Introduction

Research examining families of service members returning from the wars in Afghanistan (Operation Enduring Freedom, OEF) and Iraq (Operation Iraqi Freedom/New Dawn, OIF/OND) documents the presence of PTSD and other emotional and physical challenges in returning service members (Institute of Medicine 2010; Shea et al. 2010). Among Iraq and Afghanistan veterans, PTSD rates are as high as 20% (Seal et al. 2007, 2009; Tanielian and Jaycox 2008). In addition, there is recent research pointing to parental status as a risk factor in the development of service member PTSD. In a recent review of nearly 100,000 records from Veterans Health Administration (VA) databases of veterans seeking care, results indicated that parental status was highly associated with PTSD (44% of parents vs. 28% nonparents) (Janke-Strednosky et al. 2016). Further, these associations were more robust for veteran parents of dependent children and for fathers specifically. While these findings suggest that being a parent of a child or adolescent may convey unique risk for the development of posttraumatic stress symptoms (Janke-Stredronsky et al. 2016), another possibility is that parents with dependent children may be more motivated to seek treatment for mental health distress. Although realignment of parenting and coparenting roles is undoubtedly a central challenge of reintegration, parental posttraumatic stress is also likely to influence child distress during post-deployment (Cozza et al. 2010; Ruscio et al. 2002). Thus, while PTSD and related psychological sequelae from war have been characterized as “invisible wounds,” they can be powerful forces within the family system (Tanielian and Jaycox 2008), and may pose unique challenges in the parenting of very young children.

Trauma scholars often invoke theories of intergenerational transmission of distress through parental narrative, parent behavior and compromised parenting style (Sampler et al. 2004), and most recently, biological variables. A review of 100 studies that evaluated psychological and biological variables among children of parents with PTSD—non-specific to military or combat— concluded that parental symptoms of PTSD are uniquely related to children’s outcomes of internalizing-type problems, general behavioral problems, and altered hypothalamic-pituitary-adrenal axis functioning. These authors identified elevated family hostility and parental disengagement, which may reflect psychological avoidance and numbing responses, as key pathways through which PTSD might impact parenting (Leen-Feldner et al. 2013). As PTSD symptoms are negatively related to nearly all aspects of marital functioning, it is logical that they also negatively affect parenting alliance and coparenting support between spouses (Allen et al. 2010). The reciprocal relationship between PTSD and marital discord can put a strain on parenting resources in military families with children and may exacerbate service member difficulties with reintegrating into family roles (Sayers et al. 2009). Findings that families of veterans with PTSD are more likely to exhibit severe problems with parenting skills (54.7% in PTSD group vs. 17.3% in non-PTSD group) (Jordan et al. 1992) and concern for child-rearing practices (Khaylis et al. 2011) support this model.

Parenting research, historically, has focused on the influence of mothering and the impact of maternal-child interactions on children’s well-being. In particular, maternal depression has been found to have very negative effects on child outcomes (e.g., Kim-Cohen et al. 2005) as diminished responsiveness, inconsistent parenting, insensitivity, and rejection of the child are among the possible consequences. However, the extent to which parenting by service member fathers influences children’s status is less well known, as are the mechanisms by which service members’ trauma-related distress affects their parenting and young children’s well-being. In research on National Guard fathers, PTSD symptoms one month before returning home from OIF deployments were associated with poorer couple adjustment and greater perceived parenting challenges one-year post-deployment. The association between PTSD symptoms and parenting problems was independent of the impact on couple adjustment (Gewirtz et al. 2010). In a study of parents who had recently deployed to the Middle East, self-reported associated with less positive engagement with children (ages four symptoms of experiential avoidance among service member fathers were-13 years) and more withdrawal in a videotaped family interaction task (Brockman et al. 2015).

A growing body of research has focused on associations between mental health, parenting stress and deployment-related distress in parents of very young children (Barker and Berry 2009; Gewirtz and Zamirt 2013). Number and length of parental deployment have been consistently correlated with greater parenting stress and mental health concerns among both service member and home-front parents (Trautman et al. 2015), as well as with worrisome child outcomes (Chandra et al. 2010). In addition, home-front parents may experience increased psychological and social burdens related to anxiety about the safety of the deployed spouse, and greater role responsibilities associated with household management, financial management, and single parenting stressors (Barker and Berry 2009; DeVoe and Ross 2012; Flake et al. 2009). In families with very young children, parental deployment is associated with higher parental reports of externalizing behavior (Chartrand et al. 2008), and more behavioral health visits related to stress (Gorman et al. 2010; Hisle-Gorman et al. 2015).

Given the limited empirical literature on parenting among post-9/11 families coping with service member post-traumatic stress symptoms, the goal of this study was to examine the relationships between parental mental health status and parenting stress, quality of the couple relationships, and child well-being. Because the aim of the larger intervention study was to develop a universal rather than targeted program, we expected relatively low base rates of parent mental health distress and parent-reported child behavior problems. We considered separately the effects of mental health status of home-front parents and service members on parent-child interactions and parental perceptions of child behavioral problems. We hypothesized that higher self-reported parent mental health distress would be associated with higher rates of parenting stress (PSI scores) and perceptions of child behavior problems (CBCL scores) among both service members and home-front parents. We also hypothesized that couple relationship quality would be negatively associated with PSI and CBCL scores.

Method

Participants

A total of 169 OEF/OIF service members and 136 home-front parents participated in this research. Families were recruited in the northeast through multiple sites, including the Yellow Ribbon Programs in several states, National Guard Family Assistance Centers, and local programs associated with veteran care. Research staff obtained consent to contact forms from all potential participants who were approached in person, and later contacted each family who completed a form to explain the study and schedule baseline interviews. Families were eligible for participation if the following conditions were met: (a) they were in the reintegration phase of the deployment cycle; (b) they had at least one child age five (5) years or younger during the most recent deployment, (c) the child co-resided with one parent, (d) the family residence was no more than 90 min driving distance from our campus, and (e) they expressed willingness to participate in a home-based research protocol. Families were excluded if (a) either parent participant had current suicidal or psychotic symptoms or other concerns necessitating a higher level of care; or (b) either parent had active substance abuse that would interfere with their capacity to engage in or consent to the research protocol. Sample size varies somewhat by measure because not all families were headed by two parents and because of age requirements for some instruments, for example the Child Behavior Checklist (1.5–5 years).

Procedure

The data analyzed in this study were collected from baseline interviews with service member and home-front parents across all phases of the larger research project. Interviews and surveys were administered in participant homes by graduate level social work and psychology clinicians who received extensive training and ongoing supervision in all study protocols. Measures included in this research include family demographic characteristics, self-reported parenting stress and mental health distress, and parent report of child emotional and behavioral problems. Data were analyzed using SPSS Version 20.

Measures

The Child Behavior Check List (CBCL; Achenbach and Rescorla 2000) for children ages 1 ½ to 5 years is a parent report measure of children’s internalizing and externalizing behavior. The CBCL (1.5–5) has 13 subscales and assess two broad categories of child functioning: Internalizing (e.g., anxiety, somatic complaints, depression) and Externalizing Problems (e.g., attention problems, aggressive behavior). The CBCL total score was used in the analyses presented here. Internal consistency for service members (N = 121) and home-front parents (N = 104) were .92 and of .94 (Cronbach’s α), respectively.

The Parenting Stress Index-4, Short Form (PSI; Abidin 1997) is a 36-item self-report measure that assesses the parental distress in relation to interactions with children, children’s behaviors, and aspects of parental roles such as sense of competence. We used the 15-item Parent-Child Dysfunctional Interaction (P-CDI), which focuses on parental perceptions that the child does not meet the parent’s expectations, feelings of rejection by the child, and alienation from the child. High scores reflect inadequate parent-child bonds (Abidin 1997). Sample items include, “My child rarely does things that make me feel good,” and “Sometimes my child does things that bother me just to be mean.” Internal consistency reliability (Cronbach’s α) of 0.93 was obtained.

The Brief Symptom Inventory (BSI; Derogatis 1975/1993; Derogatis and Melisaratos 1983) is a 53-item self-report measure that assesses psychological symptom patterns in adults. We selected the Global Severity Index (GSI) scale for analysis because it provides the most comprehensive assessment of psychological distress and symptom severity. For multivariate analyses, we used the GSI only for home-front parents, because we believe it may better capture the comprehensive nature of current emotional distress, beyond trauma, experienced by home-front parents. For the current analyses (N = 132), internal consistency reliability (Cronbach’s α) was 0.95.

The Posttraumatic Stress Disorder Checklist—Military/Civilian (PCL-M/C; Weathers et al. 1994; Weathers et al. 1991) is a 17-item self-report measure of posttraumatic stress symptoms. Items correspond to the DSM-IV symptom categories, and include a 5-point response scale reflecting the degree of distress associated with each symptom. The PCL-M anchors distress to military-related events while the PCL-C asks about symptoms associated with “generic stressful experiences” rather than a specific event (National Center for PTSD 2014). Total Score means and DSM-IV positives were computed, along with subthreshold scores at the cut-off of 30 for both the service member and home-front parent. These lower bound cut-off scores reflect recommendations for identifying subthreshold distress and problem detection among primary care populations and in military screening settings (Bliese et al. 2008). For these analyses, we obtained an internal consistency reliability of .949 for the PCL-M (N = 159) and .925 for the PCL-C (N = 135).

The Dyadic Adjustment Scale (DAS) is a 32-item measure of couple relationship quality and satisfaction. The DAS includes a total score and four subscales: (1) Dyadic Satisfaction, (2) Dyadic Consensus, (3) Dyadic Cohesion, and (4) Affective Expression. Items have a variety of response scales with higher scores representing greater relationship satisfaction. Spanier reported good validity and reliability (0.96 full scale and range of 0.73–0.94 for subscales). In a more recent meta-analysis examining 91 studies, the DAS performed with acceptable to high internal consistency reliability, depending upon the subscale. In this study, internal consistency reliability (Cronbach’s α) was 0.91 for the total DAS score.

Data Analyses

Descriptive analyses examined the frequencies, means, ranges, standard deviations, and correlations for all variables and composite scores used in this study. Differences in couple report were assessed with a pooled variance t-test (i.e., CBCL, PSI, BSI, PCL, and DAS, see Table 1). Hierarchical regression analyses were conducted for service members and home-front parents to examine the impact of post-traumatic stress, as measured by the PCL, overall psychological stress, as measured by the BSI, and relationship strength, as measured by the DAS, on child behaviors, measured by the CBCL, and parental stress, as measured by the PSI. Demographic variables, including child gender and age, number of children in the household, parental age, service member pay grade (i.e., enlisted or officer), and number of out-of-country deployments, were entered in the first block. Following this, individual psychological variables were entered (i.e., PCL and BSI) in the second block. Finally, relationship strength (DAS) was entered in the third block. Measures of variable overlap were examined (i.e., variance inflation factor), and analyses were adjusted accordingly. All p values are two-tailed and p ≤ .05 was considered statistically significant.

Table 1 Descriptive statistics for military family members
Full size table

Results

Demographic Characteristics

The majority of service member parents were male (85%), and home-front parents were overwhelmingly female (96%). The average age of service member parents was 34.7 years (SD = 7.19), and of home-front parents was 32.8 years (SD = 6.39). Children were 3.71 years old (SD = 2.11) and 54% were male. Families had 1.95 (SD = 1.34) children on average. Most (79%) families of military service members have pay grades consistent with the Enlisted ranks. Service members had undergone 1.7 overseas deployments, on average, since September, 2001.

Child Emotional and Behavioral Status

Parent report versions of the CBCL show mean scores that are well below clinical ranges (70 or above) whether reported by home-front or service member parents. The parents report similar levels of problem behavoior (see Table 1)

Parenting

Assessments of parenting using the Parenting Stress Index (PSI) were analyzed to examine parenting-related stress and aspects of parental roles such as the parent’s sense of competence. Descriptive statistics presented in Table 1 show that both parents generally score within the normal range of parenting stress, though 15% of parents fell within clinically significant ranges of parental stress (≥90).

Adult Mental Health Status

Two measures of parental mental health, the Global Severity Index from the Brief Symptom Inventory (BSI) and the Post Traumatic Stress Disorder Checklist (the PCL-Military and PCL-C for civilians) were used to assess adult mental health status. Total Score means, DSM-IV positives, and subthreshold scores (30 cut-off) (Bliese et al. 2008) are reported. Results presented in Table 1 for the total BSI scores indicate that most parents do not fall within clinically diagnostic ranges for psychological disorders but service member parents have higher mean scores relative to home-front parents. These differences are consistent with expectations about psychological distress and trauma for service members. Although overall DSM-IV diagnostic positive rates are low, service members have rates (11.3%) which are triple those of home-front parents (3.7%). In addition, using a primary care cut-off on the PCL, over a third (37%) of service members, compared to 15% of home-front parents, score in the subthreshold distress range.

Additional analyses (not shown in Table) calculated the relative difference in scores on the CBCL, PSI-PCD and mental health measures for the BSI-GSI and PCL-M/PCL-C. We tested differences for a sub-sample of 100 couples on whom data were available for both members of a couple, and used a pooled variance t-test to examine the significance of differences for family member reports on parallel tests. The results showed that means do not differ significantly for any of the measures with the exception of the PCL, with higher scores found for service members on the PCL-M compared to all PCL-C results for at-home spouses, t(198) = 3.57, p < .001. The BSI-GSI did not reach statistical significance, t(198) = 1.53, p = .13 despite a substantial difference in means due to the high variability in scores.

Tables 2 and 3 provide separate correlation matrices for service members and home-front parents for variables used in the relevant multivariate analyses. Table 2 displays correlations among parenting, mental health and demographic factors for service members. Few of the demographic variables were significantly correlated with any other variables; however, the PSI, CBCL, PCL, BSI, and DAS correlated strongly with one another.

Table 2 Correlations among parenting, mental health and demographic factors for service members parents
Full size table
Table 3 Correlations among parenting, mental health and demographic factors for home-front parents
Full size table

In the next phase of analyses, we conducted a series of hierarchical regressions, examining the PSI and the CBCL for both service members and their spouses, in which we first entered a block of demographic variables, child’s gender, child’s age, number of children, paygrade, number of overseas (OCONUS) deployments, and time back since deployment, then a measure of traumatic symptoms, PCL, and finally a measure of relationship stress, DAS. For home-front parents we included the BSI as a general measure of mental health status with the PCL. For service members, we removed the BSI due to multicollinearity and retained the PCL given its salience for this population.

Table 4 gives the results of the service member regression examining the PSI. The initial regression of the PSI on the demographic variables was significant, F(6, 118) = 2.58, p = .02, R2 = .12. In the second block, the PCL significantly predicts the PSI, b = 0.50, t(117) = 3.90, p < .001. In the third step, examining relationship stress, the DAS also significantly predicts the PSI, b = −0.27, t(116) = −3.06, p = .003. Table 5 gives the results of the service member regression examining the CBCL. The initial regression of the CBCL on the demographic variables was not significant, F(6, 92) = 1.99, p = .08, R2 = .12. In the second block, the PCL significantly predicts the CBCL, b = 0.42, t(91) = 3.69, p < .001. In the third step, examining relationship stress, the DAS did not predict the CBCL, b = 0.01, t(90) = 0.08, p = .94.

Table 4 Regressions of parental stress (PSI) on service members’ symptoms (N = 125)
Full size table
Table 5 Regressions of child behavior (CBCL) on service members’ symptoms (N = 99)
Full size table

Table 6 gives the results of the home-front parents’ regression examining the PSI. The initial regression of the PSI on the demographic variables was not significant, F(6, 117) = 0.60, p = .73, R2 = .03. In the second block, the addition of the PCL and BSI together significantly predict the PSI, R2 change = .17, F(2, 115) = 11.90, p < .001. The BSI significantly predicts the PSI, b = .28, t(115) = 2.60, p = .01, though the PCL-C does not, b = 0.19, t(115) = 0.68, p = .50. In the third step, examining relationship stress, the DAS significantly predicts the PSI, b = −0.22, t(114) = −2.52, p = .01.

Table 6 Regressions of parental stress (PSI) on home-front parents’ symptoms (N = 124)
Full size table

Table 7 gives the results of the home-front parents’ regression examining the CBCL. The initial regression of the CBCL on the demographic variables was not significant, F(6, 92) = 1.56, p = .17, R2 = .09. In the second block, the addition of the PCL and BSI together significantly predict the CBCL, R2 change = .07, F(2, 90) = 3.45, p = .04; however, neither variable individually significantly predicts the CBCL, b = 0.20, t(90) = 0.55, p = .58, and b = 0.16, t(90) = 1.19, p = .24, respectively. In the third step, examining relationship stress, the DAS nearly reaches significance, b = −0.22, t (89) = −1.98, p = .05.

Table 7 Regressions of child behavior (CBCL) on home-front parents’ symptoms (N = 99)
Full size table

Discussion

This study of 169 post-9/11 service members and 136 home-front parents sought to explore the relationships between family members’ mental health status and couple satisfaction, parenting, and child well-being. We considered separately the effects of mental health status of home-front parents and service member parents and quality of the couple relationship on parent-child interactions and parents’ perceptions of young child behavior. Research from prior studies suggests that military trauma may affect all members of military families. PTSD contributes to familial distress (Glenn et al. 2002; Gold et al. 2007), poor family communication, affective responsiveness, difficulty with problem solving and impaired parenting (Gewirtz et al. 2010; Gold et al. 2007; Ruscio et al. 2002; Samper et al. 2004; Solomon et al. 2011). The severity of service members’ posttraumatic stress symptoms has been linked to more mental health service utilization and diagnoses among wives whose husbands deployed (Mansfield et al. 2011) compared to wives of non-deployed service members. Evidence from studies of prior war veterans coupled with emerging studies of post-9/11 families suggests that the severity of veterans’ trauma symptoms can impair interactions with children and create difficulties with the quality of intimacy perceived by the spouse (Brockman et al. 2015; Riggs et al. 1998).

In this study, most parents did not fall within a probable clinical range for psychological disorders. Service members also had PCL-M rates (11.3%) that were more than triple those rates (PCL-C) of home-front parents (3.7%), although mean ratings were similar. Notably, these rates are lower than those reported in the literature indicating the relative health of this community sample, though 37% of service members evidenced subthreshold posttraumatic stress symptoms. Longitudinal research is needed, initiated prior to deployment, to tease apart the temporal relationship between service member and spouse mental health distress and couple functioning through deployment cycles.

Analysis of the CBCL found that average levels of child behavior problems are well below clinical ranges, whether reported by at-home or service member parents. Based upon the longer time spent with their children, we expected that home-front parents would endorse more problematic child behaviors than service member parents who had recently returned home from a warzone. In addition, upon the return of the service member, we have observed that families sometimes experience a “honeymoon period” in which children’s challenging behaviors are less frequent and less noticeable, perhaps especially to the returning parent.

Descriptive statistics showed that generally parents have, on average, total PSI subscale scores that fall within the normal range of parenting stress, though 15% of both service member and home-front parents fell within clinically significant ranges of parental stress (≤90). Recent qualitative research describes service member parents’ (mostly fathers’) significant concerns about how they will be received by their young children upon return from deployment (DeVoe and Paris 2015; Author; Dayton et al. 2014; Trautmann et al. 2015). Further observations suggest that service member parents of young children may experience a transient lack of confidence in parenting and limited developmental knowledge in the context of early reintegration. However, our hypothesis that service member parents would endorse more challenging child behavior was not supported.

Our analyses found differences in associations between self-reported parental trauma and distress, parenting and child behavior problems for military service members and home-front parents. Bivariate and multivariate analyses conducted for service members provided evidence of a significant relationship between dysfunctional parenting (PSI-PCD), child behavior problems (CBCL) and PTSD symptom severity. This finding is consistent with the literature, as well as the notion that emotional numbing and avoidance limit opportunities for parental responsiveness and sensitivity necessary to support parent-child attachment. Further, as seen with cases of maternal depression, reduced responsiveness, inconsistent parenting, insensitivity and rejection of the child are also possible consequences (Kim-Cohen et al. 2005) of either diminished or emotionally impoverished interactions.

Analyses examining the relationship between self-reported parental trauma and parent-reported child behavior problems among service member parents found similar patterns, but some small differences as well. Once again, PTSD total symptom severity was significantly associated with an increase in parental report of child behavior problems. It is possible that the unpredictability and nature of parental responses associated with re-experiencing symptoms related to deployment or combat events are more likely to invoke challenging or intense child behaviors, which may be perceived as problematic. Another inference is that developmentally appropriate behavior among very young children, such as making loud noises, moving around, and engaging in household activity, can be provocative and serve to trigger trauma-related distress in recently returned service member parents.

The results for home-front parents differ from those noted for service member parents in that there were fewer significant associations, and relationships were more modest. Home-front parents reported lower levels of trauma and distress relative to their service member partners. Results of the series of regressions conducted for home-front parents found that current emotional distress, as measured by the BSI-GSI, significantly (p < .05) predicted parenting challenges. Further, this equation explained the most variance relative to home-front parental prior trauma. However, statistical significance aside, the analyses also show that the explanatory power of the at-home parent’s prior trauma or current emotional distress is very limited in accounting for parenting dysfunction, as we would expect given the very low base rates of distress. Subsequent analyses showed that current at-home parental emotional distress was also most strongly associated with child behavior problems (p < .001). In contrast to the analysis for parenting dysfunction, this equation accounted for a significant and meaningful portion of explained variance in child behavior problems (Adjusted R2 = .138 or 14%) reported by home-front parents.

Finally, the quality of the couple’s relationship was associated with parenting stress but not reported child behavior problems. That is, regardless of couple relationship quality, higher parental PTSD was associated with higher parent report of child problems. This finding may offer additional intervention opportunities for couples in which one partner is experiencing PTSD and parenting stress, namely couples therapy focusing on improving the quality of the relationship (Monson and Fredman 2012). For this population, exploration of parental developmental expectations and knowledge of early child development may also be useful for families who have experienced lengthy parent-child separation.

Limitations and Future Research Directions

This research has several limitations. First, because the study is cross-sectional, causality cannot be inferred. It is likely that parent and child distress and behavior have reciprocal influence within the family system. Longitudinal work will bring to light how the relationships between parental distress, couple functioning, and child well-being change and interact over time and in the specific context of military deployment and service. Second, the community sample was comprised of families who were willing to participate in an intervention development project. This self-selection may reflect altruism, resilience, and competence among military parents who may have been more likely to seek parenting support. Future research can address this limitation with multiple methods of child and family assessment, including observational measures to observe child behavior, parent-child and family dynamics, and teacher report for older preschooler children.

Consistent with a growing body of research on post-9/11 families, our analysis of cross-sectional data found that higher rates of reported child behavior problems in military families are related to parental trauma, parenting-related stress, and difficulties in parent-child interactions, and are particularly significant when the effects of combat-related trauma are examined for the military parent. These findings may point to the helpfulness of early identification and intervention to address parental traumatic stress, including avoidance and numbing, even at the subthreshold level, to support the reinstatement of nurturing and supportive parent-child relationships between service members and their young children. The significant but more moderate relationships found for home-front parents may be explained by their relatively lower level of expressed distress.

Parental perceptions of child behavior problems in military families during post-deployment are of critical concern because they may set the stage for long-term parent-child relationship challenges and later child difficulties (Petty et al. 2008). The finding that most families in this community sample did not identify behavior problems in their young children is encouraging. However, parents who reported greater distress, not surprisingly, also revealed more concerns about their young child’s behavior. These findings warrant further investigation with representative samples and may suggest the need for early interventions with military family members experiencing deployment-related and family-based life stresses.