Family Function and Post-Traumatic Stress Disorder in Children and Adolescents: A Meta-Analysis

Abstract

Family function reflects the operating status of the family system, which plays a vital role in children’s mental health. The current meta-analysis examined the association between family function and post-traumatic stress disorder (PTSD) in children and adolescents for the first time. Studies published from 1980 to 2021 were identified via searching and screening. We identified 31 studies (91 unique effects) with 8,684 children. A three-level meta-analysis revealed that overall family function was negatively associated with PTSD (r = −0.205). Among elements of family function, family affect (r = −0.251), communication (r = −0.221), and cohesion (r = −0.184) were associated with less PTSD, whereas family conflict (r = 0.228) was associated with more PTSD in children. Family flexibility (r = −0.103) was not associated with PTSD. Moderator analyses revealed differences between various types of trauma events and family function scales. The findings highlight the differences in the roles of the elements of family function and suggest that interventions should be focused on targeting specific elements of family function.

Keywords

meta-analysis post-traumatic stress disorder family function children and adolescents

Introduction

Traumatic events can trigger a range of negative psychological reactions (Baryshnikova & Pham, 2019), among which post-traumatic stress disorder (PTSD) is relatively common, particularly in children and adolescents. PTSD can negatively affect their physical health, mental health, and academic and social development (Alisic et al., 2011). To help children and adolescents recover from traumatic experiences and undergo social adjustment, developing effective psychological interventions for treating PTSD is critical. To date, numerous intervention strategies have been developed to relieve the severity of PTSD symptoms. These include cognitive-behavioral therapy (e.g., Kowalik et al., 2011), eye movement desensitization and reprocessing (e.g., Rodenburg et al., 2009), narrative exposure therapy (e.g., Mørkved et al., 2014; Robjant & Fazel, 2010), and prolonged exposure therapy (e.g., Foa et al., 2013; Mørkved et al., 2014). Although these approaches have been found to be effective, a previous study reported that more than 60% of adolescents continue to report PTSD following long-term intervention (Jaycox et al., 2010). This can be attributed to two possible reasons. First, most interventions were designed to treat children and adolescents individually and emphasize personal motivation and cognition, but overlooked their family members. Second, limited research that considered and intervened PTSD in children and family members (e.g., Berkowitz et al., 2011; Cohen & Mannarino, 2008) failed to recognize the importance of the dynamics and situations of families in children’s mental health. In fact, poor family dynamics may aggravate children’s PTSD (e.g., Deane et al., 2018; Olson, 2000a), and treatments involving the family members who contributed to the child’s PTSD may not be beneficial; rather, it may negatively affect children with initial trauma exposure (e.g., Lambert et al., 2014). Therefore, interventions should consider not only the family members of the child but also improving family function.

Family function refers to the operating status or process of the family system and is one of the most important family factors for providing the necessary conditions for the healthy development and maintenance of family members in terms of social, psychological, and biological characteristics (Guerrero-Munoz et al., 2021; Minuchin, 2018; Scully et al., 2020; Wedding & Corsini, 2019). However, recent studies have only examined the relation between family factors, such as family socioeconomic status and family structure, and PTSD (e.g., Gallo et al., 2019; Koirala et al., 2020). Research has also assessed the relation between factors outside of the family system (e.g., social class, custom, mass media, and social welfare) and post-traumatic reactions (Bronfenbrenner, 1992; Shelton, 2018). Emphasis on these factors did not significantly relieve PTSD in some traumatized individuals (e.g., Mora et al., 2022; Peverill et al., 2021). This urges us to consider family function as a proximal and underlying family factor that affects children’s and adolescents’ mental health (Minuchin, 2018; Pereda & Díaz-Faes, 2020; Wedding & Corsini, 2019).

Relevant empirical studies have found that good family functions, such as family communication and cohesion, have positive effects on relieving children's PTSD when they experience a crisis or traumatic event (Acuña & Kataoka, 2017; Mora et al., 2022), whereas dysfunctional family dynamics naturally are a risk factor of PTSD in children (e.g., Deane et al., 2018; Olson, 2000a). Therefore, clarifying the role of family function in PTSD in children is an important prerequisite for developing the most effective family interventions.

Although several recent meta-analyses have attempted to explore the relations between family factors and PTSD to address the above-mentioned issue (Birkley et al., 2016; Pinquart, 2020; Trickey et al., 2012), the studies have various limitations, such as (1) the inclusion of mainly male participants (Birkley et al., 2016); (2) small sample sizes (Trickey et al., 2012); (3) PTSD in children being ignored (Birkley et al., 2016); and (4) the inclusion of only a single type of traumatic event (Pinquart, 2020). Systematic and comprehensive evaluations of this issue are scarce; thus, a comprehensive understanding of the relation between family function and PTSD is limited. We sought to fill this gap by conducting a meta-analysis synthesizing existing theories and empirical articles to elucidate the relation between family function and PTSD in children. This would not only extend the research field of family and trauma theory, but also provide support for the development of psychological interventions and family education for children and adolescents to prevent the transmission of the negative impacts of trauma within the family, relieve children’s mental health problems, and help children achieve post-traumatic growth.

Theoretical Perspectives on the Elements of Family Function

Family function is an important variable in family research. We have summarized the main theories of family function (see Supplemental Appendix A), which can be divided into outcome orientation and process orientation, which provide different perspectives on crisis situations. The Circumplex Model (Olson, 2000a, 2000b) and the Beavers Systems Model (Beavers & Hampson, 2000) represent outcome-oriented theories and emphasize the results of family operations. According to these theories, the family system changes in response to crises, and a balanced family possesses the resources and skills to change the system appropriately to deal with crises more effectively (Olson, 2000a). In contrast to outcome-oriented theories, process-oriented theories emphasize the process of family operations, which is represented by the McMaster Model (Epstein et al., 1978) and the Process Model (Skinner et al., 2000). It has been suggested that therapists need to pay more attention to the process because the failure of the family to achieve the basic functions during the operation process can lead to the development of various mental disorders (e.g., PTSD) in family members. Both models propose that a primary goal of a family system is to complete various crisis tasks (e.g., diseases and accidents) and that the processes involved in dealing with these tasks are important. The process of completing these tasks enables the continuous development and cohesion of all family members to maintain the effective operation of their families (Skinner et al., 2000). It should be noted that each theoretical orientation focuses on only one aspect of the family, and neither fully evaluates the complete picture of family function. Furthermore, the applicability of these theories to traumatic and crisis events remains to be tested.

Nevertheless, we found that various theories of the two orientations exhibit different compositions and conceptualizations of the elements of family function, as shown in Supplemental Appendix A. These elements can be differentiated into cognitive components and affective responsiveness, which include positive components (e.g., family cohesion) and negative components (e.g., family conflict). In the current study, we conceptualized the elements proposed by various theories (Beavers & Hampson, 2000; Epstein et al., 1978; Olson, 2000b; Skinner et al., 2000) and reviewed in previous studies (e.g., Holtom-Viesel & Allan, 2014) and evaluated the status of reports included in the current analyses (e.g., sufficient effect size for analysis). In addition, we combined the relevant sub-dimensions into broader elements and coded family function at the effect size level using six elements: family conflict, family cohesion, family flexibility, family communication, family norms, and family affect. The conceptual basis of each element is shown in Supplemental Appendix B.

Family Function and PTSD in Children and Adolescents

Previous studies have established a longstanding and robust association between family function and PTSD in children and adolescents (Lee et al., 2018; Trickey et al., 2012). After suffering traumatic events, family function becomes disrupted (e.g., Walsh, 2007), and poor family function significantly positively predicts PTSD with medium to large effect sizes (Trickey et al., 2012). A family that functions poorly may show disunity in tasks, have difficulty adapting to stress factors, and have no clear family rules or boundaries (Beavers & Hampson, 2000; Skinner et al., 2000), which results in family members being prone to developing adjustment problems (Vliem, 2009), such as PTSD. However, some previous studies reported that family function was not significantly correlated with PTSD (e.g., Nelson et al., 2019), which suggests that the relations are not stable and can be easily influenced by various factors. We suggest that this inconsistency in findings may be attributed to the following two factors.

First, the relations between the key elements of family function and PTSD in children and adolescents vary in terms of direction and strength. Different elements reflect various aspects of family function and play distinct roles in coping with problems (Epstein et al., 1978). For example, positive elements of family function, such as family cohesion, enable children to feel a sense of support and connection with family members during a traumatic event and facilitate positive responses to their family members’ emotions and behaviors, thereby reducing mental problems (Gorman-Smith et al., 2004) and protecting children from the negative impacts of PTSD (Deane et al., 2018). In contrast to the effects of cohesion, greater family conflict is associated with more severe trauma reactions in sexually abused children (Bal et al., 2004), tornado survivors (Danielson et al., 2017), and survivors of multiple traumatic events (Daniunaite et al., 2021). Regarding other elements of family function, Ozono et al. (2007) conducted a univariate analysis and detected significant correlations among the elements of family function (e.g., family norms and family affect) and PTSD in children with cancer; however, these correlations disappeared in the multivariate model. Taken together, these previous findings indicate that the effects of different family function elements on PTSD are varied and unstable and are affected by numerous other factors that require further exploration.

Second, the type of traumatic event, gender, assessment sources or measurement, publication status, and other factors may have moderating effects on the relation between family function and PTSD in children and adolescents. For example, numerous previous studies have reported significant relations between family function and PTSD following different types of traumatic events, including disease (Nelson et al., 2019), abuse (Bal et al., 2004), and accidents (Lee et al., 2018). Differences in parent–child perceptions of family function (Ohannessian et al., 2000), PTSD, and family function measurement scales (Gallo et al., 2019) may also contribute to the inconsistent findings on the relations between family function and PTSD. Furthermore, the relation between family function and PTSD may vary depending on publication status, where significant associations are accepted for publication (e.g., Lee et al., 2018), whereas non-significant findings remain unpublished (e.g., Hildenbrand, 2016). However, previous quantitative and cross-sectional studies have not comprehensively explored changes in the relation between family function and PTSD in children and adolescents because of differences in moderators.

Current Study

To address the gaps described above and integrate the findings of previous studies, we conducted a systematic review and meta-analysis to assess the associations between family functions, including overall family function, general family function, and its key elements (i.e., family conflict, family cohesion, family flexibility, family communication, family norms, and family affect) and PTSD, and to elucidate the moderating effect of various factors (e.g., type of traumatic events and gender) on these relations. We sought to address the following questions: (a) how does family function affect PTSD? (b) which elements of family function are most strongly related to PTSD? (c) what are the relations between the family function of outcome/process orientations and PTSD? and (d) what are the boundaries or moderators of the relation between family function and PTSD? Our study aimed to clarify the key elements of family function that have a significant impact on PTSD, integrate previous empirical studies and family function theories, verify the applicability of these theories to traumatic and crisis events, and provide targeted and operable theoretical guidance for interventions for children and adolescents in clinical practice.

Methods

This study followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines (Moher et al., 2009; see Supplemental Appendix C).

Search Strategy

Two strategies were used to search articles. First, a systematic search of the literature was conducted in six databases: PsycINFO, Web of Science, PubMed, Medline, Embase, and PsycNet (see Supplemental Appendix D). For this process, both published literature and unpublished manuscripts (e.g., Master’s and Ph.D. theses) were included. The search terms included PTSD in children and adolescents and family function: (a) PTSD in children and adolescents: (child* OR boys OR girls OR juvenil* OR minors OR adolesc* OR preadolesc* OR pre-adolesc* OR pre-school OR preschool OR paediatric* OR pediatric* OR pubescen* OR puberty OR school* OR campus OR teen* OR young OR youth*) AND (post-trauma* OR posttrauma* OR PTSD OR PTSS OR psychotrauma*); (b) Family function: (family func* OR family conflict OR family cohesion OR family communication OR family flexibility OR family problem-solving OR family adjust* OR family maint*). We then searched for meta-analyses of family factors associated with PTSD in children and adolescents (Birkley et al., 2016; Pinquart, 2020; Trickey et al., 2012), and the references of the articles included in these meta-analyses were included in this study. All searched and included literature was limited to those published in English. The preparatory work took 1 week, the search took 1 day, and the complete search was completed on October 30, 2021. The searched articles were published between 1980 (when PTSD was first defined in the DSM) and October 30, 2021.

Screening and Eligibility Criteria

Figure 1 shows the process of article inclusion. First, 2,245 articles were obtained through a database search, and a further 34 articles were identified from other related meta-analyses (Birkley et al., 2016; Pinquart, 2020; Trickey et al., 2012). After removing 1,036 duplicate articles, 1,243 articles remained and underwent screening. Two authors independently assessed the titles and abstracts of 1,243 articles according to the inclusion and exclusion criteria, after which 145 relevant articles remained. These authors downloaded the full text of the included articles and conducted independent evaluations according to the inclusion and exclusion criteria. In this process, the inter-rater reliability was 0.92, which indicated good inter-rater agreement (Cooper et al., 2019). For any inconsistencies, the original article was reviewed and discussed until a consensus was reached. Finally, 31 articles were unanimously included in our meta-analysis (see Supplemental Appendix E). During the screening, we found that no correlation coefficients were reported. Thus, the corresponding authors of the studies were contacted via email; however, we only received one reply (Daniunaite et al., 2021).

Figure 1.

Flow diagram of identification and selection of included studies.

To be considered for inclusion, studies had to include the following: a direct measure of family function; a standardized measure of PTSD in children and adolescents (e.g., self-report); statistical testing of the association between family function and PTSD; and a sample of children and adolescents with a mean age below 18 years. Studies meeting the following criteria were excluded: review, meta-analysis, commentary, case study, interview, and studies with n < 10; studies where the sample was not exposed to a traumatic event that met DSM-5 diagnostic criteria for PTSD (American Psychiatric Association, 2013); studies not written in English.

Data Extraction

During the data extraction/coding process, we adopted a more dynamic coding approach to achieve inter-rater agreement. Specifically, we conducted three rounds of discussion to address inconsistent coding during the coding process. For example, after a few days of independent coding, two coders discussed their coding results. Inconsistencies and disagreements were discussed and resolved at each round of discussion. The following study design characteristics were extracted: (a) author information; (b) year of publication; (c) publication type; (d) country; and (e) study design (i.e., cross-sectional or longitudinal study). Child characteristics were extracted: (a) mean (M) and standard deviation (SD) of the sample age; (b) percentage of female participants; and (c) sample type (e.g., clinical or comorbid). The traumatic experiences of the family members were also extracted: (a) trauma exposure; (b) traumatic events; (c) the individual who experienced the trauma (i.e., child, parent, or both); (d) duration of the traumatic events; and (e) type of traumatic event (e.g., group or individual, or intentional or unintentional). The following PTSD and family functions were extracted: (a) assessment of variables (e.g., child, parent, or diagnostic interview); (b) scale of the variable (see Supplemental Appendix F); (c) correlation coefficient (r) between PTSD and family function; (d) correlation coefficients (rs) among PTSD and the key elements of family function; and (e) the orientation of family functions (according to the concept of family function theories and scales). The main extracted information for variable coding is shown in Table 1.

Table 1.

Characteristics of Included Studies.

Author(s) (year)	Country	Race/Ethnic Identity	Family Structure	Family Income	Design Type	Sample Characteristics			Traumatic Events	PTSD in Children and Adolescents		Family Function			Number of Correlations
Author(s) (year)	Country	Race/Ethnic Identity	Family Structure	Family Income	Design Type	N	Age M (SD)	% Females	Traumatic Events	Reporter	Measures	Reporter	Measures	Theory	Number of Correlations
Acuña and Kataoka (2017)	US	46.9% Latino, 18.4% Armenian, 18.4% White/Caucasian, 16.3% Other	NA	NA	C	98	13.05 (1.17)	44.90	Mixed	C	CPSS	C	Other	Outcome	2
Al-Krenawi and Graham (2012)	Palestine	NA	93.6% Married parents	Family economic status: 8.8% low; 42.8% average; 48.4% high	C	971	NA	57.50	Political violence	C	Other	C	FAD	Process	2
Al-Krenawi et al. (2009)	Israel/Palestine	NA	NA	NA	C	892	NA	NA	Political violence	C	Other	C	FAD	Process	2
Bal et al. (2004)	US	NA	46.1% Biological parents, 50.9% divorced, etc.	NA	C	100	14.34 (1.82)	87.00	Sexual abuse	C	TSCC	C	FES	Outcome	7
Berkman (2005) ^a	US	87% Caucasian, 13% Biracial/African American/Asian-American/Hispanic	NA	Average household income: $ 80–90,000	C	39	NA	NA	Injury	B	CSDC/CPSS	NA	FAD	Process	14
Berson (1997)^a	US	34.4% Caucasian, 50% African American, 3.1% Hispanic, 12.5% Biracial	NA	NA	C	32	11.47 (1.83)	100	Sexual abuse	C	TSCC	P	FACES	Outcome	2
Bokszczanin (2008)	Poland	NA	88% Two-parent	NA	C	533	15.96 (2.50)	60.04	Flood	C	Other	C	FES	Outcome	1
Brown et al. (2003)	US	75% Caucasian, 21.2% African American, 1.9% Hispanic, 1.9% Korean	69.2% Two-parent	Hollingshead two-factor index of social position M = 36.29, SD = 11.86	C	52	17.00 (3.44)	55.80	Childhood cancer	C	PTSD-RI	B	FES	Outcome	2
Burton et al. (1994)	US	40% Black, 40% Hispanic, 10% Caucasian, 7% Asian, 3% Other	33% Two-parent, 51.6% Divorce/Separation, 15.4% Widowed/Single	NA	C	91	16.00 (1.00)	0	Serious and repeated criminal behavior	C	Other	C	FES	Outcome	3
Coakley et al. (2010)	US	88% White	NA	54% Annual income above $60,000	C	51	NA	31.00	Mixed	B	CSDC/CPSS	P	FAD	Process	2
Cousino (2015)^a	US	73.8% White/Caucasian, 10.8% Black/African American, 6.2% Hispanic/Latino, 1.5% Asian/Asian American, 7.7% Bi/Multi-Racial	18.5% Single-parent, 64.6% Two-parent, 12.3% Mixed, 4.6% Other	Family income: 24.2% < $25,000; 24.2% $25,000–$50,000; 22.6% $50,000–75,000; 9.7% $75,000–100,000; 19.4% > $100,000	C	62	NA	NA	Cancer or brain tumor	C	CPSS	C	FAD	Process	4
Danielson et al. (2017)	US	71% White/Caucasian, 26% Black/African American, 4% Other	NA	NA	C	2000	14.50 (1.70)	50.90	Tornados	C	Other	C	Other	Outcome	1
Daniunaite et al. (2021)	Lithuania	NA	73.2% Two-parent	NA	C	97	14.34 (NA)	69.10	Mixed	C	Other	C	Other	Outcome	1
Deane et al. (2018)	US	African American	48% Single-parent	Median income = $ 19,132	L	254	12.57 (NA)	59.00	Community violence	C	Other	C	Other/FES	NA/outcome	2
D’Urso et al. (2018)	UK	NA	NA	NA	C	60	12.38 (2.85)	55.00	Cancer	C	Other	P	FAD	Process	1
Gallo et al. (2019)	Belgium	NA	75.5% Nuclear, 14.6% Blended, 9.8% Single-parent	NA	L	41	10.73 (2.11)	60.98	Accident or high levels of somatic pain	C	CPTS-RI	C	FFQ	Process	3
Green (2000)^a	US	43.8% White/Caucasian, 54.2% Black/African American, 1.0% Hispanic/Latino, 1.0% Asian/Pacific Islander	NA	9.4% < $1000/month; 81.3% $1000–<$5000/month; 7.3% $5000–<$10000/month; 2.1% missing data	C	96	12.99 (1.86)	NA	Sexual abuse	C	TSCC	P	Other	Process	1
Hall et al. (2006)	US	66% Caucasian, 11% African American, 5% Hispanic, 5% Asian/Native American/West Indian, 13% Unknown	NA	NA	L	62	11.45 (NA)	31.00	Acute burn	D	Other	P	Other	Outcome	2
Hildenbrand (2016) ^a	US	82% White, 8% Black/African American, 3% Asian, 9% Hispanic, 7% Other	NA	NA	C	76	14.08 (2.14)	79.00	Pain, disease	C	CPSS	B	FAD	Process	2
Kiliç et al. (2003)	Turkey	NA	NA	Low socioeconomic status	C	49	10.25 (NA)	46.90	Earthquake	C	CPTS-RI	P	FAD	Process	7
Lee et al. (2018)	Korea	NA	NA	NA	C	57	NA	49.10	Sewol Ferry disaster	C	Other	C	FACES	Outcome	2
Marchante-Hoffman (2018)^a	US	Hispanic/Latino: 65.8% South American, 12.5% Central American, 5.3% Cuban, etc.	14% Single, 62.2% Married, 21.7% Separated/Divorced, 2.1% Widowed	Annual household income: 33.6% < $12,000; 32.9% $12,000–$23,999......	C	111	NA	NA	Mixed	C	PTSD-RI	C	FAD	Process	1
McCarthy and Thompson (2010)	US	42% White, 37.7% Black, 10.3% Hispanic, 10.1% Other	NA	NA	C	350	15.30 (1.80)	55.90	Runaway	C	TSCC	C	FFS	Process	5
Meiser-Stedman et al. (2006)	England	NA	NA	NA	L	46	NA	NA	Assaults or MVAs	C/D	Other	P	FFQ	Process	6
Mora et al. (2022)	US	Latino/a or Hispanic: 87% Mexican, 45% Puerto Rican	91% Live with two or more adults, 9% Live with one adult	NA	C	416	15.50 (1.03)	53.00	Community violence	C	Other	P	FACES	Outcome	2
Overstreet and Braun (2000)	US	African American	NA	NA	C	70	12.50(1.00)	54.00	Community violence	C	TSCC	C	FES	Outcome	1
Rowe et al. (2010)	US	62% White, non-Hispanic, 31% African American, 4% Hispanic, 3% Other	46% Single-parent and mother-headed, 39% Two-parent	Yearly median family income $35,000	C	80	15.60 (1.00)	13.00	Hurricane	B	PTSD-RI/Other	B	FES	Outcome	4
Sadeh et al. (2020)	Isreal	88.2% Jew, 11.8% Arab	4.2% Single, 80.2% Married/Partnered, 15% Separated/Divorced, 0.5% Widowed	NA	C	196	12.80 (3.10)	38.80	Traumatic medical events	B	CPSS	B	Other	Outcome	4
Thompson (2005)	US	42.1% European American, 37.7% African American, 10.3% Hispanic, 2.6% American Indian, 0.9% Asian, 6.3% Mixed	NA	NA	C	350	15.30 (1.70)	55.90	Runaway/homeless	C	TSCC	C	FFS	Process	3
Usta et al. (2013)	Lebanese	NA	90% Two-parent	NA	C	1028	11.89 (1.67)	45.92	Maltreatment	C	TSCC	C	Other	Process	1
Vasileva and Petermann (2017)	German	NA	NA	NA	C	324	4.89 (1.29)	50.30	Foster care	P	Other	P	FAD	Process	1

Note. Design type: C = cross-sectional; L = Longitudinal. Reporter: P = parent; C = child; D = diagnostic interview; NA = not available. PTSD measures in children and adolescents: CPSS = child post-traumatic symptom scale; TSCC = trauma symptom checklist for children; CSDC = child stress disorders checklist; PTSD-RI = PTSD reaction index; CPTS-RI = child post-traumatic stress reaction index. Family function measures: FAD = McMaster family assessment device; FES = family environment scale; FACES = family adaptation and cohesion evaluation scales; FFQ = family functioning questionnaire; FFS = family functioning scale.

Dissertation.

Quality Assessment

After extracting the article data, the raters independently performed a quality assessment of each article. The raters then discussed and agreed on the final results of the quality assessment. The quality assessment of studies was conducted using a modified version of the Newcastle–Ottawa Scale (Modesti et al., 2016; Wells et al., 2001), which is commonly used for the quality assessment of cross-sectional studies (Modesti et al., 2016; Velotti et al., 2021). The quality of research methods was evaluated as good, fair, or poor, based on the three aspects of the scale (i.e., selection, comparability, and outcome) and included the representativeness and size of the sample, reliability of the measures used, and appropriateness of statistical analyses (see Supplemental Appendix G).

We included 31 studies that were evaluated as having adequate quality (M = 7.71, SD = 1.16). For the selection score (M = 4.00, SD = 0.82), 10 of the 31 included studies examined representative samples with satisfactory sample size and response rate and used verified measurements, 11 studies used unvalidated measurements (although the tool was available or described), and 10 studies had limitations in terms of sample representativeness, sample size, response rate, or measurements. For the comparability score (M = 1.71, SD = 0.46), 22 of the 31 included studies controlled for all confounding factors, whereas the remaining nine controlled for the most important factors. For the outcome score (M = 2.00, SD = 0.26), 29 of the 31 included studies implemented appropriate statistical analyses, reported relevant indicators, and provided convincing interpretations, one study provided incomplete results, and one study provided rigorous descriptions of the analysis and results.

Statistical Analyses

Main analysis and heterogeneity test

Traditional univariate meta-analytic approaches require that all effect sizes be independent of each other (Assink & Wibbelink, 2016); thus, only one effect size can be extracted from each study. However, most of the included articles in this study reported multiple effect sizes, which violated the assumption that the effect sizes are independent (Lipsey & Wilson, 2001). In such cases, applying a three-level structure to a meta-analytic model is more suitable for dealing with the dependency of effect sizes than traditional methods (Assink & Wibbelink, 2016). In this method, the sources of variance are divided into three levels: the sampling variance of the effect size (level 1), the variance between effect sizes within studies (level 2), and the variance between studies (level 3; Assink & Wibbelink, 2016; Cheung, 2014). In contrast to traditional univariate meta-analyses, the three-level meta-analysis method enables the extraction of all effect sizes from articles without reducing the number of effect sizes, thereby retaining all of the information and improving statistical efficiency (Assink & Wibbelink, 2016).

Referring to the method described by Assink and Wibbelink (2016), the metafor package in R (version 4.1.2) was used to conduct the three-level meta-analyses. The sampling variance (level 1) of the meta-analysis was estimated according to the formula reported by Cheung (2014). To determine whether the within-study variance (level 2) and the between-study variance (level 3) were significant, two separate log-likelihood-ratio tests (LRTs) were performed by comparing the deviance of the full model and the deviance of models excluding the variance at either level 2 or 3 (Assink & Wibbelink, 2016). We use the restricted maximum likelihood approach to estimate all model parameters and calculate the regression coefficients and corresponding confidence intervals (CIs) (Assink & Wibbelink, 2016).

For the main analysis, we computed the weighted mean effect sizes and heterogeneity of the included studies. The correlation coefficient (r) was used as the effect size, which was the most appropriate effect size calculation for this meta-analysis. The correlation coefficients between family functions and PTSD were transformed to Fisher’s z scores (Fisher, 1925) for analysis to control for variance (Borenstein et al., 2009). These scores can be transformed back to rs when reporting effect sizes points estimates (Lipsey & Wilson, 2001). Correlation effect sizes were interpreted as: 0.10 = small; 0.30 = medium; and 0.50 = large (Cohen, 1992). However, the effect size guidelines were based on qualitative impressions, rather than a systematic quantitative analysis of the data. Gignac and Szodorai (2016) recommended that r = 0.10, r = 0.20, and r = 0.30 are, respectively, regarded as low, medium, and strong correlation, which is more suitable for interpreting meta-analysis results and enables systematic analysis of the correlation coefficients of 708 meta-analyses.

For the heterogeneity test, the relation between family function and PTSD was easily affected by other variables and the relatively high randomness. Therefore, we used random-effects models to combine the effect sizes. This method not only accounted for sampling errors, but also the within-study and between-study variance in the three-level meta-analyses; moreover, it was stricter than the fixed-effects model (Assink & Wibbelink, 2016; Brockwell & Gordon, 2001; Kontopantelis & Reeves, 2010). We constructed a forest plot and computed heterogeneity statistics (i.e., the Q test and I²) to examine the amount of variance across studies for all analyses. Although forest plots allow research results to be presented intuitively, they are not sufficiently objective. Thus, we determined the significance of the effects using the Q test and I². Cochran’s Q statistic is widely used for assessing heterogeneity caused by sampling error, but the test efficiency is low. The I² statistic is more sensitive and is suitable for cases with small effect sizes and provides information on the percentage of the variance in the observed effects that is caused by the variance in the true effects (Borenstein et al., 2009). If the Q test is significant (i.e., p < 0.05) or I² is higher than 75%, the random-effects model can be accepted (Higgins et al., 2003). One-study-removed sensitivity analysis was carried out for the main meta-analysis to assess the potential influence of each study on the combined effect size.

Moderator analyses

To further assess the effects of other variables on the relation between family function and PTSD in children and adolescents, we included moderators (e.g., gender and trauma types) in the analysis. Here, we coded trauma type based on multiple classification considerations: (1) causes of trauma: interpersonal violence (e.g., sexual abuse), humanly caused (e.g., community violence), naturally caused (e.g., hurricane), and multiple; (2) population affected: group (e.g., hurricane), individual (e.g., acute burn), and multiple; (3) intention of injury: intentional (e.g., sexual abuse), unintentional (e.g., hurricane), and multiple; (4) ways of induction: contextual (e.g., community violence), physical (e.g., acute burn), sexual, and mixed. In the three-level meta-analysis, if there was heterogeneity within and between studies, two forms of moderator analyses were conducted to identify the variables that contributed to the heterogeneity (Assink & Wibbelink, 2016): (a) univariate meta-regression analysis for a continuous moderator and (b) a subgroup analysis for a categorical moderator, which is a specific form of univariate meta-regression. To ensure that the moderators in the subgroup analysis represented each level, the number of effect sizes at each level was required to be not fewer than five (Card, 2016).

Publication bias

Publication bias tests are conducted to address the problem of significant results being easier to publish and prevent the published literature from fully representing the overall research conducted in the field (Rothstein et al., 2006). Both published journal articles and unpublished dissertations were included to minimize the impact of publication bias on the results of the meta-analysis. To assess publication bias, we used the modified Egger’s regression method to test the asymmetry of funnel plots (Egger et al., 1997), which enables the detection of selective reporting in the presence of dependent effect sizes (Rodgers & Pustejovsky, 2021). Specifically, we conducted a multilevel meta-regression of effect size estimates on standard error (SE). Duval and Tweedie’s (2000) trim-and-fill method analysis was then used to assess whether small extreme studies or missing studies biased the summary effect.

Results

Study Characteristics

In all, 31 studies were included in the current meta-analysis, with a minimum of one effect and a maximum of 14 effects in each study, providing 91 unique effects in total (see Table 1). Of these studies, 25 were published in journals (80.64%) with publication years ranging from 1994 to 2021, and six were doctoral dissertations conducted between 1997 and 2018. Data were collected in 11 countries, and 19 of the 31 studies (61.29%) were conducted in the United States. Moreover, 27 studies were cross-sectional (87.10%). Children aged 3–18 years constituted 8,684 participants. Regarding the types of traumatic events, contextual trauma (35.48%) was studied most often, followed by mixed trauma (32.26%). Most studies included participants who experienced trauma caused by humans (54.84%), those who experienced trauma of an intentional nature (45.16%), and those who experienced trauma individually (e.g., sexual abuse, acute burns; 61.29%), rather than those who experienced trauma collectively (e.g., hurricane, political violence). Participants took part in studies 3–69 months after experiencing the traumatic event. None of the 31 studies specifically focused on participants who were diagnosed with PTSD, although some studies did report the clinical symptoms of their participants.

Association Between Family Function and PTSD in Children and Adolescents

The current study used random-effects models to test the main effects of the associations between family function and PTSD using a three-level meta-analysis. Relevant statistical results are shown in Table 2. The aggregated association between overall family function and PTSD was significant and negative (r = −0.205, p < 0.001; forest plot presented in Supplemental Appendix H). Heterogeneity analysis suggested that the heterogeneity of the correlation between overall family function and PTSD was significant (Q = 227.16, p < 0.001). The three-level model fit (Akaike information criterion [AIC]_full = −57.91, Bayesian information criterion [BIC]_full = −50.45) was better than the fit of the within-study variance constrained to zero two-level model (AIC_reduced = −30.34, BIC_reduced = −35.32, LRT = 24.59, p < 0.0001). Although the three-level model fit did not differ significantly from the between-study variance constrained to zero two-level model fit (AIC_reduced = −58.91, BIC_reduced = −53.93, LRT = 1.00, p = 0.32), among the total sources of variance, the between-study variance (I² in level 3) was 11.81%. The comprehensive consideration indicated that the three-level model was most appropriate for this study. In addition, the sampling variance (I² in level 1) was 33.23%, and the within-study variance (I² in level 2) was 54.96%. Sensitivity analysis revealed that r fluctuated between −0.214 and −0.199 after excluding any sample, indicating high stability of the results.

Table 2.

Three-Level Meta-Analysis of the Association Between Family Function and PTSD in Children and Adolescents.

Variable	s	k	Effect Size r(95% CI)	t	p Value	σ2 Level 2	σ2 Level 3	%Var. level 1	%Var. level 2	%Var. level 3
Overall family function	31	91	–0.205 [–0.241, –0.169]	–10.967	<.001	0.011	0.002	33.229	54.959	11.813
General family function	13	22	–0.235 [–0.295, –0.174]	–7.833	<.001	0.003	0.004	44.990	20.706	34.304
Element of family function
Family conflict	10	11	–0.228 [–0.376, –0.069]	–3.167	0.010	0.000	0.046	7.966	0.000	92.034
Family cohesion	8	14	–0.184 [–0.287, –0.078]	–3.717	0.003	0.023	0.000	24.815	75.185	0.000
Family flexibility	8	14	–0.103 [–0.210, 0.007]	–2.016	0.065	0.000	0.008	74.401	0.000	25.600
Family communication	7	9	–0.221 [–0.319, –0.112]	–4.581	0.002	0.005	0.004	47.422	29.332	23.246
Family norms	5	9	–0.140 [–0.247, –0.030]	–2.924	0.019	0.000	0.005	60.364	1.731	37.905
Family affect	6	12	–0.251 [–0.311, –0.191]	–8.801	<.001	0.001	0.000	93.918	6.082	0.000
Orientation of family function
Process	16	55	–0.208 [–0.257, –0.158]	–8.134	<.001	0.003	0.005	47.655	19.709	32.637
Outcome	15	35	–0.209 [–0.271, –0.145]	–6.545	<.001	0.026	0.000	18.892	81.108	0.000
Type 1: Causes of trauma
Interpersonal violence trauma	2	5	–0.026 [–0.471, 0.428]	–0.149	0.889	0.132	0.000	10.987	89.013	0.000
Humanly caused trauma	17	54	–0.216 [–0.261, –0.170]	–9.244	<.001	0.004	0.004	41.732	28.743	29.525
Naturally caused trauma	8	26	–0.202 [–0.275, –0.126]	–5.450	<.001	0.009	0.003	38.927	44.645	16.428
Multiple trauma	4	6	–0.284 [–0.417, –0.138]	–4.915	0.004	0.007	0.001	61.191	35.245	3.564
Type 2: Population affected
Group trauma	8	25	–0.231 [–0.310, –0.149]	–5.717	<.001	0.006	0.008	28.148	30.716	41.135
Individual trauma	19	60	–0.185 [–0.225, –0.144]	–8.854	<.001	0.013	0.000	35.896	63.704	0.400
Multiple trauma	4	6	–0.284 [–0.417, –0.138]	–4.915	0.004	0.007	0.001	61.191	35.245	3.564
Type 3: Intention of injury
Intentional trauma	14	38	–0.216 [–0.264, –0.166]	–8.650	<.001	0.017	0.000	21.174	78.826	0.000
Unintentional trauma	13	47	–0.186 [–0.246, –0.125]	–6.053	<.001	0.004	0.006	54.111	19.170	26.719
Multiple trauma	4	6	–0.284 [–0.417, –0.138]	–4.915	0.004	0.007	0.001	61.191	35.245	3.564
Type 4: Ways of induction
Physical trauma	7	29	–0.185 [–0.264, –0.104]	–4.614	<.001	0.000	0.005	79.135	0.000	20.865
Sexual trauma	3	10	–0.137 [–0.282, 0.014]	–2.055	0.070	0.017	0.003	38.532	52.953	8.514
Contextual trauma	11	25	–0.204 [–0.268, –0.138]	–6.279	<.001	0.008	0.004	23.842	49.824	26.334
Mixed trauma	10	27	–0.244 [–0.310, –0.176]	–7.236	<.001	0.021	0.000	21.457	78.543	0.000

Note. The assumed direction of the relationship between family conflict and PTSD is inconsistent with the direction of the relationship between the overall family function and PTSD, so the effect size of family conflict is reversed. s = number of studies; k = number of effect sizes; CI = confidence interval; %Var = percentage of variance explained; σ2 level 2 = variance between effect sizes within the same study; σ2 level 3 = variance between studies.

PTSD = post-traumatic stress disorder.

To evaluate the correlation between family function and PTSD in different conditions (i.e., general family function, family function elements, orientations of family function, and trauma types), separate meta-analyses were performed. Results revealed that general family function was significantly negatively correlated with PTSD (r = −0.205, p < 0.001). For the elements of family function, correlations ranged from large to small: family affect, family communication, family cohesion, and family norms were all significantly negatively correlated with PTSD (rs = −0.251 to −0.140, ps ≤ 0.019). The relation between family conflict and PTSD was significantly positively correlated (r = 0.228, p = 0.010); however, the association between family flexibility and PTSD was not significant (r = −0.103, p = 0.065). For the different orientations of family function, process-oriented and outcome-oriented family functions were both significantly negatively correlated with PTSD (rs = −0.208 to −0.209, ps < 0.001). Regarding different trauma types, the correlation coefficients between family function and PTSD ranged from −0.026 (interpersonal violence trauma) to −0.284 (multiple trauma), and the significance of the correlation coefficients ranged from non-significant (interpersonal violence trauma) to significant (e.g., humanly caused trauma, and group trauma).

For the publication bias tests for the above-mentioned relations described above, we conducted a three-level meta-regression of effect sizes on SEs (weighted by the inverse of the variance of the effect sizes). Results suggested that the SE significantly moderated the relation between the two variables, indicating the presence of publication bias (SE = 0.920, t = 2.935, p = 0.004). In addition, regardless of the effect sizes from within studies, there was publication bias for the relation between overall family function and PTSD (Egger’s regression intercept = −0.298, 95% CI = [−0.366, −0.231], p = 0.003; see Supplemental Appendix I).

Because of significant publication bias, we used Duval and Tweedie’s (2000) trim-and-fill method to adjust the results. The adjusted overall effect size was relatively unaffected (z = −0.244, 95% CI = [−0.279, −0.209], p < 0.0001; see Supplemental Appendix J), which indicated that publication bias had little effect on the results and that the results of the meta-analysis were accurate and reliable.

Moderator Analyses

The heterogeneity of the relation between overall family function and PTSD was significant. Thus, we conducted subsequent analyses of the moderator effects to identify the factors that caused the variation. In all, 17 moderation models were constructed to test whether moderators significantly impacted the relation between overall family function and PTSD. Different types of moderators require different analysis methods. For categorical variables (i.e., elements of family function, orientation of family function, trauma types, trauma experiencer, PTSD reporter, family function reporter, PTSD scale, family function scale, publication status, and study design), we used subgroup analyses. For continuous variables (i.e., publication year, age, proportion of female participants, and duration of traumatic events), we used univariate meta-regression analyses. The results of the moderating effect tests are shown in Table 3. Results revealed that most variables non-significantly moderated the relation between overall family function and PTSD.

Table 3.

Three-Level Meta-Analysis of the Moderator Analysis Between Family Function and PTSD in Children and Adolescents.

Moderators	s	k	Intercept/Mean z [95% CI]	Mean r	t ₀	β₁ [95% CI]	t ₁	p ₁	F _{(df1, df2)}	p Value	%Var. Level 2	%Var. Level 3
Model 1: Element of family function
Family conflict (RC)	10	11	–0.232 [–0.395, –0.069]***	–0.228	–3.167				F_{(5, 85)} = 1.736	0.135	51.501	13.479
Family cohesion	8	14	–0.186 [–0.295, –0.078]**	–0.184	–3.717	.054 [–0.047, 0.155]	1.066	0.290
Family flexibility	7	9	–0.103 [–0.213, 0.007]	–0.103	–2.016	.131 [0.017, 0.246]	2.285	0.025
Family communication	8	14	–0.225 [–0.33, –0.112]**	–0.221	–4.581	.004 [–0.109, 0.117]	0.070	0.944
Family norms	5	9	–0.141 [–0.252, –0.030]*	–0.140	–2.924	.102 [–0.012, 0.216]	1.784	0.078
Family affect	6	12	–0.257 [–0.322, –0.193]***	–0.251	–8.801	–.013 [–0.124, 0.097]	–0.243	0.808
Model 2: Orientation of family function
Process (RC)	16	55	–0.211 [–0.263, –0.159]***	–0.208	–8.134				F_{(1, 89)} = 0.389	0.534	53.989	13.247
Outcome	15	35	–0.212 [–0.278, –0.146]***	–0.209	–6.545	–.024 [–0.101, 0.053]	–0.624	0.534	F_{(1, 89)} = 0.389	0.534	53.989	13.247
Model 3: Type 1 causes of trauma
Interpersonal violence trauma (RC)	2	5	–0.026 [–0.511, 0.458]	–0.026	–0.149				F_{(3, 87)} = 1.839	0.146	54.802	11.223
Humanly caused trauma	17	54	–0.219 [–0.267, –0.172]***	–0.216	–9.244	–.172 [–0.343, –0.000]	–1.993	0.049
Naturally caused trauma	8	26	–0.205 [–0.282, –0.127]***	–0.202	–5.450	–.160 [–0.340, 0.020]	–1.764	0.081
Multiple trauma	4	6	–0.292 [–0.444, –0.139]**	–0.284	–4.915	–.246 [–0.458, –0.034]	–2.305	0.024
Model 4: Type 2 population affected
Group trauma (RC)	8	25	–0.235 [–0.320, –0.150]***	–0.231	–5.717				F_{(2, 88)} = 1.273	0.285	55.104	11.410
Individual trauma	19	60	–0.187 [–0.229, –0.145]***	–0.185	–8.854	.043 [–0.041, 0.127]	1.011	0.315
Multiple trauma	4	6	–0.292 [–0.444, –0.139]**	–0.284	–4.915	–.058 [–0.209, 0.093]	–0.759	0.450
Model 5: Type 3 intention of injury
Intentional trauma (RC)	14	38	–0.219 [–0.270, –0.168]***	–0.216	–8.650				F_{(2, 88)} = 1.410	0.250	58.266	8.447
Unintentional trauma	13	47	–0.188 [–0.251, –0.126]***	–0.186	–6.053	.043 [–0.034, 0.120]	–1.117	0.267
Multiple trauma	4	6	–0.292 [–0.444, –0.139]**	–0.284	–4.915	–.069 [–0.212, 0.074]	–0.957	0.341
Model 6: Type 4 ways of induction
Physical trauma (RC)	7	29	–0.187 [–0.270, –0.104]***	–0.185	–4.614				F_{(3, 87)} = 1.087	0.359	53.816	13.586
Sexual trauma	3	10	–0.138 [–0.290, 0.014]	–0.137	–2.055	.039 [–0.111, 0.190]	0.520	0.604
Contextual trauma	11	25	–0.207 [–0.275, –0.139]***	–0.204	–6.279	–.025 [–0.130, 0.081]	–0.464	0.644
Mixed trauma	10	27	–0.249 [–0.320, –0.178]***	–0.244	–7.236	–.071 [–0.179, 0.036]	–1.323	0.189
Model 7: Traumatic experiencer
Children (RC)	26	77	–0.209 [–0.250, –0.168]***	–0.206	–10.220			0.993	F_{(1, 89)} = 0.000	0.993	54.188	13.327
Family members	5	14	–0.203 [–0.325, –0.081]**	–0.200	–3.602	.000 [–0.104, 0.105]	0.009	0.993	F_{(1, 89)} = 0.000	0.993	54.188	13.327
Model 8: Duration of trauma happened	10	43	–0.215 [–0.308, –0.123]***	–	–4.698	.001 [–0.004, 0.006]	0.391	0.698	F_{(1, 41)} = 0.153	0.698	0.000	56.219
Model 9: PTSD reporter
Children (RC)	29	72	–0.198 [–0.237, –0.160]***	–0.195	–10.202				F_{(2, 88)} = 1.078	0.345	59.835	6.721
Parent	5	13	–0.192 [–0.292, –0.093]**	–0.190	–4.216	–.116 [–0.276, 0.043]	–1.449	0.151
Diagnostic interview	2	6	–0.326 [–0.477, –0.175]**	–0.315	–5.553	.006 [–0.098, 0.110]	0.123	0.903
Model 10: Family function reporter
Children (RC)	20	46	–0.225 [–0.271, –0.179]***	–0.221	–9.796				F_{(2, 74)} = 0.367	0.694	63.668	7.416
Parents	12	29	–0.190 [–0.262, –0.118]***	–0.188	–5.393	.022 [–0.060, 0.105]	0.534	0.595
Children and parents	1	2	–0.126 [–1.409, 1.158]	–0.125	–1.246	.098 [–0.174, 0.369]	0.717	0.476
Model 11: PTSD measures
CPSS (RC)	6	20	–0.189 [–0.318, –0.060]**	–0.187	–3.059				F_{(5, 85)} = 0.648	0.664	57.137	10.144
PTSD-RI	3	5	–0.164 [–0.416, 0.088]	–0.163	–1.807	.038 [–0.139, 0.216]	0.430	0.669
TSCC	7	20	–0.215 [–0.277, –0.153]***	–0.212	–7.228	–.013 [–0.131, 0.105]	–0.217	0.829
CPTS-RI	2	10	–0.106 [–0.215, 0.002]	–0.106	–2.224	.091 [–0.072, 0.255]	1.111	0.270
CSDC	2	8	–0.170 [–0.373, 0.034]	–0.168	–1.973	.029 [–0.143, 0.202]	0.339	0.736
Other	14	28	–0.234 [–0.301, –0.167]***	–0.230	–7.188	–.029 [–0.138, 0.080]	–0.529	0.598
Model 12: Family function measures
FAD (RC)	10	36	–0.173 [–0.240, –0.105]***	–0.171	–5.204				F_{(5, 85)} = 2.059	0.078	66.419	0.000
FES	7	19	–0.191 [–0.302, –0.081]**	–0.189	–3.635	–.042 [–0.136, 0.051]	–0.905	0.368
FFQ	2	9	–0.283 [–0.487, –0.078]*	–0.276	–3.184	–.149 [–0.291, –0.008]	–2.094	0.039
FACES	3	6	–0.085 [–0.370, 0.200]	–0.085	–0.77	.023 [–0.120, 0.167]	0.321	0.749
FFS	2	8	–0.223 [–0.298, –0.148]***	–0.219	–7.055	–.070 [–0.178, 0.037]	–1.296	0.198
Other	8	13	–0.270 [–0.358, –0.181]***	–0.264	–6.646	–.126 [–0.224, –0.027]	–2.534	0.013
Model 13: Publication year	31	91	–0.203 [–0.239, –0.168]***	–	–11.384	–.004 [–0.009, 0.001]	–1.619	0.109	F_{(1, 89)} = 2.620	0.109	60.029	5.519
Model 14: Publication status
Publish (RC)	25	67	–0.222 [–0.261, –0.183]***	–0.218	–11.375			0.105	F_{(1, 89)} = 2.682	0.105	59.888	5.997
Unpublish	6	24	–0.129 [–0.203, –0.054]**	–0.128	–3.578	.071 [–0.015, 0.157]	1.638	0.105	F_{(1, 89)} = 2.682	0.105	59.888	5.997
Model 15: Study design
Cross-sectional (RC)	27	78	–0.200 [–0.239, –0.161]***	–0.197	–10.146			0.438	F_{(1, 89)} = 0.606	0.438	56.530	10.890
Longitudinal	4	13	–0.2 36 [–0.372, –0.100]**	–0.232	–3.771	–.044 [–0.157, 0.069]	–0.779	0.438	F_{(1, 89)} = 0.606	0.438	56.530	10.890
Model 16: Age	25	67	–0.199 [–0.239, –0.159]***	–	–9.985	–.014 [–0.031, 0.004]	–1.551	0.126	F_{(1, 65)} = 2.405	0.126	73.067	0.000
Model 17: Proportion of females	26	70	–0.197 [–0.237, –0.157]***	—	–9.810	–.000 [–0.002, 0.002]	–0.155	0.877	F_{(1, 68)} = 0.024	0.877	73.941	0.042

Note. The Fisher’s z between study-wise predictor and outcome variables was the dependent variable in moderator models. The moderators were analyzed as categorical covariates as appropriate. β₀ = intercept/mean effect size (z); t₀ = difference in mean r with zero; β₁ = estimated regression coefficient; t₁ = difference in mean r with reference category; p₁ = the significance test in mean r with reference category; F_(df1,df2) = omnibus test; RC = reference category; p = the significance test for moderating effect test.

p < .05. **p < .01. ***p < .001.

PTSD = post-traumatic stress disorder.

It should be noted that the subgroup analysis revealed that significant differences between individual categories. For example, for the elements of family function, compared with the correlation between family conflict and PTSD (r = 0.228), the correlation between family flexibility and PTSD was weaker (r = −0.103, β = 0.131, p = 0.025). Among the different trauma types, the correlation between interpersonal violence trauma and PTSD was significantly weaker (r = −0.026) than that for humanly caused trauma (r = −0.216, β = −0.172, p = 0.049) and multiple trauma (r = −0.284, β = −0.246, p = 0.024). Among the different family function scales, FAD (r = −0.171) exhibited a greater change in slope relative to FFQ (r = −0.276; β = −0.149, p = 0.039) and other family function scales (r = −0.264; β = −0.126, p = 0.013). It is worth noting that the associations between process orientation (r = −0.208) and outcome orientation (r = −0.209) of family function and PTSD were not significantly different (β = −0.024, p = 0.534), and that the diagnostic interview of PTSD had a particularly strong effect size (r = −0.315).

Discussion

To the best of our knowledge, this is the first study to carry out a meta-analysis of the relation between family function and PTSD in children and adolescents. The results indicated that the relations between overall family function and general family function separately had moderate negative correlations with PTSD. Moreover, certain elements of family function (e.g., family affect, conflict, and communication) were correlated with the severity of children’s PTSD symptoms; thus, these should be considered when developing PTSD interventions for children and adolescents. A further novel finding was that the degree and direction of associations between elements of family function and PTSD are diverse and unique. Specifically, family affect, family communication, family cohesion, and family norms, as positive elements of family function, were negatively correlated with PTSD. Family conflict, a negative element of family function, was positively associated with PTSD, whereas family flexibility was not significantly related to PTSD.

Regarding the positive elements of family function, consistent with previous studies, the current results support the perspective that positive family affect (McCarthy & Thompson, 2010), family communication (Bal et al., 2004), family cohesion (Deane et al., 2018; Rowe et al., 2010), and family norms (Sadeh et al., 2020) were negatively associated with PTSD. In this regard, when a family encounters a stressful event, negative family affect is directly induced as a short-term unstable factor, such as irritable distress or worry (Gallo et al., 2019; McCarthy & Thompson, 2010), and has a temporarily strong relation with PTSD. Family cohesion allows children to feel a sense of support and connection with family members (Gorman-Smith et al., 2004) and buffers the separation and cognitive shock caused by trauma (Kaur & Kearney, 2013). In addition, open communication among family members can help when traumatic experiences are encountered by increasing the understanding of the meaning of the traumatic events (McCarty & McMahon, 2003). However, problematic family communication has a greater impact on PTSD due to the increase in avoidance and arousal and the inhibition of the potential protective effects of open communication (Acuña & Kataoka, 2017). In addition, family norms of rigidness and closure may limit disclosure, hinder trauma processing, and promote PTSD-related distress, which can worsen the severity of PTSD (Belsher et al., 2012; Whealin et al., 2015). In contrast, a stable family structure and ample family resources as family norms are correlated with a lower prevalence of PTSD, possibly because of lower sensitivity and alertness to danger among children living in environments of higher socioeconomic status (Sadeh et al., 2020).

As a negative long-term family function element, family conflict was positively associated with PTSD in children and adolescents, which is consistent with previous studies (Bal et al., 2004; Bokszczanin, 2008; Danielson et al., 2017; McCarthy & Thompson, 2010). Families that have experienced traumatic events may experience more arguments and conflict, possibly because of difficulties coping with the stress and impact of the traumatic event. Such circumstances constitute negative social support (Gaipa, 1990), which can weaken children’s family support and hinder children’s recovery from the trauma (Bal et al., 2004; Bokszczanin, 2008; Danielson et al., 2017; McCarthy & Thompson, 2010). In addition, of the long-term and stable elements, the negative element of family function (e.g., family conflict) had a stronger and longer-lasting relation with PTSD than the positive elements of family function (e.g., family cohesion, family communication, family norms). Specifically, family conflict may be associated with different stages of PTSD development and play a vital role in the onset, resolution, continuous exacerbation, or fluctuation of symptoms over time (Layne et al., 2018). Following the occurrence of a crisis or trauma, family conflict may be considered a long-term risk factor for the development of PTSD symptoms in children.

In the current study, compared with family conflict, family flexibility had a smaller relation with PTSD, which was non-significant and the most distant among the long-term elements. This finding is consistent with most previous studies and may be attributed to the differences in functional directions at various time points following trauma (Berkman, 2005; Kiliç et al., 2003). For example, overprotection, a component of family flexibility, has been reported to harm children’s emotional state 26 months after trauma (McFarlane, 1987), although it is also considered to function as a protective and supportive factor 1 month after trauma (Gallo et al., 2019). Thus, the lack of significant effects of family flexibility may still have important effects on other aspects of family members or may function differently at various times following a traumatic event.

Regarding the different orientations of family function theories (Beavers & Hampson, 2000; Epstein et al., 1978; Olson, 2000b; Skinner et al., 2000), the current results revealed that both process- and outcome-oriented family functions were significantly negatively related to PTSD. The findings support both orientations of family function theories and extend the applicability of family function theories to traumatic events. Specifically, according to outcome-oriented theories of family function (Beavers & Hampson, 2000; Olson, 2000a, 2000b), dysfunctional families tend to be either rigid or chaotic and have difficulty in obtaining the necessary resources for adapting to a crisis. This leads to more stress and problems for the long-term development of family members, which subsequently affects children’s PTSD (e.g., Mora et al., 2022). The process-oriented theories of family function (Epstein et al., 1978; Skinner et al., 2000) suggest that if families are no longer able to effectively deal with dangerous events during the process of family operations, chronic dysfunction and various mental disorders (e.g., PTSD) may emerge in children and adolescents (e.g., Al-Krenawi & Graham, 2012). Our results demonstrated that the correlations between PTSD and the two orientations of family function were consistent; thus, both orientations effectively explained family function.

For traumatic events with humanly or natural causes, whether individual or group, or intentional or unintentional, family function has a continuous impact on PTSD; such events include diseases (Nelson et al., 2019), abuse (Bal et al., 2004), and accidents (Lee et al., 2018). However, for children who have experienced interpersonal violence trauma or sexual abuse, PTSD may not be affected by family function. This phenomenon may be caused by the personal impact of these traumatic events on children being greater and more profound than the relatively weak protective effects of family function. For type 1 traumatic events (causes of trauma), trauma caused by humans and multiple traumatic experiences were more strongly related than interpersonal violence trauma to PTSD. In addition, our results revealed that, compared with a single traumatic event, family function was more strongly correlated with PTSD after experiencing multiple traumatic events, which reflected the cumulative effects of trauma, confirming previous findings (e.g., Green et al., 2000). This may be because, compared with experiencing a single traumatic event, after experiencing multiple traumatic events, children have more PTSD symptoms (Green et al., 2000), show greater alertness and sensitivity, and become more susceptible to changes in family function.

In addition to testing the theoretical orientations and elements of family function and trauma types as moderators, we also evaluated the moderating effects of measurements of PTSD and family function and investigated the characteristics of these relations. We found that most factors played a non-significant moderating role in the relation between overall family function and PTSD, which indicated that the relation is relatively stable under most conditions.

The moderating effect of duration since a traumatic event was not significant, indicating that family function has a lasting relation with PTSD and plays an important role in all stages of the development of PTSD. Our findings suggested that the individual who experienced the traumatic event (i.e., children or family members) did not impact the relation between the two. Thus, family function is associated with PTSD, regardless of which family member experiences the traumatic event. PTSD in children may originate from themselves or under the influence of family members. For example, parents who experience excessive pressure and carry significant burdens when coping with and recovering from traumatic events may exhibit negative emotions and behaviors, which may present difficulties in providing their children with necessary help, warmth, and attention (Bokszczanin, 2008). In turn, this may lead to PTSD in children and adolescents (e.g., Lambert et al., 2014). Notably, the diagnostic interview of PTSD had a particularly strong effect size, although this was based on only two studies. Taken together, our findings suggest that family function is strongly associated with PTSD.

The individual who reported PTSD and overall family function (via self-report or reported by parents) had no impact on the relation between PTSD and family function. Parents can typically understand their children’s feelings relative accurately; thus, parents’ reports of their children’s symptoms are generally consistent with children’s perceptions. For example, when children exhibit mental health problems, what they report is the same as what their parents report (e.g., Lambert et al., 2014). The moderating effects of different scales of PTSD were not significant, indicating that there was no difference in the effects measured between the different PTSD scales. This suggested that the scales have adequate reliability and validity and can serve as suitable research measurement tools. For the family function scales, we found that the correlation between family function and PTSD was stronger using the FFQ (McFarlane, 1987) than with the FAD (Epstein et al., 1983). However, the family adaptation and cohesion evaluation scale (Olson, 2000a, 2000b), which is commonly used to measure the relation between two variables, did not yield a significant effect. This finding suggests that a standardized and unified measurement tool that integrates the core points of different family function scales is needed to enable comparisons, references, and generalizations between relevant studies.

In addition, our findings revealed that, regardless of the age and gender of children, the design of the study (i.e., cross-sectional or longitudinal), the publication status, and the year of publication, the intermediate correlation between family function and PTSD was relatively stable. This illustrates the consistency of the relation between family function and PTSD, which was also unaffected by publication bias.

Limitations and Implications

Several limitations of this study should be noted. First, because only four longitudinal studies (12.90%) were included, we were unable to make any casual inferences. Future studies should use longitudinal or experimental designs to examine the bidirectional relations between family function and PTSD in children and adolescents. Second, although our moderation analysis took into consideration clinical characteristics, trauma types, reporters, and scales, it would be valuable for future studies to explore other moderators, such as clinical samples, parents’ PTSD, and different PTSD symptom clusters. Third, mediation analysis was not performed in the current study due to the small number of articles and because the underlying mechanisms between family function and PTSD were unclear. This issue requires further exploration in the future. Fourth, several subgroups (e.g., interpersonal violence trauma, k = 5; family norms, k = 9) had relatively low effect sizes; thus, these results should be interpreted with caution.

Despite these limitations, the current study has important theoretical, clinical, and practical implications. Regarding theoretical contributions, our results demonstrated that the two different orientations of family function theories similarly explain the relieving effects of family function on PTSD. Furthermore, the current study is the first to comprehensively investigate family function as a protective factor for children with PTSD. We also revealed that family affect, communication, cohesion, and norms were positively associated with PTSD, whereas family conflict exhibited a significant negative association, and family flexibility had no association with PTSD. Our findings suggest that future research should focus on specific elements of family function that have important relations with PTSD. In addition, moderating effects were not found for most conditions (e.g., trauma experiencer or the PTSD reporter), which suggested that family function and PTSD are interrelated negatively without distinction, providing support for the collection of family and trauma data in future studies. We also clarified the boundaries of the effects of family function on PTSD. Furthermore, all of these findings promoted the convergence of the fields of trauma and family research and further expanded the application of family function theories to trauma, which will be valuable for future research and the development of a more integrated perspective for future investigations.

In clinical practice, given that family function is negatively related to PTSD, it may be valuable to develop relevant intervention protocols. However, most previous studies have used more “general” intervention approaches to relieve PTSD, such as trauma-focused cognitive-behavioral therapy (Kowalik et al., 2011). Moreover, few studies have considered family function in interventions. Thus, we suggest that future studies place greater emphasis on family function when developing intervention protocols. Specifically, our findings may encourage psychotherapists to treat children and adolescents with PTSD by promoting the positive role of a well-functioning family, such as by increasing positive affect and open communication within the family, encouraging family cohesion, helping to establish appropriate family rules, and promoting the expression of emotions. In addition, psychotherapists can prevent the exacerbation of PTSD in children by limiting the negative effects of a dysfunctional family, such as by easing family conflicts and reducing problematic communication among family members. For certain types of traumatic events, interventions and treatments for severe mental disorders should consider the cumulative effects of multiple traumatic experiences.

Critical Findings

Overall family function was negatively associated with children’s PTSD.

Family function elements correlated with PTSD in different directions.

The effects of trauma types and family function scales varied in the relation between family function and PTSD.

Intervention protocols based on family function are needed.

Implications for Practice and Research

Future studies should place more emphasis on family function when developing intervention protocols.

Future research should focus on specific elements of family function (e.g., family affect, conflict, and communication) that significantly impact PTSD and develop specific intervention programs for different elements.

More attention should be paid to the negative elements of family function to prevent the exacerbation of PTSD symptoms.

Results further expanded the application of family function theories to trauma and provided evidence to inform future research.

Supplemental Material

sj-docx-1-tva-10.1177_15248380221126182 – Supplemental material for Family Function and Post-Traumatic Stress Disorder in Children and Adolescents: A Meta-Analysis

Supplemental material, sj-docx-1-tva-10.1177_15248380221126182 for Family Function and Post-Traumatic Stress Disorder in Children and Adolescents: A Meta-Analysis by Yingying Ye, Yifan Li, Shuxian Jin, Jiali Huang, Rong Ma, Xuan Wang and Xiao Zhou in Trauma, Violence, & Abuse

Footnotes

Acknowledgements

We would like to thank the authors who provided additional information and data for the meta-analysis. We also wish to thank associate professor Tengfei Wang, who shared valuable research suggestions with us.

Contributors

Y.Y. and X.Z. conceptualized and designed this meta-analysis. Y.Y. and J.H. conducted the literature search and screening. Y.Y., Y.L., R.M., and X.W. independently coded all studies and assessed the study quality. YYY and SXJ performed the data analysis. Y.Y. and Y.L. completed the first draft of the manuscript with supervision from X.Z. All authors contributed to and approved the final manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the General Project for National Social Science Fund of China (Grant No. 20BSH167).

ORCID iD

Yingying Ye

References for included studies are in .

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Yingying Ye, PhD candidate at Zhejiang University. Her research interests are focused on trauma, family function, and attachment fields.

Yifan Li is a MEd candidate at Zhejiang University. Her research focuses on the aftermath of traumatic events as well as its mechanism.

Shuxian Jin, MEd, is a PhD candidate in Social Psychology at the Vrije Universiteit Amsterdam. Her research focuses on understanding structural features of interdependent situations and institutional solutions to promote cooperation.

Jiali Huang is a PhD candidate at Zhejiang University. Her research interests include the intergenerational effect of posttraumatic stress disorder symptoms after nature disaster.

Rong Ma is a MEd candidate at Zhejiang University. His research aims at exploring the mechanism of how intrusive rumination affects PTSD.

Xuan Wang is a MEd candidate at the Zhejiang University. Her research focuses on the underlying mechanisms of PTSD and PTG in post-traumatic individuals.

Xiao Zhou, PhD, works as research fellow at Zhejiang University. He researches the family mechanism of psychological rehabilitation after major disasters, so as to bring research into practice.

References

Acuña

M. A.

Kataoka

(2017). Family communication styles and resilience among adolescents. Social Work, 62(3), 261–269. https://doi.org/10.1093/sw/swx017

Alisic

Jongmans

M. J.

van Wesel

Kleber

R. J.

(2011). Building child trauma theory from longitudinal studies: A meta-analysis. Clinical Psychology Review, 31(5), 736–747. https://doi.org/10.1016/j.cpr.2011.03.001

Al-Krenawi

Graham

J. R.

(2012). The impact of political violence on psychosocial functioning of individuals and families: The case of palestinian adolescents: Impact of political violence: Palestinian adolescents. Child and Adolescent Mental Health, 17(1), 14–22. https://doi.org/10.1111/j.1475-3588.2011.00600.x

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Author.

Assink

Wibbelink

C. J. M.

(2016). Fitting three-level meta-analytic models in R: A step-by-step tutorial. Quantitative Methods for Psychology, 12(3), 154–174. https://doi.org/10.20982/tqmp.12.3.p154

Bal

De Bourdeaudhuij

Crombez

Van Oost

(2004). Differences in Trauma Symptoms and Family Functioning in Intra-and Extrafamilial Sexually Abused Adolescents. 19(1), 108-123. https://doi.org/10.1177/0886260503259053

Baryshnikova

N. V.

Pham

N. T. A.

(2019). Natural disasters and mental health: A quantile approach. Economics Letters, 180, 62–66. https://doi.org/10.1016/j.econlet.2019.04.016

Beavers

Hampson

R. B.

(2000). The beavers systems model of family functioning. Journal of Family Therapy, 22(2), 128–143. https://doi.org/10.1111/1467-6427.00143

Belsher

B. E.

Ruzek

J. I.

Bongar

Cordova

M. J.

(2012). Social constraints, posttraumatic cognitions, and posttraumatic stress disorder in treatment-seeking trauma survivors: Evidence for a social-cognitive processing model. Psychological Trauma: Theory Research Practice and Policy, 4(4), 386–391. https://doi.org/10.1037/a0024362

10.

Berkman

J. M.

(2005). Posttraumatic stress symptoms after pediatric injury: The role of family factors (Publication No. 3189850) [Doctoral dissertation, Clark University]. ProQuest Dissertations and Theses Global.

11.

Berkowitz

S. J.

Stover

C. S.

Marans

S. R.

(2011). The child and family traumatic stress intervention: Secondary prevention for youth at risk of developing PTSD. Journal of Child Psychology and Psychiatry, 52(6), 676–685. https://doi.org/10.1111/j.1469-7610.2010.02321.x

12.

Birkley

E. L.

Eckhardt

C. I.

Dykstra

R. E.

(2016). Posttraumatic stress disorder symptoms, intimate partner violence, and relationship functioning: A meta-analytic review. Journal of Traumatic Stress, 29(5), 397–405. https://doi.org/10.1002/jts.22129

13.

Bokszczanin

(2008). Parental support, family conflict, and overprotectiveness: Predicting PTSD symptom levels of adolescents 28 months after a natural disaster. Anxiety, Stress, and Coping, 21(4), 325–335. https://doi.org/10.1080/10615800801950584

14.

Borenstein

Cooper

Hedges

Valentine

(2009). Effect sizes for continuous data. The Handbook of Research Synthesis and Meta-Analysis, 2, 221–235.

15.

Brockwell

S. E.

Gordon

I. R.

(2001). A comparison of statistical methods for meta-analysis. Statistics in Medicine, 20(6), 825–840. https://doi.org/10.1002/sim.650

16.

Bronfenbrenner

(1992). Ecological systems theory. Jessica Kingsley Publishers.

17.

Card

N. A.

(2016). Applied meta-analysis for social science research. Guilford Press.

18.

Cheung

M. W. L.

(2014). Modeling dependent effect sizes with three-level meta-analyses: A structural equation modeling approach. Psychological Methods, 19(2), 211–229. https://doi.org/10.1037/a0032968

19.

Cohen

J. A.

(1992). A power primer. Psychological Bulletin, 112(1), 155–159. https://doi.org/10.1037//0033-2909.112.1.155

20.

Cohen

J. A.

Mannarino

A. P.

(2008). Trauma-focused cognitive behavioural therapy for children and parents. Child and Adolescent Mental Health, 13(4), 158–162. https://doi.org/10.1111/j.1475-3588.2008.00502.x

21.

Cooper

Hedges

L. V.

Valentine

J. C.

(Eds.). (2019). The handbook of research synthesis and meta-analysis. Russell Sage Foundation.

22.

Danielson

C. K.

Cohen

J. R.

Adams

Z. W.

Youngstrom

E. A.

Soltis

Amstadter

A. B.

Ruggiero

K. J.

(2017). Clinical decision-making following disasters: Efficient identification of PTSD risk in adolescents. Journal of Abnormal Child Psychology, 45(1), 117–129. https://doi.org/10.1007/s10802-016-0159-3

23.

Daniunaite

Cloitre

Karatzias

Shevlin

Thoresen

Zelviene

Kazlauskas

(2021). PTSD and complex PTSD in adolescence: Discriminating factors in a population-based cross-sectional study. European Journal of Psychotraumatology, 12(1), 1890937. https://doi.org/10.1080/20008198.2021.1890937

24.

Deane

Richards

Mozley

Scott

Rice

Garbarino

(2018). Posttraumatic stress, family functioning, and externalizing in adolescents exposed to violence: A moderated mediation model. Journal of clinical child and adolescent psychology, 47(1), S176–S189. https://doi.org/10.1080/15374416.2016.1197836

25.

Duval

Tweedie

(2000). Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2), 455–463. https://doi.org/10.1111/j.0006-341x.2000.00455.x

26.

Egger

Davey Smith

Schneider

Minder

(1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629–634. https://doi.org/10.1136/bmj.315.7109.629

27.

Epstein

N. B.

Baldwin

L. M.

Bishop

D. S.

(1983). The McMaster family assessment device. Journal of Marital and Family Therapy, 9(2), 171–180. https://doi.org/10.1111/j.1752-0606.1983.tb01497.x

28.

Epstein

N. B.

Bishop

D. S.

Levin

(1978). The McMaster model of family functioning. Journal of Marital and Family Therapy, 4(4), 19–31. https://doi.org/10.1111/j.1752-0606.1978.tb00537.x

29.

Fisher

R. A.

(1925). Theory of statistical estimation. Mathematical Proceedings of the Cambridge Philosophical Society, 22(5), 700–725. https://doi.org/10.1017/S0305004100009580

30.

Foa

E. B.

McLean

C. P.

Capaldi

Rosenfield

(2013). Prolonged exposure vs supportive counseling for sexual abuse–related PTSD in adolescent girls: A randomized clinical trial. JAMA, 310(24), 2650–2657. https://doi.org/10.1001/jama.2013.282829

31.

Gaipa

J. J. L.

(1990). The negative effects of informal support systems. In Duck

(ed.), Personal relationships and social support (pp. 122–139). SAGE.

32.

Gallo

Wertz

Kairis

Blavier

(2019). Exploration of relationship between parental distress, family functioning and post-traumatic symptoms in children. European Journal of Trauma and Dissociation, 3(2), 125–133. https://doi.org/10.1016/j.ejtd.2018.06.001

33.

Gignac

G. E.

Szodorai

E. T.

(2016). Effect size guidelines for individual differences researchers. Personality and Individual Differences, 102, 74–78. https://doi.org/10.1016/j.paid.2016.06.069

34.

Gorman-Smith

Henry

D. B.

Tolan

P. H.

(2004). Exposure to community violence and violence perpetration: the protective effects of family functioning. Journal of Clinical Child and Adolescent Psychology, 33(3), 439–449. https://doi.org/10.1207/s15374424jccp3303_2

35.

Green

B. L.

Goodman

L. A.

Krupnick

J. L.

Corcoran

C. B.

Petty

R. M.

Stockton

Stern

N. M.

(2000). Outcomes of single versus multiple trauma exposure in a screening sample. Journal of Traumatic Stress, 13(2), 271–286. https://doi.org/10.1023/A:1007758711939

36.

Guerrero-Munoz

Salazar

Constain

Perez

Andres Pineda-Canar

Andres Garcia-Perdomo

(2021). Association between family functionality and depression: A systematic review and meta-analysis. Korean Journal of Family Medicine, 42(2), 172–180. https://doi.org/10.4082/kjfm.19.0166

37.

Higgins

J. P.

Thompson

S. G.

Deeks

J. J.

Altman

D. G.

(2003). Measuring inconsistency in meta-analyses. BMJ, 327(7414), 557–560. https://doi.org/10.1136/bmj.327.7414.557

38.

Hildenbrand

A. K.

(2016). Childhood trauma and posttraumatic stress in pediatric amplified musculoskeletal pain syndromes (Publication No. 10253925) [Doctoral dissertation, Drexel University]. ProQuest Dissertations and Theses Global.

39.

Holtom-Viesel

Allan

(2014). A systematic review of the literature on family functioning across all eating disorder diagnoses in comparison to control families. Clinical Psychology Review, 34(1), 29–43. https://doi.org/10.1016/j.cpr.2013.10.005

40.

Jaycox

L. H.

Cohen

J. A.

Mannarino

A. P.

Walker

D. W.

Langley

A. K.

Gegenheimer

K. L.

Scott

Schonlau

(2010). Children’s mental health care following Hurricane Katrina: A field trial of trauma-focused psychotherapies. Journal of Traumatic Stress, 23(2), 223–231. https://doi.org/10.1002/jts.20518

41.

Kaur

Kearney

C. A.

(2013). Ethnic identity, family cohesion, and symptoms of post-traumatic stress disorder in maltreated youth. Journal of Aggression Maltreatment and Trauma, 22(10), 1085–1095. https://doi.org/10.1080/10926771.2013.845278

42.

Kiliç

E. Z.

Özgüven

H. D.

Sayil

(2003). The psychological effects of parental mental health on children experiencing disaster: The experience of bolu earthquake in Turkey. Family Process, 42(4), 485–495. https://doi.org/10.1111/j.1545-5300.2003.00485.x

43.

Koirala

Soegaard

E. G. I.

Ojha

S. P.

Thapa

S. B.

Hauff

(2020). Trauma related psychiatric disorders and their correlates in a clinical sample: A cross-sectional study in trauma affected patients visiting a psychiatric clinic in nepal. PLoS ONE, 15(6), e0234203. https://doi.org/10.1371/journal.pone.0234203

44.

Kontopantelis

Reeves

(2010). Metaan: random-effects meta-analysis. The Stata Journal, 10(3), 395–407. https://doi.org/10.1177/1536867x1001000307

45.

Kowalik

Weller

Venter

Drachman

(2011). Cognitive behavioral therapy for the treatment of pediatric posttraumatic stress disorder: A review and meta-analysis. Journal of Behavior Therapy and Experimental Psychiatry, 42(3), 405–413. https://doi.org/10.1016/j.jbtep.2011.02.002

46.

Lambert

J. E.

Holzer

Hasbun

(2014). Association between parents’ PTSD severity and children’s psychological distress: A meta-analysis. Journal of Traumatic Stress, 27(1), 9–17. https://doi.org/10.1002/jts.21891

47.

Layne

C. M.

Warren

J. S.

Saltzman

W. R.

Fulton

J. B.

Steinberg

A. M.

Pynoos

R. S.

(2018). Contextual influences on posttraumatic adjustment: retraumatization and the roles of revictimization, posttraumatic adversities, and distressing reminders. In Schein

L. A.

Spitz

H. I.

Burlingame

G. M.

Muskin

P. R.

Vargo

(Eds.), Psychological effects of catastrophic disasters (pp. 235–286). Routledge. https://doi.org/10.4324/9781315821306-9

48.

Lee

S. H.

Kim

E. J.

Noh

J. W.

Chae

J. H.

(2018). Factors associated with post-traumatic stress symptoms in students who survived 20 months after the Sewol Ferry Disaster in Korea. Journal of Korean Medical Science, 33(11), e90. https://doi.org/10.3346/jkms.2018.33.e90

49.

Lipsey

M. W.

Wilson

D. B.

(2001). Practical meta-analysis. SAGE.

50.

McCarthy

M. D.

Thompson

S. J.

(2010). Predictors of trauma-related symptoms among runaway adolescents. Journal of Loss and Trauma, 15(3), 212–227. https://doi.org/10.1080/15325020903375842

51.

McCarty

C. A.

McMahon

R. J.

(2003). Mediators of the relation between maternal depressive symptoms and child internalizing and disruptive behavior disorders. Journal of Family Psychology, 17(4), 545–556. https://doi.org/10.1037/0893-3200.17.4.545

52.

McFarlane

A. C.

(1987). Family functioning and overprotection following a natural disaster: The longitudinal effects of post-traumatic morbidity. Australian New Zealand Journal of Psychiatry, 21(2), 210–218. https://doi.org/10.3109/00048678709160914

53.

Minuchin

(2018). Families and family therapy. Routledge.

54.

Modesti

P. A.

Reboldi

Cappuccio

F. P.

Agyemang

Remuzzi

Rapi

Perruolo

Parati

& ESH Working Group on CV Risk in Low Resource Settings. (2016). Panethnic differences in blood pressure in europe: A systematic review and meta-analysis. PLoS ONE, 11(1), e0147601. https://doi.org/10.1371/journal.pone.0147601

55.

Moher

Liberati

Tetzlaff

Altman

D. G.

Group

(2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLOS Medicine, 6(7), e1000097. https://doi.org/10.1371/journal.pmed.1000097

56.

Mora

A. S.

Ceballo

Cranford

J. A.

(2022). Latino/a adolescents facing neighborhood dangers: An examination of community violence and gender-based harassment. American Journal of Community Psychology, 69(1–2), 18–32. https://doi.org/10.1002/ajcp.12556

57.

Mørkved

Hartmann

Aarsheim

L. M.

Holen

Milde

A. M.

Bomyea

Thorp

S. R.

(2014). A comparison of narrative exposure therapy and prolonged exposure therapy for PTSD. Clinical Psychology Review, 34(6), 453-467. https://doi.org/https://doi.org/10.1016/j.cpr.2014.06.005

58.

Nelson

L. P.

Lachman

S. E.

S. W.

Gold

J. I.

(2019). The effects of family functioning on the development of posttraumatic stress in children and their parents following admission to the PICU. Pediatric Critical Care Medicine, 20(4), e208–e215. https://doi.org/10.1097/PCC.0000000000001894

59.

Ohannessian

C. M.

Lerner

R. M.

Lerner

J. V.

von Eye

(2000). Adolescent-parent discrepancies in perceptions of family functioning and early adolescent self-competence. International Journal of Behavioral Development, 24(3), 362–372. https://doi.org/10.1080/01650250050118358

60.

Olson

D. H.

(2000a). Circumplex model of marital and family systems. Journal of Family Therapy, 22(2), 144–167. https://doi.org/10.1111/1467-6427.00144

61.

Olson

D. H.

(2000b). Circumplex model of marital and family systems: Assessing family functioning. Journal of family therapy, 22(2), 144–167. https://doi.org/10.1111/1467-6427.00144

62.

Ozono

Saeki

Mantani

Ogata

Okamura

Yamawaki

(2007). Factors related to posttraumatic stress in adolescent survivors of childhood cancer and their parents. Support Care Cancer, 15(3), 309–317. https://doi.org/10.1007/s00520-006-0139-1

63.

Pereda

Díaz-Faes

D. A.

(2020). Family violence against children in the wake of COVID-19 pandemic: A review of current perspectives and risk factors. Child and adolescent psychiatry and mental health, 14(1), 40. https://doi.org/10.1186/s13034-020-00347-1

64.

Peverill

Dirks

M. A.

Narvaja

Herts

K. L.

Comer

J. S.

McLaughlin

K. A.

(2021). Socioeconomic status and child psychopathology in the United States: A meta-analysis of population-based studies. Clinical psychology review, 83, 101933. https://doi.org/10.1016/j.cpr.2020.101933

65.

Pinquart

(2020). Posttraumatic stress symptoms and disorders in children and adolescents with chronic physical illnesses: A meta-analysis. Journal of Child and Adolescent Trauma, 13(1), 1–10. https://doi.org/10.1007/s40653-018-0222-z

66.

Robjant

Fazel

(2010). The emerging evidence for Narrative Exposure Therapy: A review. Clinical Psychology Review, 30(8), 1030–1039. https://doi.org/10.1016/j.cpr.2010.07.004

67.

Rodenburg

Benjamin

de Roos

Meijer

A. M.

Stams

G. J.

(2009). Efficacy of EMDR in children: A meta-analysis. Clinical Psychology Review, 29(7), 599–606. https://doi.org/10.1016/j.cpr.2009.06.008

68.

Rodgers

M. A.

Pustejovsky

J. E.

(2021). Evaluating meta-analytic methods to detect selective reporting in the presence of dependent effect sizes. Psychological Methods, 26(2), 141–160. https://doi.org/10.1037/met0000300

69.

Rothstein

H. R.

Sutton

A. J.

Borenstein

(Eds.). (2006). Publication bias in meta-analysis: Prevention, assessment and adjustments. John Wiley & Sons.

70.

Rowe

C. L.

La Greca

A. M.

Alexandersson

(2010). Family and individual factors associated with substance involvement and PTS symptoms among adolescents in greater new orleans after Hurricane Katrina. Journal of Consulting and Clinical Psychology, 78(6), 806–817. https://doi.org/10.1037/a0020808

71.

Sadeh

Dekel

Brezner

Landa

Silberg

(2020). Child and family factors associated with posttraumatic stress responses following a traumatic medical event: The role of medical team support. Journal of Pediatric Psychology, 45(9), 1063–1073. https://doi.org/10.1093/jpepsy/jsaa070

72.

Scully

McLaughlin

Fitzgerald

(2020). The relationship between adverse childhood experiences, family functioning, and mental health problems among children and adolescents: A systematic review. Journal of Family Therapy, 42(2), 291–316. https://doi.org/10.1111/1467-6427.12263

73.

Shelton

L. G.

(2018). The bronfenbrenner primer: A guide to develecology. Routledge.

74.

Skinner

Steinhauer

Sitarenios

(2000). Family assessment measure (FAM) and process model of family functioning. Journal of Family Therapy, 22(2), 190–210. https://doi.org/10.1111/1467-6427.00146

75.

Trickey

Siddaway

A. P.

Meiser-Stedman

Serpell

Field

A. P.

(2012). A meta-analysis of risk factors for post-traumatic stress disorder in children and adolescents. Clinical Psychology Review, 32(2), 122–138. https://doi.org/10.1016/j.cpr.2011.12.001

76.

Velotti

Rogier

Zobel

S. B.

Billieux

(2021). Association between gambling disorder and emotion (dys) regulation: A systematic review and meta-analysis. Clinical Psychology Review, 87, 102037. https://doi.org/10.1016/j.cpr.2021.102037

77.

Vliem

S. J.

(2009). Adolescent coping and family functioning in the family of a child with autism (Publication No. 3392949) [Doctoral dissertation, University of Michigan]. ProQuest Dissertations and Theses Global.

78.

Walsh

(2007). Traumatic loss and major disasters: strengthening family and community resilience. Family Process, 46(2), 207–227. https://doi.org/10.1111/j.1545-5300.2007.00205.x

79.

Wedding

Corsini

R. J.

(2019). Current psychotherapies. Cengage Learning.

80.

Wells

G. A.

Shea

O’Connell

Peterson

Welch

Losos

Tugwell

(2001). The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Retrieved January 4, 2022, from http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

81.

Whealin

J. M.

Nelson

Stotzer

Guerrero

Carpenter

Pietrzak

R. H.

(2015). Risk and resilience factors associated with posttraumatic stress in ethno-racially diverse National Guard members in Hawai’i. Psychiatry Research, 227(2), 270–277. https://doi.org/10.1016/j.psychres.2015.02.023

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.43 MB