The mental health of Farsi-Dari speaking asylum-seeking children and parents facing insecure residency in Australia

Summary

Background

This research examined the mental health of a cohort of asylum-seeking children, adolescents and their primary caregiver affected by insecure residency while living in the community, compared to refugees and immigrants.

Methods

The project investigated the prevalence of psychosocial problems among Iranian and Afghani asylum seeker, refugee and immigrant children and adolescents, and their caregivers who arrived in Australia from 2010. In total, n=196 children and adolescents aged 5–18 years, and their primary caregiver were asked about family visa status, country of origin, level of education, parent symptoms of posttraumatic stress disorder (Harvard Trauma Questionnaire) and child wellbeing (Strengths and Difficulties Questionnaire). An additional n=362 Farsi and Dari speaking children, recruited through the Building a New Life in Australia (BNLA) study, a national comparison sample of families with permanent refugee visas, were included.

Findings

Asylum seeker children and adolescents displayed significantly more psychosocial problems compared to those with full refugee protection and immigrant background within the current sample and when benchmarked against a national sample of Farsi-Dari speaking refugee children. Higher parental posttraumatic stress disorder symptoms was associated with poorer child and adolescent psychosocial functioning. This effect was more marked in families with insecure residency.

Interpretation

Insecure visa status is associated with higher rates of children's mental health problems and a stronger association with parental PTSD symptoms compared to children with secure residency. This raises important questions about Australia's restrictive immigration policies.

Funding

This project was supported by an Australian Rotary Health Research Fund / Mental Health of Young Australians Research Grant and by the Australian Research Council (DP160104378).

Research in context.

Evidence before this study

Children in refugee families in Australia have shown positive psychosocial adjustment in the early stages of resettlement. The Building a New Life in Australia Study used a representative sample of children and their families recently resettled. Findings showed that children reported similar levels of psychosocial problems on the Strengths and Difficulties questionnaire when compared to children in the general Australian population. Evidence from Australia and other high-income countries has shown that asylum seeker children living in immigration detention and unaccompanied minors report high levels of psychosocial problems.

Added value of this study

This is the first study to systematically examine the psychosocial adjustment of asylum seeker children living with their families in the community in a high-income resettlement country. We found that asylum seeker children (with inscure residency) reported significantly more psychosocial problems compared children with temporary protection or from immigrant and refugee families (with secure residency). In addition, the relationship between worse parental posttraumatic stress disorder and child psychosocial problems was more marked families with insecure residency.

Implications of all the available evidence

Children and families subjected to restrictive immigration policies in Australia are at risk of poor psychosocial adjustment. There is an urgent need to review these policies to protect the wellbeing of these families.

Alt-text: Unlabelled box

Introduction

Approximately half of the more than 80 million displaced people worldwide are aged 18 years and under.¹ In 2019 there were approximately 3.1 million adults and children awaiting a decision on their protection claims within high income countries.¹ Australia has a long history of refugee resettlement though its offshore humanitarian program. It also has established some of the most restrictive policies for processing of onshore asylum seekers, particularly those that have arrived without entry visas. All non-citizens who enter Australia without a valid visa, including child asylum seekers, are subject to mandatory detention.² Australia also established a program of issuing only temporary protection visas to asylum seekers who have had their refugee claims upheld but arrived in Australia without proper documentation. Refugees who hold temporary protection visas (TPVs) face considerably more living difficulties compared to those with permanent protection visas.³ While awaiting a TPV, asylum seekers often hold bridging visas, which are even more short term and restrictive. Insecure residency status can have long-term consequences for mental health outcomes⁴ with length of insecure residence associated with severe mental health problems.⁵ Insecure residency is associated with higher postmigration living difficulties compared to refugees with permanent visas.⁶ For example, insecure residence can restrict an individual's employment opportunities, which affect families’ access to stable, affordable housing, and children's social integration⁷ as well as future employment prospects.⁸ A systematic review of European research found that improvements in health following conversion to secure residency were mediated by reductions in postmigration living difficulties.⁵ There is considerable complexity and variation in the conditions of various temporary protection visas internationally. The current study aimed to examine associations of parent and child mental health with insecure visas and immediate threat of removal (those still awaiting processing of their claim or who have had their applications rejected and are awaiting appeal),to those with TPVs, permanent protection and immigrants in a Australia in 2019.

Between 2008 and 2013 some 50,000 Irregular Maritime Arrivals (IMAs) arrived through people smuggling networks, the majority of whom sought political asylum in Australia. The then labour government processed 20,000 of these people. It stopped processing IMAs in August 2012 resulting in a cohort of 30,500 known as the Legacy Caseload, whose claims were frozen for up to two years and then were able to apply for TPVs. These included Temporary Protection Visas (TPV, 3 years) or Save Haven Enterprise Visas (SHEV, 5 years). One month prior to commencing data collection for the current study in January 2019, 14,603 were granted temporary protection (TPV or SHEV) while others remained with insecure status (6177 had their applications for protection refused (with some appealing) and 10,268 still awaiting the outcomes of their claims).⁹

Recent research in Australia has shown that newly arrived refugee children and adolescents with secure residency are experiencing positive mental and social well-being.¹⁰ However, insecure residency may provide stressors that impact children and adolescents in a way not evident amongst those families with access to permanent protection.¹¹ There is little information on the possible impact of restrictive immigration policies on children and adolescents. A 2019 meta-analysis in low and middle income countries found that PTSD and ADHD are higher among asylum seeker children and adolescents compared to refugees, while depression and anxiety diagnoses showed the reverse pattern.¹² Research in high income countries has tended to focus on unaccompanied minors, who show very high level of distress13, 14, 15 but it is not clear if similar difficulties will be evident amongst those living with their families. In particular, previous research has not examined the adjustment of asylum seeker children and adolescents living with their families in the community in a high-income resettlement country.

The current research focuses on Farsi and Dari asylum seekers who were members of the legacy caseload, more than 30,500 asylum seekers who arrived by boat in Australia (2009–2014) who had been living in the community with limited access to services awaiting the outcome of their protection application.⁹ We aimed to understand the prevalence of child and adolescent mental health problems, and how this is related to resettlement stressors by comparing the wellbeing of asylum seeker children and adolescents (awaiting processing of their applications) to those with temporary protection for 3–5 years (TPV or SHEV) as well as permanent refugees and immigrants. We aimed to compare outcomes across these residency categories and to examine the relationship between parent and child mental health in a representative sample of parents and children and adolescents in Australia.

We hypothesised that:

• 1.Child and adolescent asylum seekers would display a higher prevalence of psychosocial problems as assessed by the Strengths and Difficulties Questionnaire (total, emotional, peer and conduct problems, hyperactivity) and lower reporting of prosocial behaviour than those on temporary protection visas, permanent protection visas and immigrants.
• 2.Children and adolescents from families exposed to greater parental distress will have higher total problem SDQ scores and this relationship will be more marked for those families with insecure residency.

Methods

Participants

The Reassure Child and Adolescent Project (CAP) study

The Reassure CAP study is a 24-month longitudinal cohort study of children and adolescents (aged 5–17) and their primary caregiver from Farsi-Dari speaking backgrounds, arrived in Australia since 2010. We focus on findings from the first wave of assessment (January 2019 to August 2019). Families with children and adolescents (5–17 years) were identified from a larger prospective study, the Reassure study.

The reassure study

Recruitment for the adult Reassure study involved a representative multi-stage time by location sampling frame applied across 16 randomly selected ethno-specific grocery shops in Sydney from the total pool of known shops. Most newly arrived Iranian and Afghani community members visit these stores regularly to find traditional foods. This provided access to a wide range of the community and minimised recruitment bias. Customers were invited based on the time they entered the shop (randomised across the study using probability proportional to size). A Kish grid was used to randomly select a family member of the consenting person.¹⁶ This approach uses a pre-assigned table of numbers to determine which (adult) member of a household should be invited to participate. The final sample of 408 respondents (from 519 eligible customers, response rate = 79%) included 202 asylum seekers, 28 people on Temporary Protection Visas or Safe Haven Enterprise visas; 93 on permanent refugee visas; and 85 on immigrant visas. Eligibility criteria for the Reassure study were that participants spoke Farsi or Dari, identified Iran or Afghanistan as their country of birth and had arrived in Australia since 2010. Exclusion criteria: Individuals not meeting the eligibility criteria or those who were unable to provide informed consent. Participants were reimbursed $AUD20 for the initial interview. Ethics approval was obtained from the University of New South Wales Human Research Ethics Committee (HC16637).

Reassure CAP sampling

From the Reassure study, we invited n=44 families (n=66 children and adolescents), to the Reassure CAP study. Inclusion criteria for Reassure CAP were that adult participants satisfied the Reassure eligibility criteria and had children or adolescents living with them who were aged 5–18 years. Exclusion criteria were severe developmental and/or cognitive impairment of children and adolescents identified by the parent. This sample includes n=31 parents with n=48 children and adolescents aged (5–18 years) who consented to participate. The second stage of recruitment involved a linkage sample. Reassure CAP participants were asked to nominate up to two eligible families in their network to participate. Nominated families (n=135) were invited using an opt-in approach with the initial invitation made via the nominating Reassure CAP participant and follow up by bilingual research assistants. The family nominated primary caregiver was asked to complete the questionnaires for themselves and their children. In addition, adolescents (aged 11–18 years) were invited to complete self-report questionnaires regarding their own wellbeing. Across both sampling approaches, n=195 children and adolescents from n=135 families were recruited. See Figure 1. Ethics approval was obtained from the University of New South Wales Human Research Ethics Committee (HC180562). Data will be made available with a signed data access agreement through contact with the corresponding author.

Figure 1.

Flow of participants through the Reassure CAP study.

The Building a New Life in Australia (BNLA) sample

The Building a New Life in Australia study (BNLA) is a representative sample of 2399 humanitarian migrating units.¹⁷ The five waves of data collection of the BNLA study were conducted by the Australian Government Department of Social Services and the Australian Institute of Family Studies. Participants were adults with permanent refugee visas obtained offshore (78%) or onshore (22%) received May–December 2013,recruited 3-6 months after being granted a permanent visa across 11 sites nationally.

In the third wave of the BNLA study 694 children and adolescents of the 426 primary caregivers were included in the study (October 2015–February 2016), up to two children and adolescents per family¹⁰ and answered questions about child and adolescent physical health and activity, school absenteeism and achievement, family structure and parenting style, and community and neighbourhood environment.¹⁰ We extracted data for variables which could be directly compared to the Reassure CAP for n=27 Iranian and n=154 Afghani families from the BNLA sample as a benchmark to compare to the Reassure CAP sample of permanent refugees. The BNLA dataset is publicly available and accessible by authorised researchers who have obtained permission from the Department of Social Services. Ethics approval to analyse the wave three data was obtained from the Australian Institute of Family Studies Human Research Ethics Committee (LSHM 13/03).

Measures

Where available, previously validated translations in Farsi and Dari were used¹⁸ otherwise the World Health Organisation guidelines for translation of measures was used for individual items, including pilot testing.

Parent items

Reassure CAP parent items

The nominated primary caregiver completed demographic (visa status, country of origin, age, gender, marital status, whether they were separated from their spouse, the highest level of education completed and period of residence in Australia). Symptoms of posttraumatic stress disorder (PTSD) were measured using the Harvard Trauma Questionnaire (HTQ)^19,20 has been used extensively in international refugee populations.^21,22 The HTQ includes 16 symptom items based on DSM-IV PTSD criteria with response options ranging from: 1 (not at all) to 4 (extremely).

BNLA parent items

The BNLA study used the PTSD-8 to provide a brief measure of PTS symptoms which has been derived from the HTQ. Demographic items which matched those collected in Reassure CAP included parent gender, country of origin and age.

To compare adult reporting of PTSD symptoms between parents in the Reassure Cap and BNLA studies, the 8 items from the HTQ which matched those in the PTSD-8 were used. A symptom severity score was calculated as an average score across all items. The PTSD-8 asked “People sometimes have bad reactions after experiencing hurtful or terrifying events in their lives. Please think about each reaction listed below and decide whether, and how much, it bothered you in the past week.” and referred to “events” in each item, The HTQ asked “The following are symptoms that people sometimes have after experiencing hurtful or terrifying events in their lives. Please read each one carefully and decide how much the symptoms bothered you in the past week” and referred to “hurtful or terrifying events” in the individual items. The items included 1. Intrusive recollection (B1), 2. Event recurring (B2), 3. Recurrent dreams (B3), 4. Psychological and physiological distress (B4/5), 5. Efforts to avoid activities (C1), 6. Efforts to avoid thoughts (C2), 7. Exaggerated startle response (D4), 8. Hypervigilance (D5) (Hansen, 2010).

Child items

Child items included in this analysis were identical between the Reassure CAP and BNLA studies. Child wellbeing was measured using the Strengths and Difficulties Questionnaire (SDQ). The SDQ is a 25-item scale that has been used extensively^23,24 to measure 5 subscales of conduct problems, hyperactivity-inattention, emotional symptoms, peer problems and pro-social behaviours. Higher scores indicate greater problems, except for prosocial behaviours. In this study, as well as examining all subscales, the first four subscales were summed to create a total problem subscale (range 0–40) to match that used in the BNLA study. Parent report was collected for all children. For Reassure CAP adolescents, where available, parent report was replaced with self-report for adolescents.

Statistics

We applied Mixed Models Analysis, clustered by family unit to account for non-independence of association within family units and controlled for family country of origin, child gender, child age and whether adolescents self-reported. Demographic variables were compared across groups using chi square analyses for categorical outcomes and one-way ANOVA for continuous outcomes. First, a four-group visa variable (Asylum, TPV/SHEV, permanent refugee, immigrant) was used to compare among Reassure CAP participants. Second, for variables where BNLA data was available, these groups were again compared to the BNLA group (who were all on permanent refugee visas). To examine differences among visa groups in child SDQ scores, a five-group variable compared each of the Reassure Cap groups as well as BNLA participants. Two sets of pairwise comparisons were conducted. Firstly, each Reassure CAP migration group (SHEV/TPV, Permanent Refugee, Immigrant) was compared to the group on asylum visas. Second, each Reassure CAP group was compared to the BNLA refugee cohort. Bonferroni correction was used to control for type 1 error. When comparing within the Reassure CAP sample, we corrected for three comparisons with asylum seekers as the reference group (p<0.017). When comparing BNLA to the Reassure CAP groups there were four comparisons with BNLA as the reference group (p <0.0125). To examine the association between parent PTSD symptoms and child SDQ problems subscale, a Mixed Models Analysis, clustered by family was conducted, controlling for child age and gender, country of origin, whether adolescents self-reported or were from the BNLA sample. The model further tested whether there was an interaction between parent PTSD symptoms and whether the family was on insecure (asylum) or secure (all others) visas.

Role of the funding source

The funders played no role in the design or analysis of this study.

Results

Demographics

Family/Parent demographics

Demographic data is described in Table 1. In the Reassure sample, there were significantly more families from Iran (n=112) than Afghanistan (n=20). The reverse was true for the BNLA sample (Afghanistan, n=154; Iran, n=27). Most primary caregivers were married for those with secure (74%) and insecure (89%) visa status, as well as for those in the BNLA Wave 3 (74%). There were no significant differences between parent gender or age and immigration status.

Table 1.

Sociodemographic characteristics.

		Asylum		TPV/SHEV		Refugee		Immigrant		χ2 Reassure	BNLA		χ2 BNLA vs Reassure
Primary Care Givers		n = 47		n = 20		n = 27		n = 41			n = 181
		n	%	n	%	n	%	n	%	p	n	%	p
Gender	Male	15	32	7	35	8	30	13	32	0.985	41	23	0.336
	Female	32	68	13	65	19	70	28	68		140	77
Country of Origin	Iran	44	98	20	100	12	44	36	90	<0.001	27	15	<0.001
	Afghanistan	1	2	0	0	15	56	4	10		154	85
Finished High School		37	82	17	90	13	57	34	90	0.019
Completed Tertiary Study		16	36	9	53	4	22	27	75	<0.001
Marital status	Single	2	4	0	0	1	4	0	0	0.476
	Married/ de facto	35	74	14	74	22	85	38	93
Divorced / widowed / separated		10	21	5	26	3	11	3	7
Geographically separated from partner*		6	13	2	11	2	7	3	7	0.315
		M	SD	M	SD	M	SD	M	SD	p	M	SD	p
Number of Children in Family		2	0.8	1.67	0.7	2.77	1.7	2.1	1.4	0.019
Age of parent		40	8.1	40	7.2	39.1	9	38	6.4	0.498	38.7	8.6	0.391

Children		n = 64		n = 30		n = 44		n = 57			n= 296
		n	%	n	%	n	%	n	%	p	n	%	p
Gender	Male	36	56	16	53	26	59	26	46	0.538	64	22	<0.001
	Female	28	44	14	47	18	41	31	54		232	78
Self-report	Parent report	40	64	21	70	40	91	53	93	<0.001	296	100	<0.001
	child-report	23	36	9	30	4	9	4	7		0	0
		M	SE	M	SE	M	SE	M	SE	p	M	SE	p
Age		12.4	0.52	11.9	0.75	10.9	0.63	9.2	0.55	<0.001	12.4	0.48	<0.001

Note: Refers to a person currently in a relationship (married or de facto) who is separated from their partner due to living in a different country.

Child demographics

On average, BNLA children and adolescents were significantly older than Reassure CAP children and adolescents (CAP M = 11.95% CI 10.45–11.61; BNLA M = 12.25 95% CI 11.85, 12.65). Amongst Reassure CAP children the oldest group were the asylum seekers which were comparable to the BNLA sample. There were no significant differences between Reassure CAP immigration groups in gender, but there were significantly more females in the BNLA study.

Strengths and difficulties questionnaire

Table 2 shows the SDQ total problems and subscales. All analyses controlled for family country of origin, child gender, child age and whether adolescents self-reported. Results are broken down by gender to allow comparison to the previously published BNLA study.¹⁰

Table 2.

SDQ scores Reassure Farsi-Dari child sample by Migration Stream compared to BNLA Farsi-Dari national sample.

	Reassure Cap Children							BNLA
	Asylum	TPV/SHEV	Refugee	Immigrant	Asylum vs TPV/SHEV	Asylum vs Refugee	Asylum vs immigrant	BNLA	Asylum vs BNLA	TPV / SHEV vs BNLA	Refugee vs BLNA	Immigrant vs BNLA
BOYS	n = 36	n = 16	n = 26	n = 26	p	p	p	n = 162	p	p	p	p
	M (SE)	M (SE)	M (SE)	M (SE)				M (SE)
Emotional Problems	7.59 (.44)	4.37 (.59)	3.20 (.48)	2.68 (.48)	<0.001	<0.001	0.001	2.74 (.31)	<0.001	0.013	0.349	0.903
Conduct Problems	3.22 (.34)	3.00 (.49)	1.94 (.49)	3.11 (.43)	0.697	0.010	0.833	1.77 (.30)	0.002	0.032	0.690	0.007
Hyperactivity	6.09 (.39)	4.18 (.51)	4.14 (.42)	3.61 (.42)	<0.001	<0.001	<0.001	3.38 (.28)	<0.001	0.17	0.079	0.607
Peer Problems	5.03(.26)	3.85 (.39)	3.76 (.32)	3.49 (.32)	<0.001	<0.001	<0.001	2.92 (.22)	<0.001	0.036	0.011	0.091
Prosocial behaviour	8.07(.41)	7.40 (.58)	8.94 (.44)	8.17 (.46)	0.311	0.137	0.851	8.26 (.35)	0.722	0.205	0.177	0.867
Total Problems	22.3 (1.1)	15.1 (1.4)	13.0 (1.16)	12.1 (1.15)	<0.001	<0.001	<0.001	11.4 (.74)	<0.001	0.019	0.174	0.575

GIRLS	n = 28	n = 14	n = 18	n = 31	p	p	p	n = 116	p	p	p	p
Emotional Problems	7.66 (.44)	4.44 (.59)	3.27 (.50)	2.74 (.47)	<0.001	<0.001	<0.001	2.80(.31)	<0.001	0.013	0.349	0.903
Conduct Problems	3.67 (.36)	2.08 (.49)	1.69 (.45)	1.95 (.39)	<0.001	<0.001	<0.001	2.17 (.23)	0.001	0.865	0.316	0.603
Hyperactivity	6.05 (.39)	4.14 (.51)	4.10 (.44)	3.57 (.41)	<0.001	<0.001	<0.001	3.34 (.28)	<0.001	0.17	0.079	0.607
Peer Problems	5.17 (.30)	3.99 (.39)	3.90 (.33)	3.63 (.32)	<0.001	<0.001	<0.001	3.06 (.22)	<0.001	0.036	0.011	0.091
Prosocial behaviour	7.85 (.43)	9.56 (.57)	8.99 (.53)	8.74 (.42)	0.01	0.920	0.960	8.05 (.26)	0.684	0.018	0.093	0.143
Total Problems	22.1 (1.07)	14.9 (1.42)	12.8 (1.2)	11.9 (1.13)	<0.001	<0.001	<0.001	11.2 (.74)	<0.001	0.019	0.174	0.575

Note: Results are shown for (Asylum-TPV/SHEV-Refugee-Immigrants-BNLA) along with pairwise comparisons between Asylum and all other Reassure CAP groups, and BNLA and all Reassure CAP groups. Means are adjusted for covariates.

Total problems

Groups were significantly different (F=22.12, p <0.001) with scores highest for the asylum group, which were significantly higher than all other Reassure CAP groups and the BNLA refugee child sample (See Table 2). TPV/SHEV scores were significantly higher than BNLA. Self-reported scores were significantly higher, F = 6.632, p = 0.01; parent report (M = 13.2, SE = 0.5); self-report (M = 16.2 SE = 1.2).

Emotional problems

Asylum seekers reported significantly higher emotional problems compared to all Reassure CAP groups (F = 27.22, p <0.001). The asylum group was significantly higher than BNLA. Self-report was significantly higher, F = 5.13, p = 0.024; Parent report (M = 3.6, SE = 0.2); self-report (M = 4.7, SE = 0.5); and older children and adolescents displayed higher scores, B = 0.06, SE = 0.03, t = 1.9, p = 0.046.

Conduct problems

There was a main effect for group, F = 4.5, p = 0.001, and a significant interaction between gender and all groups, F = 3.3, p = 0.012. For boys, asylum seeker scores were significantly higher than refugees. Asylum and immigrants groups were significantly higher than BNLA. For girls, the asylum seeker group was higher than all Reassure CAP groups and BNLA. There was also a main effect for self-report, F = 6.14, p = 0.14; parent report (M = 1.9, SE = 1.6); self-report (M = 2.9, SE = 0.34), and an interaction between age and group, F = 2.4 p = 0.35, with all groups except Reassure sample immigrants showing a negative correlation between conduct scores and age.

Hyperactivity symptoms

The asylum seeker group was significantly higher than all other Reassure CAP groups and BNLA., F = 9.9, p <0.001. There was a significant effect for age, B = -0.7, SE = 0.03, t = -2.42, p = 0.016.

Peer problems

Asylum seekers were significantly higher than all other Reassure CAP groups, and BNLA was significantly lower than the asylum seeker and refugee groups, F = 9.1, p <0.001. There was a significant effect for self-report, F = 9.1, p = 0.001 Parent report (M = 3.4, SE=0.15); Self-report (M = 4.4, SE = 0.32).

Prosocial behaviour

On the Prosocial subscale there was no main effect for group, F= 1.5, p = 0.212. There was a significant effect for gender, F = 3.9, p = 0.047, an interaction between gender and group, F = 2.8, p = 0.027, and a significant effect for age, B = 0.05, SE = 0.02, t = 2.17, p =0.031.

Association between parent PTSD and child SDQ

After controlling for country of origin, child gender, age, self-report and whether they were from the BNLA sample, parent PTSD symptoms was significantly associated with child SDQ score, F=25.4, p<0.001. Higher parent PTSD symptoms were associated with higher child SDQ, B = 4.9, p < 0.001. There was a significant interaction between parent PTSD symptoms and family visa status, F = 5.31, p = 0.022. When compared to those with secure visa status, the association of parent PTSD symptoms with child SDQ was significantly more marked among families with insecure visa status, B = -3.08, p = 0.022.

Discussion

Our findings demonstrate a clear pattern of asylum-seeking children and adolescents displaying heightened symptom scores compared to language matched children with similar lengths of stay in Australia on temporary protection visas, refugees, and immigrants. This raises significant concerns about the impact of restrictive immigration policies on families. There were significant differences between asylum seekers and all other Reassure CAP visa groups and the refugee children and adolescents in the BNLA study for total problems and for all subscales except for prosocial behaviour. The validity of these findings is supported by the congruence between the Reassure CAP immigrant and refugee SDQ scores and the BNLA sample who had access to permanent residency. We further note that SDQ scores in the BNLA sample have been previously demonstrated to be largely consistent with Australian community norms,¹⁰ indicating that refugee children can show positive adjustment when provided with secure residency. In contrast, children with insecure residency in this study reported elevated psychosocial problems. The findings support a compounding situation for asylum families whereby parental PTSD is associated with child psychosocial problems. Across all residency categories, higher parental PTSD symptoms were associated with significantly higher child total problems. There was an interaction effect with residency status in the form of a stronger relationship between parent PTSD and child SDQ score among asylum seeker (with insecure residency) families compared to those with comparatively secure residency.

We focus on the challenges faced by asylum seekers living for long periods in the community with restricted access to rights and services.⁹ The hold on processing of protection claims, which resulted in exclusion from status resolution and associated benefits, created a context where even a temporary visa outcome (i.e. TPV and SHEV) became a preferable alternative. Given that our analyses controlled for gender, age and country of origin of the children and adolescents, a key feature which appears to distinguish asylum-seeker children and adolescents is that they live with ongoing visa insecurity. From this cross-sectional analysis, we cannot establish exactly why asylum seekers report more problems, only the association with visa status. Similar differences in SDQ scores have been found in other studies comparing host community European and parent-accompanied asylum-seeking children and adolescents living in asylum centres.⁴ Our findings strongly suggest that children and adolescents in asylum-seeking families may develop psychiatric symptoms in response to ongoing uncertainty. However, a prospective analysis is required to determine a causal link. In addition, families with insecure visas may have been exposed to greater adversity before arriving in Australia.²⁵

Conduct problems among girls

We note that female asylum-seeking children and adolescents displayed significantly more conduct problems than girls in the other Reassure CAP groups. Conduct problems have previously been associated with experience of traumatic events and other adversities²⁶ and may be shaped by cultural expectations surrounding appropriate social behaviour.²⁷ Previous research among unaccompanied children and adolescents has also shown that the level of conduct problems for girls was high.²⁶ Similarly, in the BNLA study,¹⁰ mean scores were generally similar between girls and boys on conduct scores, except for 14–17 girls. Of note is that the mean age in the asylum seeker group in Reassure CAP was higher than other groups, it may be that conduct problems are scored higher by parents of older girls. It is possible that older girls undergoing acculturation may come into conflict with their parents, and this may be understood as behaviour problems by the parents. We note that we did not test for differences between genders and only report within gender effects.

Parent PTSD symptoms and child SDQ

The findings regarding the relationship between parent PTSD and child SDQ in this study add to those reported from a prospective analysis of the BNLA study in which parent PTSD resulted in a greater reliance on harsh parenting that was directly associated with the poorer mental health in the children.²⁸ Although the existing evidence is consistent on the association between trauma exposure and a range of child mental health problems, more is needed to understand the mechanisms in relation to refugee parents.²⁹ A mixed method study among conflict-affected families in Timor-Leste suggested that parental explosive anger was related to women's traumatic experiences, and this had an impact on women's parenting.³⁰ The relationship between parenting style and child behaviour is likely bi-directional: recent investigation of coping among displaced Syrian refugee families revealed that problems increased in parenting style as children develop increasingly problematic behaviours.³¹

Limitations

We note that other factors (such as stressors in country of origin or during travel to Australia) may have impacted on parent and child mental health. We controlled for parent trauma events to account for this. The Reassure CAP sample drew on the representatively sampled Reassure study, supplemented with a linkage framework that could have introduced non-representativeness. The finding of similar psychosocial difficulties with age and language matched children provides reassurance that the linkage sample in the Reassure CAP study did not introduce sampling bias towards greater severity. Both surveys relied on interview delivered self-report, which may overestimate base rates of distress. There was a gap of four years between the BNLA and Reassure studies. There may have been cultural differences between Afghani and Iranian participants. To address this, we controlled for country of origin. However, we note that culture and ethnicity are far more complex phenomena than is captured by country of origin and so these factors may remain unaccounted for. Finally, it is important to note that the cross-sectional nature of the reported analyses does not allow for the sequence of the identified associations to be determined. In addition, the PTSD measures between the two studies were not completely identical.

Conclusion

This is the first study to examine the mental health of accompanied asylum-seeking children living in the community in a resettlement country. The findings demonstrate that children and adolescents living with insecure residency report higher psychological challenges than age and language matched refugees with refugee status resolution and immigrants. These results build on findings of previous studies of the association between stressors and unaccompanied and detainee asylum-seeking children's mental health.32, 33, 34 In addition, these symptoms are associated with their parents’ psychiatric symptoms, indicating that restrictive asylum policies are associated with greater mental health problems across the whole family. Asylum seeker children have access to only limited clinical services; these findings raise considerable concern that asylum seeker children have large unmet mental health needs that need to be addressed.

Contributors

RR Conceptualisation, data curation, project administration, writing-original draft

RW - Conceptualisation, formal analysis, supervision, writing - review and editing

JS data curation, project administration

DB writing - review and editing

DHP writing - review and editing

DS writing - review and editing

AN writing - review and editing

MOD writing - review and editing

RB writing - review and editing

AF writing - review and editing

ZS - conceptualisation, funding acquisition, supervision, writing - review and editing

Data sharing statement

Due to the nature of this research, participants of this study did not agree for their data to be shared publicly, so supporting data is not available.

Declaration of interests

AM serves as consultant to Australian Department of Veterans Affairs and South Australian Department of Health and Ageing; received honoraria from Servier Pharmaceuticals and Royal Australian and New Zealand College of Psychiatrists; No other authors have conflicts to disclose.