety, and stress symptoms. Mothers’ depressive, anxiety and stress symptoms were assessed with the Portuguese version of Depression Anxiety Stress Scales (DASS-21; Pais-Ribeiro et al., 2004), which was used to assess the convergent validity of the PPQ-II. It was originally developed by Lovibond and Lovibond (1995) and comprised 42 items (DASS-42) organized into three dimensions (depression, anxiety, and stress). The DASS-21 is a shortened version of the DASS-42 and evaluated the same dimensions that the DASS-42. Each subscale is constituted by 7 items answered on a 4-point Likert scale ranging from 0 (did not apply to me at all) to 3 (applied to me very much, or most of the time). Participants are instructed to rate the items thinking about their experience over the past week. A total score for each subscale and overall scale can be computed by summing up the respective items. Higher scores indicate higher levels of symptoms. To allow comparison with the DASS-42, participants’ scores were multiplied by two (range 0–42). Cronbach’s alphas in our sample were 0.88 (depression subscale), 0.82 (anxiety subscale), and 0.89 (stress subscale).

Data preparation and analysis

Statistical analyses were performed using SPSS version 28.0 and JASP open-source software. Descriptive statistics were computed for sample characterization. We used a two-stage process to explore and validate the factorial structure of the European Portuguese version of the PPQ-II. The total sample was randomly split into two subsamples through the random sample selection procedure in SPSS. The first subsample (n = 229; calibration sample) was used to compute an exploratory factor analysis (EFA) and the second subsample (n = 222; validation sample) was used to validate the observed factor structure through a confirmatory factor analysis (CFA). Both subsamples showed an appropriate size (Costello & Osborne, 2005) and they also presented similar socio-demographic, clinical and obstetric characteristics (all ps > 0.05).

To determine the suitability to conduct an EFA, we first examined the Kaiser–Meyer–Olkin’s (KMO) coefficient and Bartlett’s test (Tabachnick & Fidell, 2013). An EFA was therefore performed on JASP, using the Principal Axis Factoring (PAF) method to explore the factorial structure underlying the European Portuguese version of the PPQ-II, with an oblique oblimin rotation because the factors were expected to be correlated. The Parallel Analysis (PA) method, combined with the Guttman-Kaiser criterion (i.e., eigenvalue > 1.00), was applied to determine the number of factors to retain. Factor loadings of 32 or above and items with communalities of 0.3 or above were considered meaningful (Tabachnick & Fidell, 2013); items that did not full these criteria were eliminated. A CFA using maximum-likelihood estimation was performed in the validation sample (n = 222) using JASP to confirm the factor structure found in the calibration sample. Beyond considering a nonsignificant chi-square as indicative of good model fit, additional criteria were used to examine the overall fit of the model: a Comparative Fit Index (CFI) and Tucker-Lewis index (TLI) ≥ 0.95 and Root-Mean-Square Error of Approximation (RMSEA) ≤ 0.05 (Hu & Bentler, 1998).

The internal consistency was estimated in both the calibration and validation subsamples, through the McDonald’ omega coefficient (ωt) (Deng & Chan, 2017). We also examined the average inter-item correlation coefficients, the average item-rest correlation, and McDonald’ omega if the item deleted. A coefficient of McDonald’s ω > 0.70, item-rest correlation > 0.30, and inter-item correlation between 0.15 and 0.50 (and < 0.80) are indicators of good internal consistency (Field, 2005; Kline, 1998). Convergent validity of the PPQ-II was tested in the validation sample through computing Pearson correlations (r) between the PPQ-II and depressive, anxiety and stress symptoms. Pearson correlations were interpreted as follows: small: r ≥ .10; medium: r ≥ .30; large: r ≥ .50 (Cohen, 1988). Besides, high-risk vs. low-risk subsamples based on well-established risk factors for CB-PTSS (i.e., prematurity, mode of delivery – vaginal delivery, instrumental vaginal delivery, planned cesarean section, emergency cesarean section – and baby’s postnatal health complications; yes/no scale) were compared on the PPQ-II to estimate the known-groups validity of the scale (i.e., the extent to which the scale discriminated those subgroups) through independent t tests computed in the whole sample (N = 451). For all analyses, significance was set at the level p < .05.

Results

Exploratory and confirmatory factor analysis

The KMO’s coefficient (0.83) and the Bartlett Sphericity test (χ2(55) = 963.99, p < .001) suggested that the data were appropriate for EFA. The PA, in conjunction with the Guttman-Kaiser criterion, suggested a two-factor structure, explaining 48.3% of the total variance. Three items with factor loadings < 0.32 and/or communalities < 0.30 were subsequently excluded from the further analyses (items 3, 6, and 14 in the original version), resulting in a 11-item scale (Table 2).

Table 2 Exploratory Factor Analysis: Factor loadings and communalities for each item (n = 229)
Full size table

Item loadings ranged from 0.52 to 0.84. Factor 1 was labeled “Arousal” and includes 7 items that explain 30.3% of the variance. Factor 2 was labeled “Avoidance and intrusion” and includes 4 items that explain 18.0% of the variance. The CFA applied to the two-factor structure model obtained in the calibration sample showed an acceptable fit to the data (χ2(43) = 103.53, p < .001, CFI = 0.95; TLI = 0.94; RMSEA = 0.08, 90% CI [0.06, 0.10], p < .01). As shown in Fig. 1, standardized factor loadings were high (ranging between 0.66 and 0.84) and statistically significant (all p < .001).

Fig. 1
figure 1

Confirmatory factor analysis of the European Portuguese version of the PPQ-II. Note. Standardized factor loadings are presented (all p < .001)

Full size image

To obtain more information about the validity of this factorial structure, we choose testing an additional EFA and CFA in the whole sample (N = 451). We found similar results as the ones obtained in the calibration and validation subsamples, demonstrating that the 11-item two-factor structure (with comparable factor loadings for each item) was the solution that best fitted the data, accounting for 53.3% of the total variance, with good values of goodness-of-fit on the CFA (χ2(43) = 154.41, p < .001, CFI = 0.95; TLI = 0.93; RMSEA = 0.076, 90% CI [0.06, 0.09], p < .01).

Reliability

The values of McDonald’s ω demonstrated that both Factor 1 and Factor 2 of the PPQ-II have very good internal consistency in both the calibration and validation subsamples. Concerning McDonald’s ω if item deleted, statistics did not suggest that the removal of any item would increase the total McDonald’s ω value. The inter-item correlations in the validation sample exceeds the recommended range of 0.15 and 0.50; however, they were both below.80, so they are still acceptable (Kline, 1998). All the item-rest correlations were above the recommended cut-off point of 0.30, ranging from 45 (item 1) to 75 (item 11). The McDonald’s ω value for the total scale of the mPPQ was 0.86 and 0.88 in the calibration and validation samples, respectively (Table 3).

Table 3 Reliability of the PPQ-II in the calibration and validation samples
Full size table

Convergent and known-groups validity

Supporting convergent validity, strong, significant, and positive correlations were found between the PPQ-II (total score and Factor 1) and depressive, anxiety, and stress subscales of the DASS-21, ranging from 0.62 to 0.74, p < .001. The correlations between Factor 2 of the PPQ-II and all subscales of DASS-21 were also significant and positive, but small in magnitude (Table 4).

Table 4 Convergent validity: Correlations between PPQ-II (total and subscales) and DASS-21 subscales (n = 222)
Full size table

Overall, the PPQ-II also demonstrated substantive known-groups validity, as it discriminated among mothers who have babies with postnatal health problems from those with healthy babies (except Factor 2, p > .05), with the first reporting higher levels of CB-PTSS, as well as those who had vs. didn’t have an instrumental vaginal birth, with the first reporting higher overall levels of CB-PTSS (Table 5).

Table 5 Known-groups validity: Comparison of high-risk vs. low-risk mothers on the PPQ-II factors and total score (N = 451)
Full size table

Overall, the PPQ-II did not discriminate between mothers who had a preterm vs. postterm birth (except Factor 2; M preterm birth: 1.44, SD = 3.45 vs. M postterm birth: 0.60, SD = 1.82, t(449) = -2.18, p < .05), as well as those mothers who had a planned vs. unplanned cesarean section (except Factor 2; M planned cesarean: 0.15, SD = 0.81 vs. M unplanned cesarean: 0.77, SD = 2.13, t(449) = -2.61, p < .01). Mothers with preterm babies and who had an unplanned cesarean section reported higher scores on Factor 2.

Discussion

The main goal of the present study was to explore the factor structure and psychometric properties of the PPQ-II in European Portuguese mothers. Following original and prior validations studies, the results showed adequate validity and reliability of the scale and raise some important issues that are discussed below.

First, contrary to the original studies (Callahan et al., 2006), our results suggest that the scale comprises two rather than three factors. Inconsistencies regarding the factor structure of the PPQ-II have been observed, in terms of number of factors and corresponding items. For instance, while other validation studies support the original three-factor structure (Hernández-Martínez et al., 2021; Park et al., 2016; Zhang et al., 2017), the items that pertain to each factor vary between studies. Even in the original studies, authors reported unexpected item loadings (e.g., item 2 referring to upsetting memories loaded on the avoidance factor; Callahan et al., 2006). Methodological issues could explain these disparities across studies. For instance, variation may be attributed to differences in sample selection (targeted vs. non-targeted samples), version of the PPQ (original dichotomous scale vs. Likert-scale), and measurement time point (up to 6 months vs. 18 months postpartum). Since our aim was to validate the PPQ-II as a universal screening measure, we choose to conduct the analyses in a non-targeted sample.

Our observed two-factor structure is aligned with the Turkish (Akik & Batigun, 2020) and French (Pierrehumbert et al., 2004) versions, though in these versions the factor Intrusion and avoidance includes two more items (which were excluded in our study). In line with the conclusions of Callahan et al. (2006), it may be that avoidance behaviors may be more predominant in those women with persistent reexperiencing as an attempt to cope with such intrusions, which could explain that a unique factor, comprising intrusion and avoidance behaviors, emerged from the analyses in our and prior studies. Consistent with this, two symptom clusters reflecting childbirth-related symptoms of re-experiencing and avoidance and general symptoms of hyperarousal have emerged in the factor structure of measures specifically developed to assess CB-PTSD (Stramrood et al., 2010), including according to DSM-5 criteria (e.g., the City Birth Trauma; Ayers et al., 2018). Inconsistencies between diagnostic clusters specified in mental disorders classifications such as the DSM and symptom structure observed in measures assessing CB-PTSD symptoms have been reported (Ayers et al., 2018), supporting the complexity and distinctiveness of CB-PTSS/PTSD (Boorman et al., 2014; Horesh et al., 2021). Perhaps for this reason, as highlighted previously by Akik and Batigun (2020), most studies using the PPQ/PPQ-II have relied on its total score rather than sub-scales (for a review see Heyne et al., 2022).

Our results also suggest the exclusion of three items, including item 6 (difficulties to remember details of the birth), which has shown to be particularly problematic in other validation studies of the PPQ-II (Akik & Batigun, 2020; Callahan et al., 2006; Zhang et al., 2017). Interestingly, the same item content did not load in neither factor in other instruments of CB-PTSS/PTSD (Ayers et al., 2018), which suggest that the nature of childbirth, as a highly physiological, sensory event might lead to stronger traumatic memories that made it easily remembered, in comparison to other traumatic experiences (Harrison et al., 2021). Interestingly, the factor Arousal was the most consistent across studies, being composed by the same 7 items in the Spanish (Hernández-Martinez et al., 2021), Turkish (Akik & Batigun, 2020), Korean (Park et al., 2016), and Chinese (Zhang et al., 2017) validations. Besides, this is the factor with the strongest association with postpartum depressive, anxiety, and stress symptoms, which is in line with prior research showing the strong relationship between symptoms of PTSD, more specifically regarding the symptoms of hyperarousal and numbness, and postpartum depression (Andersen et al., 2012; Grekin & O’Hara, 2014; Watson, 2009). In fact, the correlations between the total score and both factors of the PPQ-II and psychological symptoms were found to occur in the hypothesized directions.

Concerning the scale’s reliability, we used an alternative and more sensitive indicator (i.e., the McDonald’ omega) (Deng & Chan, 2017) to estimate the internal consistency of the PPQ-II total score and factors. Our results indicated good homogeneity of the PPQ-II, with reliability values for the total score closer to the original (α = 0.90; Callahan et al., 2006) and validation (e.g., Turkey and Spanish versions: α = 0.89 and 0.90, respectively) studies. The values for the factors were also good and closer to prior validation studies. These results seem to support the conclusion of Callahan et al. (2006) that the Likert scale format response improves the internal consistency of the scale, as the unique validation with poor reliability values is the French version (Pierrehumbert et al., 2004), which used the dichotomous scale.

Like studies of risk factors for CB-PTSS/PTSD (Andersen et al., 2012; Grekin & O’Hara, 2014), deliveries involving complications and health issues of the infant were also significantly associated with CB-PTSS as measured by the PPQ-II. Our results suggested that mothers who had an instrumental vaginal birth and babies with postnatal health problems are at increased risk of developing CB-PTSS; however, in contrast to previous findings (e.g., Akik & Batigun, 2020; Callahan et al., 2006; Quinnell & Hynan, 1999), experiencing a preterm birth and an unplanned cesarean section were not associated with overall CB-PTSS neither hyperarousal. Rather, such birth characteristics seem to constitute specific risk factors for mothers experiencing nightmares and intrusive thoughts about childbirth, as well as avoiding behaviors related to childbirth and the child. Considering that a preterm birth is commonly followed by the stay of the infant in intensive neonatal care and related uncertainty and helplessness regarding the infant’s life, and that preterm infants have a higher likelihood of neurodevelopmental problems, we could argue that these factors can serve as constant reminders of a traumatic birth and then specifically triggers re-experiencing and avoidance symptomatology, which may last several years postpartum (Ahlund et al., 2009). In a similar way, intrusive traumatic memories are particularly prevalent among women who undergo an emergency cesarean section, which combines an unexpected surgery and delivery (Horsch et al., 2017).

Implications for practice and research

This study entails relevant implications for midwifery care. Following prior conclusions across countries, the present study suggested that the PPQ-II is a short, valid, and reliable scale for measuring CB-PTSS experienced by postpartum women, encouraging its routine use in Portuguese maternity and primary care settings. The immediate period following childbirth represents a window of opportunity to screen for the presence of CB-PTSS as women are routinely followed to 6 weeks after childbirth and services providing psychological support are available in major Public Maternity Hospitals and General Hospitals upon women request. The PPQ-II is brief, simple, and easy to administer, which makes it possible to be easily integrated into routine screening practices of perinatal mental health problems already implemented (e.g., jointly with the Edinburgh Postnatal Depression Scale that is widely used for postpartum depression screening). This is of particular importance considering the high comorbidity between postpartum depression and CB-PTSS/PTSD; therefore, women reporting clinically relevant depressive symptoms after childbirth should be more wisely assessed for possible PTSS related to childbirth. This would allow earlier recognition of CB-PTSS and targeted interventions to prevent the pervasive consequences of CB-PTSS/PTSD on mothers, infants, and the entire family.

Some issues are needed to be taken into account when considering the routine use of the PPQ-II into obstetric settings. First, while the Portuguese version of the scale supports the computation of two subscales, in accordance with some other validations studies of the PPQ-II (Akik & Batigun, 2020; Pierrehumbert et al., 2004), as well as the increased evidence of two clusters of symptoms (Ayers et al., 2018; Stramrood et al., 2010), the use of a total score of CB-PTSS is advised attending to the diversity in the factor structure of the scale observed worldwide. Besides, although the PPQ-II contains items assessing symptoms of the four clusters defined in DSM-5-TR (APA, 2022), it should be viewed as a measure that evaluates specific and general CB-PTSS rather than a screening tool that mirrors the PTSD diagnostic framework. Second, our results suggest that health professionals such as midwives might be wise to apply the PPQ-II among mothers with preterm infants and/or with medical complications, as well as those with operative births, as they are at increased risk of developing CB-PTSS/PTSD. Guidelines about how to conduct this assessment and further steps to guarantee appropriate follow-up and referral, namely the extent to which the PPQ-II could be included as part of a stepped care model approach for the identification, prevention, and treatment of perinatal mental health problems, are beyond the scope of this article but research addressing these issues is encouraged to draw more solid conclusions for clinical practice.

This study comprises some limitations that should be acknowledged. First, it relies on a sample of postpartum women recruited from the Internet, which may limit the generalization of the findings. However, this study comprised an ample and heterogenous sample size in terms of sociodemographic and clinical characteristics, which allowed conducting comparisons across groups. Second, its cross-sectional nature did not allow us to draw solid conclusions regarding the development of and underlying risk factors for CB-PTSS. Third, it did not examine reliability over time nor the predictive validity of the PPQ-II for identifying women with a clinical diagnosis of CB-PTSD. The original validation study of the PPQ-II suggests a cut-off score of 19 as evidence of clinically relevant distress that may justify the need of referral to treatment (Callahan et al., 2006). While this value has been used in other countries, the establishment of culturally sensitive cut-off for screening measures has been advised (Kozinszky & Dudas, 2015). Further research considering a longitudinal, prospective design including a gold standard (namely a structured, diagnostic interview) would allow to establish cut-off/diagnostic criteria for clinically relevant CB-PTSS or PTSD among other populations, including for Portuguese ones. This would enhance the clinical application of the scale.

Despite these limitations, this study extended prior research through unique contributions to this research field. First, along with the versions available around the world, this study adds support to the cross-cultural adequacy of the PPQ-II. Its Portuguese version is of utmost relevance for non-Portuguese researchers to establish comparison across countries and cultures (e.g., regarding the prevalence of CB-PTSS). Second, it provides further evidence concerning the conceptualization of CB-PTSS/PTSD involving two-symptom clusters, which support the hypotheses that this symptomatology may represent a unique condition (Horesh et al., 2021), with a symptom structure that should be understood beyond the clusters specified in common classifications of mental health disorders. Third, due to its briefness, the PPQ-II can be incorporated into research protocols, such as research trials aimed to study the efficacy of interventions for CB-PTSS/PTSD, without adding substantial burden to participants.