This chapter investigates the physical, mental and family health of the partners of ADF members who deployed to Timor-Leste. No statistically significant differences were found between those partners and comparison partners on measures of physical health, mental distress, Posttraumatic Stress Disorder symptoms, mental health status, drinking, smoking, family functioning, relationship satisfaction, work–family conflict or partner abuse. All analyses were adjusted for partners’ age (18–29, 30–39, 40–49 and 50+ years), sex and education level and ADF members’ rank and Service (Navy, Army or Air Force).
Overall, the majority of partners scored within the healthy range on almost all measures. The 2007–2008 National Health Survey (Australian Bureau of Statistics 2009) provides some information on self-reported health, smoking and psychological distress. The findings for females aged between 25 and 44 years were similar to those reported here. The data have not, however, been matched for age and sex or statistically compared.
When answering questions about work–family conflict, partners reported that their ADF member’s work created conflict in some aspects of their family life but not in others, highlighting the complex effects military life has on families.
Measuring partner abuse is difficult because such abuse tends to be under reported, the questions can be confronting for the participant, and domestic abuse can take many forms. Choosing the most suitable and sensitive way of measuring abuse was therefore difficult.
The screening tool used for this study—the Woman Abuse Screening Tool—was chosen because it was short, participants had reported being ‘comfortable’ or ‘very comfortable’ when the WAST was administered in other research (Brown et al. 2000), it measured multiple facets of intimate partner violence (physical, emotional and sexual), it was less intrusive than other measures that ask for more details about the kinds of abuse, and it has been shown to be scientifically reliable.
Participants were classified as screening positively for intimate partner violence without having to endorse any specific items about abuse. They had only to agree that there was a lot of tension or that arguments were resolved with great difficulty. These two questions have been shown to correctly identify more than 90 per cent of abused women (Brown et al. 2000). The WAST does not, however, reveal the duration or frequency of any abuse or whether the individual had experienced abuse in earlier relationships. As a result, no further information is available about the proportion of partners of ADF members who were experiencing abuse at the time or had sought help for this problem, or whether any children in the relationship were also suffering, or for how long they had been in this situation. Nevertheless, the WAST provided a baseline measure of how many partners might have experienced IPV in their current relationship. About 10 per cent of partners screened positively. Further exploration to better understand IPV in the military community is warranted, particularly in relation to how the military community compares with the civilian community in Australia.
It is perhaps puzzling that no statistically significant differences were found between the partner groups. Some measures, such as the PCL-C and the AUDIT, assess outcomes that are comparatively rare in the community: one would not expect large percentages of the population to have Posttraumatic Stress Disorder or to be high-risk drinkers. The confidence intervals associated with the statistics reflect this: they are very wide and include the possibilities of the Timor-Leste partners doing better or worse than the comparison partners.
There are other plausible reasons for a finding of no statistically significant difference between the groups, the most obvious being that there is no difference. In contrast to earlier conflicts, such as World War I or the Vietnam War, contemporary ADF members are likely to have been on more than one operational deployment. Both comparison partners and Timor-Leste partners might have experienced their ADF member’s deployment to another location. Extrapolating from Table 3.8, 46.7 per cent of all partners had been in a relationship with their ADF partner for two or more deployments. Consequently, as ADF members experience more deployments, each of them different, isolating the specific impact of an individual deployment becomes more difficult. It is possible that the absence of the serving member, rather than the location of their deployment, has the greatest impact (Andres & Moelker 2011). Further, in the current operational environment it is not clear whether those who have not deployed are, from an epidemiological point of view, equivalent to those who have deployed. They might have different training, skills or duty requirements that require them to remain in Australia, or there might be some other health or family reason that makes them systematically different from those who have deployed. It is thus difficult to isolate a particular deployment experience from any other deployment or non-deployment experiences.
Overall, the sample was reasonably healthy. The literature generally finds that military families constitute a robust and resilient population (Cozza et al. 2005). Alternatively, it is possible that partners who were experiencing health problems did not participate in the study, with the result that healthy, well-functioning partners are over-represented.
For some partners up to 12 years have passed since Timor-Leste deployment and any unique impacts might have since dissipated. More than a quarter of the partners of Timor-Leste veterans were not with their ADF member at the time of that deployment. The East Timor Health Study (McGuire et al. 2009b) found that East Timor veterans who deployed between September 1999 and January 2000 had a higher mean number of symptoms on the PTSD Checklist – Civilian Version than did ADF members who deployed later; although the mean was higher it was not above 50, indicating a positive screen. It is reasonable to infer that if there was any secondary traumatisation of the veteran’s partner, this event is comparatively rare and consequently not apparent in statistical analyses. Prospective research designs are better able to answer questions of specific effects at specific times.
More than 75 per cent of all partner participants completed more than 90 per cent of the questionnaire, the most frequent missing answers being those for free text fields such as ‘Please list below any benefits that you gained from your partner’s deployment’. In contrast, fewer than 10 per cent of partner participants completed less than 20 per cent of the questionnaire. All participants who responded were included in the analysis and as a result there were missing data on most measures. This could relate to partner health.
Lead statements to questionnaire scales that participants might have found distressing (for example, questions about abuse) included the statement ‘If you would prefer not to answer any of these questions, please leave them blank’. It is realistic to assume that this advice was taken by some participants.
Most of the measures reported require the calculation of a final score from a set of questions. For example, in order to calculate an individual’s consumption of alcohol category (AUDIT), answers to 10 questions were needed. If the participant missed an item, calculating their score was not possible and they were reported as having missing data. Where statistical techniques for replacing missing data were available, they were used. For example, the Kessler-10 measure reports outcomes in categories. If a participant missed one question, and assuming any response to that question did not change which category they belonged to, that individual’s category outcome was included.
Longer measures have an increased likelihood of having data missing from them. The FACES-IV (Family Adaptability and Cohesion Evaluation Scale) is the longest scale in the questionnaire and family functioning scores were not able to be calculated for more than 400 partners. This is unlikely to represent a difference between Timor-Leste and comparison partners because the amount of missing data was about the same between the two groups. Nonetheless, it is possible that partners in greater distress did not complete this measure, meaning they are not appropriately represented by the data.
Chapter 5 also deals with research aim 1 but focuses on whether there are any differences in outcomes for the children of ADF members who deployed to Timor-Leste compared with children of ADF members who had not deployed to Timor-Leste. Chapter 6 begins the analysis of risk and protective factors associated with health outcomes.