Abstract
Individuals with posttraumatic stress disorder (PTSD) suffer distressing and disabling symptoms. Clinicians working with Vietnam veterans suffering PTSD have reported high levels of tension, conflict and aggression within their families (see below).
Most of these reports, however, are based on clinical impression and opinions rather than controlled research and are largely based on US veterans.
Clinicians in the United States have noted that partners of Vietnam veterans with PTSD had problems coping with the veterans' hypersensitivity, verbal abuse and outbursts of anger [1,2]. Sexual dysfunction or sexual disinterest among the veterans was common [1], and the marriages suffered a lack of intimacy [3]. Partners reported living in an atmosphere of fear, social isolation and self-doubt; they experienced feelings of loneliness and isolation in the face of reduced communication, emotional withdrawal and psychological numbing of the veteran. These factors also placed various social restrictions on the lives of family members [1,3].
Some reports noted that avoidance behaviours (for instance, those listed in Criterion C of the DSM-IV diagnosis) led Vietnam veterans with PTSD to withdraw from society. Families of these veterans tended to adjust their lifestyles around this behaviour: for instance, leaving the partner with no social contact other than the veteran [1,3]. The picture is not entirely clear however, as one major study found no differences in the degree of social isolation between Vietnam veterans with PTSD and a comparison veteran sample without PTSD [2].
The literature suggests that partners were more likely to be anxious, suffer from low self-esteem, feel confused and overwhelmed, suffer from loss of identity and self-doubts and feel out of control of their lives [1,3,4]. Verbosky [3] found a significant relationship between severity of the Vietnam veteran's PTSD symptoms and the partner's low self-esteem, limited coping skills and ineffective use of various strategies to deal with problems.
Some partners had come to believe that they were responsible for ‘healing’ the veteran and suffered feelings of guilt and helplessness for not being able to [1,3]. In some cases, partners identified with the symptoms of the veterans with PTSD. Thus, they themselves, suffered hostility, social and emotional withdrawal, tearfulness, lack of concentration and stress related symptoms such as headaches, sleep disturbance and sexual dysfunction [5,6].
Long-term family effects of combat trauma have also been found among families of Israeli veterans of the 1982 Lebanon war. There were increased feelings of loneliness through impaired marital and family relationships and a lack of satisfaction with the wider social network [7]. Families of veterans with PTSD experienced more problems in parenting as well as marital relationships [2]. For instance, increased levels of violence within the family were perpetrated not only by veterans [1,5], but also their partners [2].
Reflecting marital and parenting difficulties, children showed symptoms in association with parental PTSD [1,8,9]. Examples of such symptoms include having low self-esteem, being aggressive, having impaired social relationships, developmental difficulties at school and ambivalent feelings towards their mothers [1]. Rosenheck and Nathan [8], who treated children of Vietnam veterans with PTSD noted that the children displayed symptoms similar to children of Holocaust survivors. The symptoms included identification with and intense involvement in the emotional life of the father, guilt, anxiety and aggressiveness. The effects of having a parent with PTSD varied from family to family, but common responses to father's irritability, aggressiveness, depression and withdrawal were secondary traumatisation with depression, guilt, anxiety and irritability [8,10], or becoming ‘rescuers’ of their parents [9]. Second generation effects included an increased susceptibility to the development of PTSD in the children themselves [11,12]. This is similar to findings from families of survivors of the Holocaust [13].
Among the few Australian papers, a review of veteran's services found that Vietnam veterans and their partners believed their children had suffered because of the veteran's emotional and adjustment problems [14]. In a recent book, Raftery and Schubert [15] reported families of World War II veterans had to deal with multiple problems including chronic illness, drinking problems, social withdrawal, anxiety, mood swings, negative war-focused behaviour and lack of empathy. In some of these families, symptoms consistent with secondary traumatisation of partners and children occurred.
This is comparable to families of Holocaust survivors, where the home atmosphere was often characterised by pervasive depression, worry, symbiotic clinging, outbursts of rage, mistrust and fear of the outside world [16].
Overseas studies give clues to what may be expected among families of Australian veterans with PTSD. However, to date no quantitative research has looked at these Australian Vietnam veterans' families. This study was designed to assess the personal adjustment and family relationships of partners and children of Australian Vietnam veterans with PTSD.
Method
A random sample of Vietnam veterans with PTSD who had contact with St John of God Hospital, North Richmond, New South Wales, were asked to give permission for members of their families (partners and children over the age of 15) to be surveyed. The veteran's PTSD had been diagnosed through clinical interviews by the principal psychiatrist and the administration of the Clinician Administered PTSD Scale (CAPS) [17] by a clinical psychologist. Only those who had current relationships were eligible for inclusion in the sample.
Control comparison groups were sought by advertising for volunteers from university or hospital staff. Those eligible were either (i) aged between 40 and 60 years (the same age range as the veteran group) with a child over the age of 15; or (ii) themselves over the age of 15 and with parents between the ages of 40 and 60. This meant that the control group would be similar in family pattern to the veterans' families.
Packages of tests were sent by mail with a covering explanatory letter and included a return free post envelope. Some errors were found in returned forms (e.g. a few packages contained omissions such as two pages turned over at once). These were not included in the analysis.
All those contacted were informed that the purpose of the research was to investigate the impact of PTSD on families, were reassured about confidentiality, and offered further information if required. Written consent was obtained from all those participating.
Measures
Four self-report psychometric inventories were administered: the General Health Questionnaire (GHQ 28) [18]; the Family Environment Scale [19]; the Coopersmith Inventory–Adult Form [20]; and the Lifestyle Questionnaire, questions on health and lifestyle practices.
The Lifestyle Questionnaire was developed by the authors to provide information about gender, family status, health and wellbeing, prior traumatic experiences, past contact with counsellors, general health and fitness practices such as exercise, use of cigarettes and alcohol.
The GHQ 28 is an instrument designed to assess the rate of psychological distress in a population. It concerns itself with two major classes of phenomena: inability to continue to carry out normal ‘healthy’ functions, and the appearance of new phenomena of a distressing nature. It examines symptoms present for the few weeks prior to testing and compares current function with prior levels of wellness in the areas of somatic symptoms, anxiety and insomnia, social dysfunction, severe depression.
The Family Environment Scale [19] provides a measure of the social climate within families and has been recommended for use when studying effects of trauma [21]. It has 10 subscales that fit along three major dimensions. They are cohesion, expressiveness, conflict in the relationship dimension; independence, achievement orientation, intellectual–cultural orientation, active-recreational orientation, moral-religious emphasis in the personal growth dimension; and organisation, and control in the system maintenance dimension.
The Coopersmith Inventory [20] is a short selfreport inventory designed to measure self-esteem through attitudes toward the self in social, academic, family and personal areas of experience. High scores correspond to high self-esteem. In normal populations, the means have generally been reported to be within the range of from 70 to 80 with a standard deviation from 11 to 13 [20].
Results
Sample and response rate
There was a 61% return rate for veteran family members. The control group were volunteers who replied to advertisements or verbal requests. They came from a broad spectrum of hospital and university staff, ranging from domestic staff to nurses and lecturers.
The samples were similar in age but differed a little in gender. Ideally, the groups should have been controlled for socioeconomic status. This was not done. Most veteran families were receiving disability pensions and an estimate of the veteran's highest level of prior functioning was not obtained. Many partners and children gave their own current occupational status. A few indicated they had completed tertiary education or were senior managers, but approximately half the veteran's partners listed ‘home duties’ or ‘carer’ as their main occupation. It is probable that the control group, on average, had a higher socioeconomic status and this is a limitation when interpreting the results.
The adult groups were: partners (32 females; mean age = 48.1 years, SD = 5.66); and adult control group (five males, 10 females; mean age = 49.7 years, SD = 5.60). There was no significant difference in ages (t = 86, p = 0.40).
The child groups were very similar: children of veterans (six males, 16 females; mean age = 20.9 years, SD = 4.56); children in control group (four males, 10 females; mean age = 22.8 years, SD = 6.21). There was no significant differences in ages (t = 1.0, p = 0.33).
Results are reported separately for partners and children.
Test results for partners
On the GHQ 28 (Table 1), results showed that on all scales there was a significant difference between partners and controls (i.e. the partners of veterans with PTSD showed significantly higher levels of somatic symptoms, anxiety and insomnia, social dysfunction and depression). In the GHQ28, the cut-off point for serious distress is usually set at a score of four to five for the total score on the scale. Setting five as the cut-off point, 67.7% of partners were in this case range. This can be compared with just 13.4% of the adult control group.
General Health Questionnaire (GHQ): veterans' partners compared with control group
The results from the GHQ show that Veteran's partners were distressed and having difficulty coping. This was confirmed by their responses to the Family Environment Scale (Table 2). Veteran's partners rated their families as having less cohesion; a lower level of expressiveness; and a higher degree of conflict.
Family Environment Scale: veterans' partners compared with control group
The personal growth dimension and system maintenance dimensions did not show significant differences, although there was a difference in intellectual/cultural orientation in the personal growth dimension. This may have been due partly to the fact that the control comparison group probably had a higher level socioeconomic status on average. Moos and Moos [18] noted that people with higher education and occupational status were more likely to establish families orientated toward personal growth.
The personal emotional problems experienced and the difficult family environment may be expected to have an effect on self-esteem. This was confirmed by the results from the Coopersmith Inventory. There was a significantly lower level of self-esteem (t = −4.8, p = 0.000) in the veterans' partners (mean = 50.3, SD = 26.8) compared to controls (mean = 80, SD = 14.9). The partners' mean score is well below the usual normal range of between 70 and 80.
The Lifestyle Questionnaire gave some insight into lifetime experiences and coping strategies. Responses showed that 56% of veterans' partners experienced a life-threatening or dangerous situation, compared to 40% of controls. One-third of the partners reporting prior trauma said this was due to the veterans' violence, others cited life threatening events such as motor vehicle accidents or serious health problems.
Partners reported significantly more stress-related responses than controls (t = 3.33, p = 0.002) as measured by such things as regular nightmares, panic attacks, being easily startled, excessive sweating, poor concentration, irritability and reduced sexual desire.
Sleep problems and depressive feelings were assessed in the GHQ section B which measures anxiety and insomnia, where the results indicated that the veteran's partners had significantly more of these problems (see Table 1).
Their lifestyle practices in terms of smoking, drinking and exercise were not dissimilar to those of the control group, so these healthy lifestyle elements could not be regarded as being responsible for the differences in symptoms. The major differences were in the family relationships and the past trauma experiences.
Summary of partners results
The partners of Australian veterans had high levels of somatic symptoms (measured in GHQ section A and the Lifestyle Questionnaire). They suffered from anxiety, depression and insomnia (measured in GHQ section B, GHQ section D and the Lifestyle Questionnaire). They reported high levels of social dysfunction (measured in GHQ section C and Coopersmith Inventory); and that they lived in a family environment with poor relationships, low expressiveness and high conflict (Family Environment Scale).
Tests results for children
The results on the GHQ for veterans' children did not show the same degree of psychological distress or dysfunction as their mothers. That is, there was no significant mean difference between the two groups of children (veterans' children: mean = 3.50, SD = 4.39; control: mean = 2.50, SD = 4.70). Nevertheless, 36.4% of veterans' children scored above five on this scale, whereas only 14.3% of the control group scored that high. A score of 5 is generally used as the cut off level indicating distress [18].
On the Family Environment Scale, the veterans' families were rated by their children as being significantly higher in conflict (t = 3.53, p = 0.002), with a tendency to be lower in cohesion although this result did not reach statistical significance (t = −2.05, p = 0.05).
In the Coopersmith Inventory, there were no significant differences found between the mean scores of the two groups of children. The range of scores reflected differences at the extreme end of the scale between the two groups (veteran's children mean = 69.5, range = 16–96; control mean = 80.86, range = 48–100). Four veterans' children scored lower than the lowest person in the control group; and eight (almost 50%) were below the ‘normal’ score range of 70–80 compared with only four of the control group. Clearly, there were individuals identified as suffering from low self-esteem and this was more common among veterans' children.
Reports on healthy lifestyle practices showed there were no significant differences between the two groups.
Summary of children's results
The veterans' children lived in families with significantly higher levels of conflict. There was a small number of veterans' children who had very low self-esteem and some had high levels of distress as measured by the GHQ. There was a non-significant trend for more of them to report stress symptoms similar to those of their fathers (Lifestyle Questionnaire).
Conclusions
The sample size in this research is small and it should be taken as a pilot study indicating areas for further study. A larger sample, matched for age and background, would provide a better basis for assessing the impact of PTSD on veterans' families.
The survey was carried out by mail and this imposes limitations. Some possible problems, especially with the children, relate to whether the respondents adequately understood the test instructions; and whether they were prepared and could be honestly and accurately reflective in their responses. However, the GHQ, Coopersmith and Family Environment Scale have all been used over wide age ranges, including young adults [18,20]. A further limitation relates to the fact that the socioeconomic status was not controlled between the control and experimental groups. It is possible that differences noted here could be impacted by this fact. Future research should control for SES between groups.
Given these limitations, this study of Australian families of PTSD sufferers found results similar to those overseas. Like the wives of Vietnam veterans with PTSD studied by Matsakis [1], Jordan et al. [2] and Verbosky and Ryan [3], these partners of Australian Vietnam veterans with PTSD were distressed. They were anxious, depressed, had poor relationships within the family and saw themselves as being dysfunctional socially. Similarly, Solomon et al. [13] found that wives of Israeli veterans with PTSD or combat stress reactions reported lower family cohesion, reduced expressiveness and increased conflict, and they had higher rates of depression, anxiety and somatic complaints. In this study, the Australian women were struggling with low self-esteem in an environment of conflict and reduced intimacy. Their level of distress was increased because of the physical problems they report experiencing.
In summary, this study shows that partners of Australian Vietnam veterans with PTSD show a similar pattern of secondary traumatisation to the American and Israeli partners of veterans with PTSD [1,4,22]. They reported being stressed, experiencing frequent nightmares, insomnia, panic attacks, high startle response, reduced sexual desire, poor concentration and heavy sweating. They were anxious, irritable, depressed and socially dysfunctional. They had little support within the family as they saw their families as being high in conflict, with low cohesion and a reduced freedom of expression.
Whereas the veterans have been diagnosed with a specific disorder (PTSD) and offered help, the partners usually have not sought counselling. Sainsbury et al. [14] reported that partners complained about the difficulties they faced within the family and there were frequent requests for respite care or time out, varying from a few hours for themselves through to hospitalisation for the veteran. A common request was for more contact with the partner's psychiatrist, but this was to be better able to assist the veteran, rather than seeking help for themselves. A similar pattern was observed in World War II veterans' families when Raftery and Shubert examined strains on the marriages and found that almost 80% of the wives did not seek counselling. They commented: ‘It is clear that for up to 50 years many wives were involved in a significant struggle, with minimal outside intervention’ [15, p.89]. Likewise, Solomon [23] found that families of veterans with PTSD were reluctant to seek or accept psychological help.
Recognition of their suffering and the provision of information about the symptoms of PTSD may help partners understand and cope with their own and the veterans' problems. Partners could benefit from peer group meetings that would allow them to share their experiences with other women, with a view to improving social support, self-esteem and developing coping or problem solving skills. In cases where distress is extreme, individual or marital counselling could also be provided. Feedback from partners who attend Spouse Days at the St John of God PTSD program shows they appreciate the opportunity to meet and talk to other women with similar problems.
The veterans' children saw their families as being significantly higher in conflict and tending to be lower in cohesion. While average results for the children showed no significant differences, a small number of the veterans' children reported distressing symptoms on the Lifestyle questionnaire and extreme scores on the GHQ and self-esteem scale. These trends, while not as marked are similar to other studies showing similarity between the veteran fathers' symptoms and that of their children [12] and veterans' children being distressed [2].
There were some differences between the two groups of children in exercise programs and smoking behaviour. These should be investigated further. Also, this study did not specifically ask about physical violence within the family, and further studies should investigate this. Harkness [10] found that children of a violent father were more likely to have behaviour problems, and less social competence than children with a non-violent father.
As a group, the children were not as distressed as their mothers, though, there was a small number who had very low self-esteem and a few who reported high levels of distress. This finding is similar to Major [24], who found great variation in functioning among second generation Norwegian-born adolescent children of Jewish survivors of the Holocaust. They found there was increased depression and more school problems, but no evidence of overall serious psychopathology among the children they studied.
The Australian children of veterans reported their families as being high in conflict. Jordan et al. [2] found there was more violence in families of veterans with PTSD. The present study did not ask about violence within the family and future studies should.
The children were in their teens or early adulthood and had been living in the family environment for many years, so it is too late for early intervention. However, counselling (individual or family therapy) seems advisable for those who are most distressed. This may prevent serious long-term reactions later in life. Of particular concern is the potential for the continuation of the cycle of abuse into their own families [22].
Group or family therapy could improve and reinforce good parenting strategies. In turn, this would provide a better home environment for the children. In many cases, given the long-standing problems that have been existing in some families, comprehensive family therapy would be required. For a few of the more disturbed children, individual therapy may also be required.
Clinicians [1,5] have described the multiple caretaking roles assumed by wives of veterans with PTSD. The wives often act as a buffer between the veteran and the rest of the world, for example by taking on responsibility for explaining the father's actions to the children and attempting to protect both husband and child from the effects of PTSD. Effective buffering may have been the reason so few of the Australian children showed signs of serious problems. This too should be the subject of future research, as well as an assessment of the emotional cost of that role to the mother.