How are experiences and acceptability of child maltreatment related to resilience and posttraumatic growth: a cross cultural study

ABSTRACT

Background: Post-traumatic growth (PTG) and resilience, regarded as positive psychological change following a traumatic experience, are under-researched across cultures in people exposed to child maltreatment (CM).

Objective: We investigated how experiences and the perceived acceptability of CM are related to resilience and PTG in countries with different cultures, living standards, and gross national income.

Method: A total of 478 adults from Cameroon (n = 111), Canada (n = 137), Japan (n = 108), and Germany (n = 122) completed an online survey with self-reported questionnaires, including the Brief Resilience Scale and the Post Traumatic Growth Inventory-Short Form.

Results: Across countries, self-reported male gender and age were positively associated with resilience, while experiences of physical abuse and emotional maltreatment were negatively associated with resilience. Experiences of emotional maltreatment were positively associated with PTG. Higher levels of PTG and resilience were found amongst Cameroonian participants as compared to other countries.

Conclusion: Our results suggest that positive changes following CM can vary significantly across cultures and that experiences of specific CM subtypes, but not the perceived acceptability of CM, may be important for a deeper understanding of how individuals overcome trauma and develop salutogenic outcomes. Our findings may inform CM intervention programmes for an enhanced cultural sensitivity.

HIGHLIGHTS

• Across the four countries (Canada, Cameroon, Germany, Japan), more experiences of physical abuse and emotional maltreatment were associated with lower resilience; more experiences of emotional maltreatment were associated with greater post-traumatic growth.

• Higher levels of post-traumatic growth and resilience were found in Cameroon as compared to other countries.

• Positive changes following child maltreatment vary across cultures and experiences of specific child maltreatment subtypes, but the perceived acceptability of child maltreatment did not exert an influence on salutogenic post-traumatic outcomes.

Antecedentes: El crecimiento postraumático (PTG por sus siglas en ingles) y la resiliencia, considerados como un cambio psicológico positivo después de una experiencia traumática, no se han investigado lo suficiente en todas las culturas en personas expuestas a maltrato infantil (MI).

Objetivo: Investigamos cómo las experiencias y la aceptabilidad percibida del MI están relacionadas con la resiliencia y el PTG en países con diferentes culturas, niveles de vida e ingreso nacional bruto.

Método: Un total de 478 adultos de Camerún (n = 111), Canadá (n = 137), Japón (n = 108), y Alemania (n = 122) completaron una encuesta en línea con cuestionarios de auto-reporte, incluyendo la Escala Breve de Resiliencia y la versión corta del Inventario de Crecimiento Postraumático.

Resultados: En todos los países, el género masculino y la edad auto-informada se asociaron positivamente con resiliencia, mientras que las experiencias de abuso físico y maltrato emocional se asociaron negativamente con resiliencia. Las experiencias de maltrato emocional se asociaron positivamente con PTG. Se encontraron niveles más altos de PTG y resiliencia entre los participantes cameruneses comparado con los otros países.

Conclusión: Nuestros resultados sugieren que los cambios positivos después del MI pueden variar significativamente entre las culturas y que las experiencias de subtipos específicos de MI, pero no la aceptabilidad percibida del MI, pueden ser importantes para una comprensión más profunda de cómo las personas superan el trauma y desarrollan resultados salutogénicos. Nuestros hallazgos pueden informar los programas de intervención de MI para una mayor sensibilidad cultural.

背景：创伤后成长（PTG）和心理韧性被视为创伤经历后的积极心理变化，但在不同文化背景下，对遭受儿童虐待（CM）者的研究还不够。

目的：我们调查了在不同文化、生活水平和国民总收入的国家中，CM 的经验和感知可接受性与心理韧性和 PTG 之间的关系。

方法：来自喀麦隆 (n = 111)、加拿大 (n = 137)、日本 (n = 108) 和德国 (n = 122) 的总共 478 名成年人完成了一项自我报告问卷的在线调查，包括简要心理韧性量表和创伤后成长量表-简表。

结果：在各个国家中，自我报告的男性性别和年龄与心理韧性呈正相关，而身体虐待和情感虐待的经历与心理韧性呈负相关。情感虐待经历与 PTG 呈正相关。与其他国家相比，喀麦隆参与者表现出了更高水平的 PTG和韧性。

结论：我们的结果表明，CM 后的积极变化在不同文化中和特定 CM经历亚型可能存在显著差异，而不是 CM 的可接受性，对于更深入地了解个体如何克服创伤和发展有益结果可能很重要。我们的研究结果可以为 CM 干预计划提供信息，以增强文化敏感性。

KEYWORDS:

• Child maltreatment
• acceptability of child maltreatment
• perceived social norms
• resilience
• posttraumatic growth
• culture

PALABRAS CLAVE:

• Maltrato infantil
• aceptabilidad del maltrato infantil
• normas sociales percibidas
• resiliencia
• Crecimiento Postraumático
• cultura

关键词:

• 儿童虐待
• 儿童虐待的可接受性
• 感知到的社会规范
• 心理韧性
• 创伤后成长
• 文化

1. Introduction

Millions of children are estimated to be victims of child maltreatment (CM) (physical abuse, sexual abuse, emotional maltreatment, neglect, and exposure to domestic violence) each year across the globe (Stoltenborgh et al., 2015). CM is a major public health issue worldwide (Alenezi et al., 2022) given its strong association with several poor short and long-term outcomes including impairments in social functioning (Pfaltz et al., 2022), disordered psychological development and behavioural problems (Al Odhayani et al., 2013). Culture, which is defined as a way of life of a group of people (e.g. behaviours, beliefs, values, and symbols) and usually considered to be within the confines of a country (Taras et al., 2016), is central in the understanding of CM. Although CM is prevalent across cultures and societies, the definition of CM varies widely (Lansford et al., 2015) depending on perceived social norms (Wadji et al. 2023). For example, the use of physical punishment can be perceived as acceptable and necessary for child discipline in one culture but considered abusive elsewhere (Klika et al., 2019). Moreover, culture is associated with the potential for resilience and post-traumatic growth (PTG) in CM survivors (Gunnestad, 2010; Ungar, 2013). Research examining CM and its correlates from a cross-cultural perspective, including how CM is viewed (i.e. acceptability of CM) and how survivors of CM recover and even thrive in different cultures is necessary, but scarce. The current cross-cultural study aims to fill some of the aforementioned gaps by examining how experiences and the perceived acceptability of CM are related to resilience and PTG.

1.1. Resilience and post-traumatic growth

Beside adverse consequences, people may recover and even thrive after traumatic experiences (Shakespeare-Finch & Lurie-Beck, 2014), showing resilience and PTG (Tedeschi & Kilmer, 2005). Resilience is defined by the American Psychological Association Dictionary of Psychology as ‘successfully adapting to difficult or challenging life experiences, especially through mental, emotional, and behavioral flexibility and adjustment to external and internal demands’ (American Psychological Association, n.d.). PTG refers to individuals not only adapting to, but even experiencing positive psychological changes after trauma such as an increase in appreciation for life, in meaningful interpersonal relationships, and in perceived personal strength, and changed priorities (Tedeschi & Calhoun, 2004). PTG and resilience are distinct, yet correlated constructs (Kaye-Tzadok & Davidson-Arad, 2016).

Sociodemographic factors (e.g. gender, age, marital status, income level and education) appear to be associated with resilience and PTG, but mixed results have been found to date (Fox et al., 2009; Koutrouli et al., 2012; Weitzel et al., 2022). For example, some studies found that men are more resilient than women, while women reported higher PTG levels than men (Adjorlolo et al., 2022; Cohen-Louck, 2022). Other studies found higher resilience in women as compared to men (Linnemann et al., 2020). The cultural context in which the previous studies were carried out may explain this discrepancy.

Several studies have showed that individuals with histories of CM reported diminished resilience (Campbell-Sills et al., 2009; Nishimi et al., 2020) and PTG (Quan et al., 2022). Some people, on the other hand, show high resilience and PTG following CM (Louis, 2018; Ulloa et al., 2016). Variation remains, however, in the definition of CM from one culture to another and may account for some of the mixed results obtained to date. Research shows that it is important to take the cultural context into consideration, as it can influence how trauma is defined or perceived, as well as the ability to achieve positive outcomes such as resilience and PTG (Zheng et al., 2020a). Cross-cultural research is increasingly important to understand how people who have experienced CM develop resilience and PTG in different cultures, and to provide practitioners with tools to foster positive coping in individuals with histories of CM.

1.2. Resilience, post-traumatic growth, and child maltreatment through a cross-cultural lens

Evidence suggests culturally and contextually specific aspects that contribute to resilience (Ungar, 2006) and PTG (White et al., 2021). For example, positive religious coping and belief have been associated with greater PTG and resilience whereas negative religious coping increased distress (Schaefer et al., 2008). A previous study showed significant cultural differences in resilience capacity and the baseline of stress severity in the aftermath of trauma among populations from three regions: America, Hong Kong, and Mainland China (Zheng et al., 2020a). Familialism, a cultural value which emphasizes warm, close and supportive family relationships, was identified as a protective factor in non-Western cultures, likely to promote resilience and growth (Zheng et al., 2020b). Furthermore, in many low- and middle-income countries, people face harsh living conditions (e.g. poverty, war, violence) and struggle for survival on a daily basis (Lund et al., 2010). Therefore, resilience and PTG in the aftermath of potentially traumatic events may vary across low- and high-income countries (Vázquez et al., 2014).

Furthermore, parents in different cultures may harbour different beliefs about how a child should be raised (Bornstein, 2015). These social norms, or perceived acceptability of childrearing practices, can either prevent or perpetuate CM (Fleckman et al., 2019). For example, the use of physical punishment or violence (e.g. spanking, slapping, pushing, punching) is a commonly accepted way of disciplining a child in many low- and middle-income countries, like Sub-Saharan Africa and South Asia (Cuartas et al., 2019; Pace et al., 2019) whereas in high income countries, like USA, Germany, or Canada, it is prohibited and considered a violation of human rights (Zolotor & Puzia, 2010). Equally, the physical manifestation of healing practices resulting in skin marks (e.g. applying heated objects on the skin to treat and heal) can be confused with CM or misinterpreted depending on the culture (Kolhatkar & Berkowitz, 2014; Lupariello et al., 2020; Vashi et al., 2018). Certain cultural values (e.g. shame, taboos, respect, and patriarchy) may reduce the likelihood of CM disclosure, increasing the risk of prolonged exposure to abuse and neglect (Fontes & Plummer, 2010). Therefore, it is necessary to consider the cultural context in which CM occurs to better understand its effects. In this context, the perceived acceptability of CM as it relates to resilience and PTG was examined in the current study.

1.3. The present study

Cultural values and beliefs can influence both an individual's interpretation of a traumatic event and their coping strategies and abilities. So far, resilience and PTG have been in most cases investigated as mediators or moderators affecting the relationship between CM and post-traumatic outcomes (Jankovic et al., 2022; Kim et al., 2017). There is a paucity of research examining factors associated with resilience and PTG. Previous studies conducted in a few specific countries (e.g. USA) found that some sociodemographic factors (e.g. gender, education, income) (Bonanno et al., 2007) and reported histories of CM were associated with resilience (Campbell-Sills et al., 2009). However, the extent to which these findings generalise across countries, especially non-western ones, is unknown. Moreover, little is known about how experiences and the perceived acceptability of specific CM subtypes are associated with resilience and PTG. Such research would improve intervention efforts within diverse populations.

The current study aims at examining how CM-related factors may be associated with resilience and PTG across countries (Cameroon, Japan, Canada, Germany) representing different continents and cultures, living standards, and gross national income. We expect to find: (1) differences across countries in levels of resilience and PTG, and (2) that maltreatment-related and demographic factors will be significantly associated with resilience and PTG. No specific hypotheses are provided regarding cross-country differences in resilience and PTG, given the lack of previous research.

2. Method

2.1. Participants

General population samples were recruited in Cameroon, Japan, Canada, and Germany. Inclusion criteria were to reside in one of the four target countries and being 18 years and older. The languages used were French, English, Japanese, and German.

In Cameroon, Canada, and Germany, recruitment was conducted through postings on social media (e.g. Facebook, Messenger, LinkedIn, and Instagram) and flyers circulated through the networks of the investigators. In Japan, both convenience sampling through flyers posted in physical locations (e.g. universities, hospitals), and a survey through an online survey company were conducted. This additional recruitment strategy was needed to reach our target sample size in Japan.

2.2. Procedure

An online survey using Qualtrics took place in all four countries. In Japan, the additional panel survey was conducted using the Freeasy service. Initially, 728 participants completed the survey. However, prior to data analysis, we conducted a data quality check, given the online format of the study. Thus, we excluded participants who: (1) did not provide consent (n = 121), (2) completed < 75% of the survey (n = 114), (3) completed the survey in less than half of the modal time (6 mins) (n = 6), (4) reported a current country of residence not from the participating countries (n = 7), and (5) provided contradictory answers related to their country of residence (n = 2), resulting in a final N = 478. This study, conducted as part of the Global Collaboration on Traumatic Stress, a group of researchers and clinicians from around the world, was approved by local ethics committees in the four countries.

2.3. Measures

2.3.1. Experiences of child maltreatment

The neglect subscale of the ISPCAN Child Abuse Screening Tools Retrospective version (ICAST-R) (Dunne et al., 2009) was used (5 items). A sample item is: ‘have you ever been hurt or injured because no adult was supervising you?’ The original English, French, and German versions of the ICAST-R are validated. Back-translation was used for Japanese. Cronbach’s alpha for the neglect subscale of the ICAST-R in our sample was 0.68.

Physical abuse (5 items; e.g. ‘were you ever punched or kicked by a parent or caregiver’), emotional maltreatment (5 items; e.g. ‘were you often put down or ridiculed by a parent or caregiver?’) and sexual abuse (6 items; e.g. ‘were you ever forced or coerced to perform oral sex on someone against your will?’) were assessed using the Early Trauma Inventory Self-Report – Short version (ETISR-SF) (Bremner et al., 2007). The original English version of the ETISR-SF is validated; French and German versions were already available. Back-translation was used for Japanese. The internal consistency of the three subscales was high across the four countries: emotional maltreatment (α = .85), physical abuse (α = .73), and sexual abuse (α = .81).

In addition, we used a three-item scale for exposure to domestic violence, adapted from the Revised Conflict Tactics Scales 2 (CTS2) (Straus et al., 1996). A sample item is: ‘Prior to your 18th birthday was your mother or stepmother, often or very often pushed, grabbed, slapped, or had something thrown at her?’ The original English version of the CTS2 is validated; a French version was already available. Back-translation was used for Japanese, and German. Cronbach’s alpha across the four countries was 0.73.

CM items were answered in a yes/no format, and they were dummy coded to 1 = yes and 0 = no. Total scores were calculated by adding each item and used as continuous count scores for each CM subtype.

2.3.2. Acceptability of child maltreatment

Acceptability of CM was measured by asking participants to rate how acceptable certain behaviours are within their community on a five-point Likert scale from ‘never acceptable’ to ‘always acceptable’. Examples of some items are ‘When caregivers can afford it, how acceptable is it [in your community] that children are not taken care of when they are sick or injured?’ (neglect); ‘How acceptable is it that children are being punched or kicked by a parent or caregiver?’ (physical abuse); ‘How acceptable is it that children are being ignored or made to feel that they didn’t count (by an adult)?’ (emotional maltreatment); ‘How acceptable is it that someone under 18 years old is experiencing another person rubbing their genitals against them against their will?’ (sexual abuse); and ‘How acceptable is it that children are often or very often exposed to their mother or stepmother being pushed, grabbed, slapped, or having something thrown at her?’ (exposure to domestic violence). A mean score of acceptability (1–5) was calculated for each subscale. The Cronbach’s alphas in this study were high for all CM subtypes (α = .84 to .96).

2.3.3. Resilience

Resilience was assessed using the Brief Resilience Scale (BRS) (Smith et al., 2008), which includes six items rated on a five-point Likert scale from 1 – strongly disagree to 5 – strongly agree. The mean scores range from 1 to 5, with higher scores indicating higher levels of resilience. The English, French, German, and Japanese versions of the BRS are validated (Jacobs & Horsch, 2019; Chmitorz et al., 2018; Tokuyoshi & Moriya, 2015). Cronbach’s alpha of the current sample across the four countries was .88 (range from .77 and .92 in individual countries).

2.3.4. Post-traumatic growth

The Post Traumatic Growth Inventory-Short Form (PTGI-SF) (Cann et al., 2010) was used to measure PTG. PTGI-SF consists of 10 items, each rated on a six-point-Likert scale (0–5). The total sum score ranges from 0 to 50, with higher scores reflecting higher levels of PTG. Cronbach’s alpha across the four countries was .91 (range from .87 to .92 in individual countries). Although only the original English version of PTGI-SF is validated, the PTGI (21 items) is validated in French, German, and Japanese (Cadell et al., 2015; Maercker & Langner, 2001; Taku et al., 2007).

2.4. Data analysis

Analyses were performed using JMP version 16.2 and SPSS 29. For preliminary analysis, we examined the association between resilience and PTG using a Spearman correlation test that was conducted between BRS and PTGI-SF across countries. We also compared groups based on BRS and PTGI-SF scores and used a one-way ANOVA to calculate differences between countries in levels of resilience and PTG.

For our main objective – to examine factors associated with resilience and PTG - since BRS and PTGI-SF scores had normally distributed residuals, we conducted multiple linear regression analyses. In the country comparisons, dummy variables were used with either Japan or Cameroon as the reference. Thus, the multiple regressions were run twice. With regards to gender, male was the reference category (male = 0, female = 1). In Block 1 of the multiple regression, we entered the demographic variables and country as independent variables. In Block 2, we added CM acceptability scores and CM subtypes scores. The block approach allowed to quantify the added explained variance when introducing the CM variables (experiences and perceived acceptability). Variance Inflation Factors (VIF) and Tolerances were examined to confirm the absence of collinearity in the data.

3. Results

3.1. Preliminary analyses

3.1.1. Sociodemographic characteristics and of child maltreatment -related variables

Sociodemographic characteristics by countries are shown in Table 1. The mean age of the participants was significantly different among the four countries (χ² (3, N = 471) = 168, p < .01). Multiple comparisons using the Steel-Dwass method showed that the mean age in Japan was higher than that of the other three countries. Gender distributions differed between countries (χ² (12, N = 477) = 116.5, p < .01). In Canada and Germany, about 80% of the respondents were women; in Cameroon and Japan, the proportions of men and women were relatively equal.

Table 1. Sociodemographic characteristics, resilience, and PTG scores by country.

	Cameroon (n = 111)		Canada (n = 137)		Germany (n = 122)		Japan (n = 108)		Kruskal-Wallis H test
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	H	p
Age	35.65	8.34	34.39	10.81	28.86	9.75	52.45	14.13	167.9	<.000***
Education years	15.39	8.48	16.82	6.02	14.29	3.95	15.62	3.21	31.0	.001**
Number of children	2.24	1.92	1.15	1.22	0.29	0.69	1.12	1.05	107.99	<.000***
BRS	3.44^a	0.86	3.40^a	0.96	3.05^b	0.86	3.09^b	0.91	20.50	<.001***
PTGI-SF	21.98^a	13.80	15.33^b	13.24	13.29^b	10.59	10.23^b	11.03	31.41	<.001***
	n	%	n	%	n	%	n	%
Gender
Female (1)	47	42.73	123	89.8	97	79.51	45	41.67	95.0	<.000***
Male (0)	58	52.73	14	10.2	19	15.57	63	58.33
Status
Single (1)	29	26.13	25	18.25	46	37.70	26	24.07	46.1	.001**
In a relationship (2)	17	15.32	46	33.58	39	31.97	4	3.70
Cohabiting (3)	4	3.60	22	16.06	20	16.39	0	0
Married (4)	57	51.35	34	24.82	12	9.84	68	62.96
Separated/Divorced (5)	4	3.60	10	7.30	2	1.64	8	7.41
Widow/widower (6)	0	0	0	0	1	0.82	2	1.85
Current occupation
Full-time (1)	64	57.66	62	45.26	21	17.21	56	51.85	45.6	.001**
Part-time (2)	13	11.71	25	18.25	14	11.48	14	12.96
Unemployed (3)	13	11.71	6	4.38	5	4.10	8	7.41
Pupil/student/training/unpaid work (4)	13	11.71	30	21.90	77	63.11	3	2.78
Retirement (5)	1	0.90	1	0.73	1	0.82	10	9.26
Parental Leave (6)	2	1.80	7	5.11	2	1.64	1	0.93
Domestic Tasks (7)	2	1.80	5	3.65	1	0.82	16	14.81
Other (8)	1	0.90	1	0.73	0	0	0	0

Note. DV: domestic violence, CM: child maltreatment, SD: standard deviation, PTGI-SF: Post Traumatic Growth Inventory-Short Form, BRS: Brief Resilience Scale, Scores with the same subscript are not significantly different from each other but are significantly different from those with different subscripts (p < .05).

**p < .01, ***p < .001.

Participants from all countries reported experiencing CM. As presented in Table 2, the less frequent forms of CM included neglect, sexual abuse, and exposure to domestic violence (overall mean < 1) and the most frequent form was physical and emotional abuse (overall mean between 1 and 2). Average acceptability scores for the different CM subtypes were between 1 and 2 (between never and almost never acceptable).

Table 2. Spearman’s correlations between CM variables and resilience and PTG.

	M (SD)	Correlations
		Resilience	PTG
Acceptability: Neglect	1.50 (.64)	−.118*	.130**
Acceptability: Physical abuse	1.39 (.64)	−.039	.177***
Acceptability: Emotional malt.	1.61 (.75)	−.100*	.116*
Acceptability: Sexual abuse	1.16 (.54)	−.088	.164**
Acceptability: Exposure to DV	1.25 (.61)	−0.085	.244***
CM: Neglect	.79 (1.18)	−.138**	.170***
CM: Physical	1.41 (1.47)	−.191***	.196***
CM: Emotional	1.68 (1.85)	−.248***	.243***
CM: Sexual	.56 (1.24)	−.061	.114*
CM: Exposure to DV	.28 (.71)	−.068	.203***

Note. DV: domestic violence, CM: child maltreatment, M: mean, SD: standard deviation.

* p < .05, **p < .01, ***p < .001.

3.1.2. Association between resilience and PTG across countries

Overall means of BRS and PTGI-SF were 3.25 (SD = 0.92) and 15.86 (SD = 12.47). An ANOVA showed that the mean BRS scores differed between the four countries (F (3, 441) = 5.40, p < .001). Post hoc comparisons showed that BRS scores in Cameroon were significantly higher than in Japan and Germany; BRS scores in Canada were higher than in Germany (see Table 1 for M and SD in each country).

The mean PTGI-SF scores differed significantly between countries (Welch’s F(3, 234.69) = 9.53, p < .001). Post hoc comparisons showed that the mean PTGI-SF score was higher in Cameroon than in the other countries. A significant gender difference emerged for the BRS scores (Welch’s F (1, 305.70) = 6.22, p < .05), but not for the PTGI-SF scores (F(1, 423) = 0.075, p = .78).

The Spearman correlation coefficient between BRS and PTGI-SF across countries was r = −0.04, p = .37. However, this correlation coefficient varied widely by country. A weak and non-significant positive correlation was found in Japan (r = .19, p = .06), whereas a medium significant and negative correlation was observed in Cameroon (r = −.30, p < .01). Other countries showed weak non-significant correlations (r = −.08, p = .37 in Canada; r = −.09, p = .35 in Germany).

While resilience was negatively correlated with acceptability and experiences of CM, PTG was positively and, overall, more often significantly correlated with CM variables (see Table 2 for details).

3.2. Main analyses: multiple linear regressions

3.2.1. Factors associated with resilience

Results of the multiple linear regressions are presented in Tables 3 and 4. Being a woman as compared to a man was associated with significantly lower resilience scores; age was positively associated with resilience scores in the final block of the regressions. Countries comparisons with Japan as the reference showed a significant effect; participants from Canada, Cameroon, and Germany had higher resilience scores than Japanese participants. Canadian and Cameroonian participants did not significantly differ in the regression with Cameroon as the reference, while German and Japanese participants had significantly lower resilience scores than Cameroonian participants. In both analyses (with Japan and Cameroon as reference), none of the CM acceptability scores were significantly associated with resilience. However, experiences of physical abuse and emotional maltreatment were significantly associated with lower resilience scores.

Table 3. Multiple linear regression analyses of factors associated with resilience.

	Japan as the reference				Cameroon as the reference
	Block 1 (R^2 = 6.8%)		Block 2 (R^2 = 16.6%)		Block 1 (R^2 = 6.8%)		Block 2 (R^2 = 16.6%)
Predictors	B (SE)	[95% CI]	B (SE)	[95% CI]	B (SE)	[95% CI]	B (SE)	[95% CI]
Gender	−0.290** (0.107)	−0.500, −0.079	−0.338** (0.106)	−0.547, −0.130	−0.290** (0.107)	−0.500, −0.079	−0.338** (0.106)	−0.547, −0.130
Age	0.010* (0.004)	0.002, 0.208	0.010** (0.004)	0.002, 0.017	0.010* (0.004)	0.002, 0.018	0.010** (0.004)	0.002, 0.017
Country: Japan	Reference	Reference	Reference	Reference	−0.496*** (0.136)	−0.785, −0.208	−0.860*** (0.164)	−1.182, −0.539
Country: Canada	0.629*** (0.144)	0.347, 0.911	0.777*** (0.150)	0.483, 1.071	0.133 (0.136)	−0.133, 0.399	−0.083 (0.158)	−0.394, 0.228
Country: Cameroon	0.496*** (0.147)	0.208, 0.785	0.860*** (0.164)	0.539, 1.182	Reference	Reference	Reference	Reference
Country: Germany	0.339* (0.158)	0.028, 0.649	0.536** (0.162)	0.217, 0.855	−0.158 (0.140)	−0.433, 0.117	−0.324* (0.158)	−0.634, −0.014
Acceptability: Neglect			−0.036 (0.085)	−0.204, 0.131			−0.083 (0.158)	−0.204, 0.131
Acceptability: Physical abuse			0.049 (0.129)	−0.206, 0.303			0.049 (0.129)	−0.206, 0.303
Acceptability: Emotional malt.			0.016 (0.092)	−0.165, 0.197			0.016 (0.092)	−0.165, 0.197
Acceptability: Sexual abuse			0.096 (0.156)	−0.209, 0.402			0.096 (0.156)	−0.209, 0.402
Acceptability: Exposure to DV			−0.110 (0.148)	−0.401, 0.181			−0.110 (0.148)	−0.401, 0.181
CM: Neglect			−0.058 (0.053)	−0.162, 0.045			−0.058 (0.053)	−0.162, 0.045
CM: Physical			−0.093* (0.042)	−0.176, −0.010			−0.093* (0.042)	−0.176, −0.010
CM: Emotional			−0.091** (0.033)	−0.156, −0.026			−0.091** (0.033)	−0.156, −0.026
CM: Sexual			0.001 (0.039)	−0.076, 0.079			0.001 (0.039)	−0.076, 0.079
CM: Exposure to DV			0.026 (0.072)	−0.115, 0.167			0.026 (0.072)	−0.115, 0.167

Note. CI: confidence interval; CM: child maltreatment; DV: domestic violence Block 1: included demographics variables and country parameter; Block 2: included demographics variables, country parameter, perceived social norms of CM scores, and CM subtypes scores.

* p < .05, **p < .01, ***p < .001.

Table 4. Multiple linear regression analyses of factors associated with post traumatic growth.

	Japan as the reference				Cameroon as the reference
	Block 1 (R^2 = 8.7%)		Block 2 (R^2 = 14.8%)		Block 1 (R^2 = 8.7%)		Block 2 (R^2 = 17.9%)
Predictors	B (SE)	[95% CI]	B (SE)	[95% CI]	B (SE)	[95% CI]	B (SE)	[95% CI]
Gender	1.601 (1.453)	−1.255, 4.458	0.772 (1.444)	−2.068, 3.611	1.601 (1.453)	−1.255, 4.458	0.772 (1.444)	−2.068, 3.611
Age	−0.083 (0.054)	−0.190, 0.024	−0.086 (0.053)	−0.191, 0.019	−0.83 (0.054)	−0.190, 0.024	−0.086 (0.053)	−0.191, 0.019
Country: Japan	Reference	Reference	Reference	Reference	−7.643*** (2.2002)	−11.579, −3.708	−4.337 (2.244)	−8.748, 0.074
Country: Canada	−0.844 (1.955)	−4.687, 2.999	−0.365 (2.053)	−4.400, 3.671	−8.487*** (1.852)	−12.129, −4.846	−4.702* (2.189)	−9.007, −0.398
Country: Cameroon	7.643*** (2.002)	3.708, 11.579	4.337 (2.244)	−0.074, 8.748	Reference	Reference	Reference	Reference
Country: Germany	−3.001 (2.146)	−7.220, 1.217	−2.822 (2.215)	−7.176, 1.533	−10.645*** (1.909)	−14.397, −6.893	−7.159** (2.172)	−11.430, −2.888
Acceptability: Neglect			1.624 (1.154)	−0.645, 3.894			1.624 (1.154)	−0.645, 3.894
Acceptability: Physical abuse			1.049 (1.752)	−2.395, 4.493			1.049 (1.752)	−2.395, 4.493
Acceptability: Emotional malt.			−0.937 (1.246)	−3.387, 1.513			−0.937 (1.246)	−3.387, 1.513
Acceptability: Sexual abuse			0.078 (2.105)	−4.060, 4.216			0.078 (2.105)	−4.060, 4.216
Acceptability: Exposure to DV			3.327 (2.005)	−0.614, 7.268			3.327 (2.005)	−0.614, 7.268
CM: Neglect			−0.875 (0.714)	−2.279, 0.529			−0.875 (0.714)	−2.279, 0.529
CM: Physical			0.018 (0.573)	−1.108, 1.143			0.018 (0.573)	−1.108, 1.143
CM: Emotional			1.532*** (0.450)	0.647, 2.416			1.532*** (0.450)	0.647, 2.416
CM: Sexual			0.960 (0.533)	−0.087, 2.007			0.960 (0.533)	−0.087, 2.007
CM: Exposure to DV			0.492 (0.972)	−1.420, 2.404			0.492 (0.972)	−1.420, 2.404

Note. CI: confidence interval; CM: child maltreatment; DV: domestic violence Block 1: included demographics variables and country parameter; Block 2: included demographics variables, country parameter, perceived social norms of CM scores, and CM subtypes scores

* p < .05, **p < .01, ***p < .001.

3.2.2. Factors associated with PTG

For PTG, gender and age were not significant variables in both final steps with Japan and Cameroon as reference. Countries comparison with Japan as the reference showed that participants from Cameroon had significantly higher PTG scores in Block 1, but this difference was no longer significant in Block 2 when accounting for CM experiences and acceptability scores. When including Cameroon as the reference, participants from Canada, Japan, and Germany had significantly lower PTG scores in Block 1, but only the difference with Canada and Germany remained significant in Block 2. With regards to acceptability of CM, no significant effect was found in both regression models. However, experiences of emotional maltreatment were associated with higher PTG scores in both final blocks. Results for the other CM subtypes were non-significant.

4. Discussion

The aim of this study was to examine factors associated with positive post-traumatic outcomes in a diverse sample from four countries in four continents with different cultures, living standards, and gross national incomes. Our results partly confirmed our hypotheses by showing cross-country differences and associations between some CM variables and positive outcomes. However, acceptability of CM did not appear to exert a significant influence on either resilience or PTG.

4.1. Factors associated with resilience and PTG

4.1.1. Sociodemographic factors, resilience and PTG

Gender was significantly associated with scores of resilience, but not with PTG. Consistent with some other studies (Bonanno et al., 2007; Campbell-Sills et al., 2009), resilience was higher for men than women even after adjusting for country and experiences and acceptability of CM. This gender difference may be related to the biological vulnerability implicated in the increased mood and stress-related disorders in women (Boardman et al., 2008). However, it is possible that gender differences are related to the type of traumatic experience. A study assessing veterans showed that men's resilience scores were higher than women's, yet this association was no longer significant after accounting for trauma types (e.g. more sexual abuse and interpersonal violence in women, more accidents in men) (Portnoy et al., 2018). Another explanation may be that men perceive themselves as more capable than women to handle stressful situations (Verma et al., 2011) which could impact their answers in self-report measures such as the one we used.

In addition, age was positively associated with scores of resilience but not PTG in our multivariate model. Since we explored CM, it is possible that the greater time span between the traumatic events and the study participation in older adults resulted in these increased resilience scores (Colombo et al., 2020). So far, findings on the relationship between age and positive outcomes are inconsistent (Kunzler et al., 2018; Linnemann et al., 2020; Rodríguez-Rey et al., 2016). Regarding PTG, the nature and the severity of the traumatic experience may be worth consideration. Previous researchers that found associations with age have examined trauma types such as severe distress in a war zone (Kılıç et al., 2016), COVID-19 hardships (Cohen-Louck, 2022), serious illness, natural disaster or terrorism (see the meta-analysis: Vishnevsky et al. [2010]). Our findings showing associations between specific CM subtypes and positive outcomes provide further evidence of the importance of considering trauma types in the study of positive outcomes. However, more studies are required to elucidate the relationship between age, resilience, and PTG in the context of CM and identify potential moderators.

4.1.2. Association between resilience and PTG: cultural differences

A positive correlation between resilience and PTG was found in Japan whereas a negative correlation was observed in Cameroon, and no correlation in Canada and in Germany. This result is consistent with the literature on the relationship between resilience and PTG which appears to be ambiguous, with some studies reporting no significant association between these positive outcomes (Ogińska-Bulik & Kobylarczyk, 2016), while others found a negative (Duan et al., 2015) or a positive association (Bensimon, 2012). A possible explanation may be that there are cultural differences in how undergoing changes in values, beliefs, and perceptions of the self and others following adversity is considered positive and an indicator of personal strengths and coping abilities or, on the contrary, perceived as a sign of weakness and maladaptation following adversity. Since this is the first study to explore these questions, this interpretation should be considered with great caution and replication studies are needed with larger and representative samples from various cultures. Qualitative studies may also help uncover the meaning of these correlational patterns.

The most notable country-specific difference in this study was a higher level of PTG and resilience in Cameroon compared to other countries despite high levels of CM (see Wadji et al. 2023 for cross-country comparison in CM among participants of this study). Given the association between PTG and traumatic experiences, it is likely that the high prevalence of CM experiences in Cameroon is related to the higher PTG scores in Cameroon. However, this does not fully explain the higher PTG scores in Cameroon than in other countries, because there was a significant difference even in the model which adjusted for perceived acceptability of CM and experiences of CM. Another explanation may be that participants from low and middle income countries, like Cameroon, are used to stressful situations since they face harsh conditions and struggle for survival on a daily basis and as such they may be more resilient (Vázquez et al., 2014). PTG and resilience are important constructs that may contribute to explaining how people living in cultural settings characterised by extremely precarious conditions can continue to maintain good psychological functioning and ability to adjust to trauma.

The resilience scores by country were significantly lower in Japan, and this result did not change when adjusted for perceived acceptability of CM and experiences of CM. This result may be due to a lower self-serving bias (Hymes & Akiyama, 1991). It has also been reported that Japanese individuals do not strongly affirm positive words, and they tend to suppress positive emotions more than American individuals (Iwata et al., 1995; Iwata & Buka, 2002). A study showed that individuals of European American and Asian American in the US value high-arousal positive affect (e.g. excitement, enthusiasm) more than do Hong Kong Chinese. On the other hand, participants of Hong Kong Chinese and Asian American value low-arousal positive affect (e.g. calm, quiet) more than do European American participants (Tsai et al., 2006). Lower resilience scores among the Japanese participants might thus reflect cultural differences in ideal affect. There may thus have been a paucity of response to positive statements by participants from Japan in our study.

4.1.3. Child maltreatment, resilience and PTG

While significant associations were found between scores of CM acceptability and positive outcomes in bivariate correlations, none of these associations remained significant in our multivariate models considering sociodemographic factors, country of residence, and CM experiences. These findings indicate that actual experiences of CM may be more important than subjective perceptions regarding the acceptability of CM within one’s community in understanding resilience and PTG. Another explanation for the absence of effects with acceptability is the lower variability present in these scores, limiting our statistical power to detect significant differences. In previous studies, some researchers argued that subjective impressions of trauma and life events stressors carry more weight as compared to the occurrence of the trauma per se (Maschi et al., 2011). Future studies are needed to replicate our findings and explore potential explanations.

Interestingly, emotional maltreatment and physical abuse remained the only significant CM experiences associated with resilience when all CM subtypes were considered. Furthermore, the associations of CM experiences with lower resilience but higher PTG deserve further examination. Emotional maltreatment is highly understudied and has long been considered less damaging, but recent studies tend to invalidate this assumption (Kumari, 2020; Norman et al., 2012). Making sense of these deleterious experiences of emotional maltreatment may be a vehicle for growth. Higher PTG scores mean that many participants expressed having undergone transformative changes when recovering from the devastation brought on by their traumatic experience (Jankovic et al., 2022; Tedeschi & Calhoun, 2004). Additionally, emotional maltreatment, and physical abuse to a lower extent (Cheng & Langevin, 2022; Kim et al., 2023), have been identified as the most damaging subtypes of CM for emotion regulation, and upon close examination, the BRS items could be construed as indicators of emotion regulation abilities in high stress situations (e.g. ‘It does not take me long to recover from a stressful event’). This could explain why emotional maltreatment was associated with higher PTG but lower resilience in our sample, and why physical abuse was negatively related to resilience in our final models. Future studies should examine mechanisms underlying these associations and further explore why other CM subtypes did not yield significant findings.

5. Limitations

This study has several limitations. First, we used a cross-sectional design and causal relationships cannot be assumed. Second, we used a convenience sample and there were baseline differences in some sociodemographic parameters (e.g. age, gender). This makes it difficult to generalise our results. Third, we created and used a non-validated measure for the perceived acceptability of CM since none were already available. In the future, it would be necessary to conduct validation studies for this measure. Fourth, the internal consistency of the neglect measure was low, which may have an impact on the results. However, it was similar to that of validation studies, which also had Cronbach’s alpha ranging from moderate to high (between .61 and .82) (Dunne et al., 2009; Fung, 2020; Soer et al., 2019). Fifth, some cultures (e.g. Japan) may be less willing to disclose CM, which may influence the results of this study. Lastly, this study was conducted after the start of the COVID-19 pandemic; COVID-19 infection status varied by country and time periods, which may have affected resilience and PTG scores.

6. Practical implications

The present findings have implications for intervention efforts with CM survivors. First, our results showed that positive changes following CM, such as resilience and PTG, can vary significantly across cultures. Cultural sensitivity is therefore vital in CM intervention and prevention programmes, as there may be specific cultural aspects (e.g. a common understanding of trauma; a common pattern of adaptation to difficult life experiences) that can have a buffering effect on stress and therefore help to promote positive outcomes. We also found some correlates of resilience (e.g. age and gender) that remained significant across all cultures. This result sheds new light on the literature on positive outcomes, as it supports the idea that demographic variables, such as gender and age, are somehow linked to resilience and PTG in both non-Western and Western cultures. Second, experiences of emotional maltreatment and physical abuse appear to be the only CM subtypes related to positive post-traumatic outcomes when all CM subtypes and culture are considered. These findings, if replicated and enhanced by identifying underlying mechanisms, must be taken into consideration when reviewing CM intervention programmes (e.g. targeting these underlying mechanisms to foster positive outcomes across CM subtypes and cultures). Finally, the present investigation supports the presence of sociocultural determinants of resilience and growth which clinicians can be mindful of in order to enhance well-being and prevent psychological suffering in people exposed to CM.

7. Conclusion

Our results provide valuable insight into similarities and differences across cultures in the experience of CM as well as positive outcomes. Our results showed that self-reported male gender and age was associated with greater resilience while more experiences of physical abuse and emotional maltreatment were associated with reduced scores. On the other hand, more experiences of emotional maltreatment were associated with higher PTG. A seemingly paradoxical finding was the higher level of PTG and resilience in Cameroon as compared to other countries despite high levels of CM (Wadji et al. 2023). Our findings encourage further research on potential moderators and mediators of the relationship between culture, CM experiences, the perceived acceptability of CM, resilience, and PTG.

Acknowledgements

We wish to thank the participants without whom this project would not have been possible.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author (Dr. Rachel Langevin). The data is not publicly available due to privacy or ethical restrictions.

Additional information

Funding

This work was supported by Japan Society for the Promotion of Science KAKENHI [grant number JP23K02998] and a grant from McGill University. Dr. Langevin is supported by a Chercheur-Boursier Junior 1 Award from the Fonds de recherche du Québec – Santé (#310809). Dr. Wadji is supported by a Postdoc . Mobility grant by the Swiss National Science Foundation [grant number P500PS_214332, 2023]. M. Sc. Bartoli is supported by a doctoral scholarship funded by the German Studienstiftung des deutschen Volkes (2019–2022).