The Indigenous Queensland Family Cohort (I-QFC) study: a feasibility birth cohort study of Aboriginal and Torres Strait Islander families – cohort profile

Abstract

Purpose

The Developmental Origin of Health and Disease framework emphasises the intergenerational influence of both maternal and paternal factors on the perinatal and longer-term outcomes of offspring. Aboriginal and Torres Strait Islander communities have driven major advances in family and health research across Australia, including longitudinal studies grounded in community priorities. Despite this progress, few birth cohort studies have focused on Indigenous families and even fewer have examined the health and well-being of Indigenous fathers.

Participants

The Indigenous Queensland Family Cohort (I-QFC) study is a prospective, longitudinal birth cohort study that commenced in 2021, recruiting Aboriginal and/or Torres Strait Islander families from the Mater Mothers’ Hospital in Brisbane, Australia. This feasibility study forms part of the larger QFC study to understand the acceptability and practical considerations of including Indigenous families in a birth cohort study. Participants include pregnant individuals, their partners (where applicable), and their infants, followed from pregnancy to 6 weeks post partum. Data collection involved self-administered and researcher-facilitated surveys, alongside biological sample collection.

Findings to date

The study enrolled 103 pregnant participants and 64 partners and outlined the study design and baseline characteristics. Survey completion rates were high for antenatal and birth-related data collection when facilitated by Indigenous research staff (>94%). In contrast, completion declined substantially at later time points and among partners, with only 35 mothers completing the 6-week survey post partum and progressive participant attrition observed across the study period. Variation in biological sample provision and longitudinal follow-up provided important insights into cultural, logistical and relational factors influencing engagement in family-based cohort research. Biological sample collection was more limited, with few participants contributing across multiple time points. Participants were influenced by cultural considerations, including the spiritual significance of placental tissue, as well as perceptions among male partners’ belief that their involvement in perinatal research was unnecessary.

Future plans

By adopting a family-centred approach, I-QFC provides insights into culturally grounded research with Indigenous families, with the potential to strengthen health and well-being across generations. As such, these findings are currently used to inform the development of a forthcoming co-design birth cohort study of Aboriginal and Torres Strait Islander families in Queensland, Australia.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The Indigenous Queensland Family Cohort study is a feasibility study embedded within an established birth cohort, enabling assessment of recruitment, retention and data collection processes for Aboriginal and Torres Strait Islander families.

• Recruitment and data collection processes were adapted to promote culturally appropriate and respectful study participation.

• The study incorporated repeated questionnaires and biological sample collection across pregnancy and the early postnatal period for mothers, partners/fathers and infants, enabling assessment of the acceptability of intensive longitudinal study procedures.

• Variable survey completion, loss to follow-up and limited biological sample provision constrained longitudinal analyses and necessitate cautious interpretation of descriptive findings.

• The absence of early community consultation and Indigenous governance structures during study design may have affected the cultural acceptability and feasibility of some study components.

Background

It is now clear that there are multiple factors including parental health before conception and during pregnancy that are critical determinants of pregnancy outcomes as well as long-term health and well-being of offspring.^{1 2} Perinatal and early postnatal environments represent critical windows that can nurture healthy development and lay strong foundations for well-being throughout life. This perspective, known as the Developmental Origin of Health and Disease (DOHaD), postulates that early-life environmental factors influence genetic function and determine future health outcomes.³

Central to the DOHaD paradigm is intergenerational health, which recognises the enduring influence of behaviours, societal conditions and environmental exposures across generations and over the life course. These influences operate through interconnected biological, psychological and cultural pathways, including strong connections to land, family and community, which are central to health and well-being among Indigenous peoples.^{3 4} However, the ongoing impacts of colonisation and other forms of dispossession and trauma have deeply disrupted these connections and contributed to intergenerational inequities in health and social outcomes.^{5 6} For Aboriginal and Torres Strait Islander (hereafter respectfully Indigenous) communities in Australia, the ongoing effects of colonisation are compounded by systemic inequalities in healthcare, education and economic opportunity, which continue to shape health outcomes and access to well-being.^{7 8}

For Indigenous communities, birth cohort studies offer a powerful way to examine how biological, environmental and cultural determinants of health, including connection to Country and family, shape well-being. A Two-Eyed Seeing approach within DOHaD research underscores the value of Indigenous knowledge systems, which conceptualise health as a relational process deeply rooted in land, culture and community connections, offering a holistic lens that complements biomedical perspectives.⁹ However, in Australia, birth cohort studies focusing on Indigenous peoples remain limited. Among the few available, the Aboriginal Birth Cohort study in the Northern Territory (NT), established in 1987, remains the largest and longest running Indigenous birth cohort study that tracks the health and development of children through to adulthood.¹⁰ A comparable study, the Gudaga Study,¹¹ is an Aboriginal birth cohort study conducted in an urban community in Sydney, New South Wales (NSW) that followed mothers and their infants from pregnancy through to the first year of the child’s life. More recently, additional studies have emerged, including the Gomeroi gaaynngal study, a prospective longitudinal study that follows Indigenous mother–child dyads in rural and regional NSW from pregnancy through childhood,¹² and the Pregnancy and Neonatal Diabetes Outcomes in Remote Australia Study which focuses on Indigenous women with diabetes in pregnancy in the NT.¹³

Despite progress in Indigenous cohort studies, there remains a significant gap in birth cohort research in Australia, particularly studies that meaningfully include fathers within Indigenous families. This gap in research overlooks the influence of paternal health, lifestyle and socioeconomic factors on children’s well-being and is of crucial importance when assessing the influence of both parents on the intergenerational impact of health trajectories of children. While some birthing practices are traditionally regarded as ‘women’s business’,¹⁴ Aboriginal and Torres Strait Islander fathers increasingly see themselves as active and essential contributors to the pregnancy, birth and early parenting journey. Their involvement is shaped by cultural identity, emotional connection and a strong sense of responsibility toward family well-being.¹⁵ In particular, the role of culture, family and community before and after birth is vital for Indigenous children and mothers. These influences have been shown to improve resilience and overall health and well-being, particularly for Indigenous children.¹⁶ Supporting Indigenous fathers in their parenting roles not only strengthens family bonds but also enhances child development outcomes, reinforcing the need for a birth cohort study that adopts a family-centred approach.¹⁷ These principles provide an essential ethical and methodological context for Indigenous health research and inform ongoing efforts to improve the design of longitudinal studies involving Aboriginal and Torres Strait Islander families.

To address this gap, the Indigenous Queensland Family Cohort (I-QFC) study was initiated in 2021 as a feasibility study embedded within the QFC¹⁸ in Brisbane, Australia. The study aimed to assess the feasibility, acceptability and implementation considerations of a longitudinal, family-centred birth cohort, particularly one involving the collection of biological samples. Accordingly, this paper focuses on describing the study design, recruitment processes, participation pattern and learnings generated through this feasibility initiative.

The I-QFC study was designed to address several interrelated categories of research questions relevant to Indigenous family health across the perinatal period. These included (1) assessing the acceptability, cultural appropriateness and practical considerations of recruiting and retaining Indigenous families, including fathers, (2) examining maternal and paternal health and well-being during the perinatal period and (3) assessing some family, social and cultural determinants of health that may influence family well-being and early childhood outcomes.

Cohort description

Setting

The I-QFC is a prospective longitudinal study that followed pregnant individuals in Queensland, their partners (if they have one), and their babies from pregnancy to 6 weeks post partum. This feasibility study formed part of the broader QFC study (qldfamilycohort.org).¹⁸ I-QFC participants were recruited from the Mater Mothers’ Hospital in Brisbane, Australia, the state’s largest quaternary level perinatal centre, with ~10 000 births annually.¹⁹ At the study hospital, Aboriginal and Torres Strait Islander women account for approximately 3%–4% of all annual births. Study recruitment commenced in September 2021 and the study concluded in mid-2025.

Study recruitment and follow-up

Eligible participants were pregnant individuals and/or partners of babies identified as Aboriginal and/or Torres Strait Islander. Between 18 and 24 weeks of gestation, potential participants were identified by Indigenous researchers and non-identified research midwives through review of routinely collected hospital medical records to confirm pregnancy status and potential eligibility based on recorded Indigenous identification of the pregnant individual. Antenatal clinics and clinical staff also notified the research team of potentially eligible participants during routine antenatal care. Eligibility was confirmed through discussion and self-identification prior to consent. It was imperative that participants received culturally appropriate support from the outset and throughout their involvement in the study. Initial contact was primarily made via phone, during which a thorough discussion occurred where study aims and protocol were shared. Verbal consent was obtained after consideration, followed by subsequent written consent. Research information was collected at various time points (24, 28 and 36 weeks of gestation) and following birth and 6 weeks post partum. Participants received SMS and phone call reminders before each scheduled visit, and appointments were arranged for data collection.

Partner participation was optional. Where a pregnant participant reported having a partner and consented to their contact, the partner was invited to provide biological samples and health and well-being questionnaires at the 24-week gestation study visits. If this was not feasible, samples and questionnaires were collected at alternative time points or opportunistically when partners accompanied their pregnant partner to other visits. For each study visit, participants were provided with a A$20 grocery gift card and a $A20 parking voucher (if applicable). At the 6-week postpartum visit, participants were presented with a ‘Jarjum’ doll (Little Jarjums), as an appreciation for their involvement in the study (figure 1). ‘Jarjum’ is an Aboriginal word for children in the Bundjalung and Yugambeh languages, used on the Eastern Coast of Australia.²⁰

Figure 1. Jarjum doll received by participants during their 6-week postpartum study visit.

Data and biological sample collection

Health and well-being questionnaires and biological samples were structured around multiple health themes to enable meaningful analyses of health outcomes and experiences of participating families (figure 2).¹⁸ Biological sample collection in the I-QFC study was designed to align with the QFC protocol and the DOHaD framework. The data collection tools and instruments used in I-QFC were adopted directly from the published QFC study protocol, which provides comprehensive descriptions of all questionnaires and measures.¹⁸ Samples were collected to enable the investigation of biological, environmental, microbial and psychosocial mechanisms underpinning parental and early childhood health, with a particular focus on pathways underlying the development of chronic disease across the life course (summarised in online supplemental table S1). Repeated sampling across pregnancy, birth and the early postnatal period was undertaken to capture the expansive changes occurring during the critical developmental windows that may shape disease risk.

Figure 2. Overview of measures and outcomes used in I-QFC across multiple study time points. *Only collected from pregnant participants/mothers. Note: For a detailed and complete overview of the questionnaires and tools used, see the published QFC protocol by Borg et al.¹⁸ I-QFC, Indigenous Queensland Family Cohort.

Data were collected and recorded electronically, either by the Indigenous researcher/research midwife or self-reported by participants (figure 2). Those surveys completed by a midwife were referred to as ‘midwife-use’ surveys and focused on questions that could be accessed from the participants’ medical record, with minor additional questions directly to the participant if clarification was needed. This approach aimed to minimise repetition and reduce the burden for participants.

As a feasibility study embedded within the broader QFC study, I-QFC was not powered to detect associations or generate population estimates. Rather, the sample size was determined to support assessment of recruitment strategies, data completeness, retention across time points, and the acceptability of repeated questionnaires and biological sample collection among Indigenous families.

Patient and public involvement

Patients or the public were not involved in the design phase, reporting or dissemination of the study.

Statistical analyses

Descriptive statistics summarised baseline data including age, education, employment, marital status and Indigenous identity, among others. When data were missing, the research midwife retrieved it from the participant’s medical record, if available.

Overall participation

Results are presented descriptively, focusing on recruitment, participation patterns, data completeness and biological sample collection across study time points.

A total of 139 pregnant participants and 79 partners were initially contacted and recruited via phone to participate in the I-QFC study (figure 3). Of these, 105 pregnant participants (75.5%) and 64 partners (81.0%) were provided with online enrolment and consent forms. 34 pregnant participants (24.5%) and 15 partners (19.0%) declined to participate. In addition, eight pregnant participants (5.8%) withdrew after enrolment; two disallowed the use of their data, while six (4.3%) permitted further data linkage. These six pregnant participants and four partners who allowed continued data collection were included in the final analysis. Therefore, the final study cohort comprised 103 pregnant individuals (74.1% of those initially contacted) and 64 partners (81.0%).

Figure 3. Participation in the I-QFC study. I-QFC, Indigenous Queensland Family Cohort.

At the time of recruitment, the majority of pregnant participants were receiving antenatal care through Mater Mothers’ Hospital’s Midwifery Group Practice Birthing in Our Community programme, with 54 participants (52.4%) enrolled in this model of care. This is a culturally grounded, Indigenous-led maternity care model that provides continuity of midwifery care and holistic support for Aboriginal and Torres Strait Islander families.^{21 22}

The number of pregnant participants was higher at the start of the study but progressively declined as pregnancy progressed, with a significant drop observed at the 6-week postpartum visit (figure 3 and table 1). Completion rates varied across participant groups and survey types. Overall, midwife-administered surveys yielded more complete data than self-administered surveys, particularly at later gestational stages (eg,96.8% at 36-week gestation), enhancing data completeness. However, partner participation and completion rates were notably lower with each data collection.

Table 1. The types of I-QFC surveys at each study time point, either self-administered by the participants or conducted by the research midwife to participants and the corresponding rate of surveys with complete data.

Survey type	n	Survey with complete responses, n (%)
		Participant completed	Midwife completed
Admission form:
Pregnant participant	103	na	45 (43.7)
Partner	64	na	23 (35.9)
Baby	76	na	73 (96.1)
Enrolment form:
Pregnant participant	103	69 (67.0)	56 (54.4)
Partner	56	33 (58.9)	32 (57.1)
24-week survey:
Pregnant participant	101	46 (45.5)	95 (94.1)
Partner	30	11 (36.7)	17 (56.7)
28-week survey:
Pregnant participant	101	43 (42.6)	97 (96.0)
36-week survey:
Pregnant participant	94	36 (38.3)	91 (96.8)
During labour and at birth:
Woman	99	na	96 (97.0)
Baby	76	na	73 (96.1)
On ward survey:
Woman	99	52 (52.5)	74 (74.7)
Baby	74	na	58 (78.4)
6-week postpartum survey:
Woman	35	na	16 (45.7)
Baby*	37	23 (62.2)	na

^*Survey completed by the mother.

I-QFC, Indigenous Queensland Family Cohort; na, not applicable.

Demographic characteristics

Table 2 and online supplemental table S2 summarise participant characteristics at recruitment. Mean gestational age was 22.4 weeks (SD = 2.8). Pregnant participants were relatively young (mean age 27.9 years, SD = 6.6), while partners averaged 29.3 years (SD = 6.0). Most pregnant participants were married or in de facto relationships (85%). As only one parent was required to identify as Indigenous for study inclusion, 86.9% of pregnant participants identified as Indigenous. Of these, 76.2% identified as Aboriginal, 9.5% identified as both Aboriginal and Torres Strait Islander and only one pregnant participant (1.2%) identified solely as Torres Strait Islander. Only 35.7% were non-Indigenous. In contrast, fewer partners reported being Indigenous (table 2). Some participants identified their Indigenous Nation, while others were unsure. One attributed this uncertainty to colonisation’s impact on cultural identity and continuity.

Table 2. Summary of the demographic characteristics of participants in the I-QFC study.

Demographic characteristics	Mothers, n (%) n=103	Partners, n (%) n=64
Age at enrolment, years
Mean (SD)	27.9 (6.6)	29.3 (5.9)
Total	97	44
Missing	6	20
Indigenous identity, n (%)
Aboriginal	64 (76.2)	27 (44.3)
Torres Strait Islander	1 (1.2)	5 (8.2)
Aboriginal and Torres Strait Islander	8 (9.5)	1 (1.6)
None	30 (35.7)	28 (45.9)
Total	103	61
Missing	.	3
Gestation at the time of consent, weeks
Mean (SD)	22.3 (3.02)	.
Total	61	.
Missing	42	.
Marital status, n (%)
Single	10 (12.7)	.
Married	23 (29.1)	.
De facto (not legally married but in a committed long term relationship)	44 (55.7)	.
Other	2 (2.5)	.
Total	79	.
Missing	24	.
Country of birth, n (%)
Australia	66 (94.3)	36 (90.0)
New Zealand	2 (2.9)	1 (2.5)
Others	2 (2.9)	3 (7.5)
Total	70	40
Missing	33	24
Highest tertiary qualification completed, n (%)
Trade/vocational training/certificate/diploma	40 (69.0)	29 (82.9)
Bachelor’s Degree	15 (25.9)	5 (14.3)
Masters	3 (5.2)	.
PhD	.	1 (2.9)
Total	58	35
Missing	45	21
Currently employed/studying, n (%)*
Yes	54 (67.5)	42 (93.3)
No	26 (32.5)	3 (6.7)
Total	80	45
Missing	23	11
Total weekly household income (after tax) in AUD $ †, n (%)
<$A1500	29 (46.8)	.
$A1501–$A2499	19 (30.6)	.
>$A2500	14 (22.6)	.
Total	62	.
Missing	41	.
Is this the first pregnancy?
Yes	27 (36.0)	.
No	48 (64.0)	.
Total	75	.
Missing	28	.
Gravida (total number of pregnancies including current pregnancy)
1	27 (26.2)	.
2	24 (23.3)	.
3	16 (15.5)	.
4+	36 (35.0)	.
Total	103	.
Parity (total number of births >24 weeks)
0	38 (36.9)	.
1	28 (27.2)	.
2	14 (13.6)	.
3	8 (7.8)	.
4+	15 (14.6)	.
Total	103	.
Type of pregnancies in current pregnancy
Singleton	74 (98.7)	.
Twins	1 (1.3)	.
Total	75	.
Missing	28	.
BMI category at enrolment
<18.5 Underweight	7 (7.9)	.
18.5–24.9 Ideal weight	31 (34.8)	15 (45.5)
25.0–29.9 Overweight	19 (21.3)	11 (33.3)
>30.0 Obese	25 (28.1)	6 (18.2)
>40.00 Morbidly obese	7 (7.9)	1 (3.0)
Total	89	33
Missing	14	23
Diagnosed with at least one chronic disease at enrolment‡
Yes	88 (85.4)	46 (82.1)
No	15 (14.6)	10 (17.9)
Total	103	56

^*In the partner’s enrolment survey, this question was reworded to ask only, ‘Are you currently employed?’.

^†From enrolment survey.

^‡Participant reported to research midwife. Chronic diseases referred to asthma, autoimmune disease, blood coagulation defect, cancer, cystic fibrosis, gestational diabetes, type 1 diabetes, type 2 diabetes, endocrine disease, endometriosis, heart disease, hypertension/high blood pressure, low blood pressure, infectious disease, kidney/renal disease, mental health disorder, musculo-skeletal disease, neurological disease, phenylketonuria, respiratory disease (not asthma), skin disease/eczema, sleep disorder, thalassaemia, thyroid conditions, none, other, anaemia, polycyctic ovary syndrome (PCOS), rheumatic heart disease. Partner’s survey included all except gestational diabetes, endometriosis, anaemia, PCOS and rheumatic heart disease.

BMI, body mass index; I-QFC, Indigenous Queensland Family Cohort.

Over 90% of participants were Australian-born. Among pregnant participants, 31% held a bachelor’s degree or higher compared with 17.1% of partners. Employment was higher among partners (93.3%). Over half of pregnant participants reported weekly household incomes below $A1500, and 70.7% lived in rental housing (online supplemental table S2).

Health and behavioural characteristics

In terms of parity, 36.9% of participants were nulliparous (no previous births beyond 24 weeks), 27.2% were primiparous (one previous birth), and 14.6% had given birth four or more times (table 2). Additionally, 35% of participants were multigravida with four or more lifetime pregnancies (table 2). In the current pregnancy, there were 61.1% (n=58) spontaneous vaginal births, 6.3% (n=6) instrumental vaginal births, 17.9% (n=17) emergency caesarean births and 14.7% (n=14) elective caesarean births (online supplemental table S2).

A large proportion of families (34.8% of pregnant participants and 45.5% of their partners) were classified within the ideal healthy weight category (table 2). At the time of enrolment, most participants reported health-promoting behaviours, with the majority being non-smokers (71.4% of pregnant participants and 68.2% of partners) (online supplemental table S2). Alcohol consumption was also reported by just over half of the pregnant participants and by most partners (53.7% of pregnant participants and 76.7% of partners) (online supplemental table S2). At 3 months preconception, most pregnant individuals reported occasional alcohol use, with 39.5% drinking less than once a week and only 7% three or more times weekly. By weeks 1–12 of pregnancy, 55.3% reported no alcohol use, and by the most recent 2-week period, 94.6% reported abstinence. Among partners, alcohol use remained relatively frequent across all time points. The majority of participants reported no prior asthma diagnosis, with 37.8% of pregnant participants and 27.3% of partners indicating a history of asthma (online supplemental table S2). Gestational diabetes was uncommon, with only four pregnant participants reporting a diagnosis during their current pregnancy at enrolment (online supplemental table S2). Additionally, 84.4% of pregnant individuals and 82.1% of partners reported having at least one diagnosed chronic condition.

Neonatal outcomes

Neonatal outcomes are presented in table 3. The mean gestational age at birth was 38.4 weeks (SD=2.8). Preterm birth occurred in 10 (13.5%) of the births. 20% of neonates were admitted to the neonatal intensive care unit or special care nursery, the majority of whom were born preterm. The mean birth weight of babies born was 3273.7 grams (g) (SD=669.3), with most babies falling in the normal birth weight range (table 3). In addition, only 5.7% of all infants had birth weights below the 10th percentile according to the Fenton growth chart, indicating appropriate growth for their gestational age.

Table 3. An overview of the neonatal outcomes in the I-QFC study.

Variables	All babies		Females (n=30)		Males (n=45)
	Mean (SD) or %	n	Mean (SD) or %	n	Mean (SD) or %	n
Gestational age at delivery, mean (SD), weeks	38.4 (2.8)	76	38.1 (3.6)	30	38.5 (2.1)	45
Preterm birth (<37 weeks), %	13.5	10	10.0	3	16.3	7
Birth weight, mean (SD) (g)	3273.7 (669.3)	71	2956.3 (574.3)	29	3492.8 (647.4)	42
Low birth weight, <2500 (g), %	8.5	6	13.4	4	4.8	2
Normal birth weight, 2500–4499 (g), %	88.7	63	86.2	25	90.5	38
High birth weight, ≥4500 (g), %	2.8	2	0	0	4.8	2
Admitted to NICU/special care, %	20.0	14	19.2	5	20.5	9
Small for gestational age, %	5.7	4	7.4	2	4.7	2

I-QFC, Indigenous Queensland Family Cohort; NICU, neonatal intensive care unit.

Biological samples

Overall, biological samples that were easier to collect or minimally intrusive were more likely to be collected from participants, for example, cheek swab, skin swab and hair samples (table 4). This was similar for partners at the 24-week gestation visit (online supplemental table S3). Also, biological samples from newborns were rarely collected for samples such as heel prick, urine and cord blood (online supplemental table S3). At 6 weeks post partum, cheek swabs and hair samples were more likely to be obtained from infants. Across all the study visits, biological samples from pregnant participants were commonly collected at a single study visit (table 4). This was mostly observed for blood (60.7%), skin swab (67.4%), toenail (68.4%) and vaginal swab (66.7%) (table 4). Fewer participants provided samples across multiple study visits, and a small proportion contributed samples at all four visits, for example, 13.3% for saliva and 7.1% for blood (table 4).

Table 4. Distribution of biological samples collected from pregnant participants/mothers over I-QFC study visits, n (%).

Biological sample	1 time point only	2 time points only	3 time points only	All 4 time points	Total samples	Total enrolled pregnant participants with ≥1 sample
Blood	17 (60.7)	6 (21.4)	3 (10.7)	2 (7.1)	46	28 (27.2)
Urine	22 (53.7)	10 (24.4)	5 (12.2)	4 (9.8)	73	41 (39.8)
Skin swab	31 (67.4)	11 (23.9)	4 (8.7)	0	65	46 (44.7)
Cheek swab	16 (35.6)	14 (31.1)	10 (22.2)	5 (11.1)	94	45 (43.7)
Saliva	16 (36.4)	9 (20.5)	13 (29.5)	6 (13.6)	97	44 (42.7)
Microba/stool	10 (58.8)	5 (29.4)	2 (11.8)	0	26	17 (16.5)
Hair	27 (64.3)	15 (35.7)	0	0	57	42 (40.8)
Toenail	26 (68.4)	12 (31.6)	0	0	50	38 (36.9)
Vaginal swab*	24 (66.7)	12 (33.3)	0	na	48	36 (35.0)
Breastmilk	6 (85.7)	1 (14.3)	na	na	8	7 (6.8)

^*Vaginal swab collected only during pregnancy.

I-QFC, Indigenous Queensland Family Cohort.

Some participants did not provide biological samples due to refusal, telehealth visits, insufficient sample (eg, short nails, lack of hair), missed appointments or equipment unavailability. After birth, early hospital discharge was a key reason, along with limited research staff availability and births at non-affiliated hospitals.

Discussion

This paper reports findings from the I-QFC, a feasibility birth cohort study designed to explore how Aboriginal and Torres Strait Islander families engage with a longitudinal, family-centred research study. The study enrolled 103 pregnant participants and 64 partners. Recruitment occurred at a mean gestational age of 22.4 weeks, with strong engagement and high survey completion observed when facilitated by a research staff member. Participants represented diverse life experiences and reported a range of health-promoting behaviours. Birth outcomes were largely positive, with most infants born at term and within a healthy birth weight range.

Survey completion and participant engagement were notably higher when interactions were supported by a research midwife and/or Indigenous research assistant. In particular, the involvement of Indigenous research staff highlighted the importance of culturally safe communication and continuity of engagement with the same research staff member across multiple visits. These approaches were associated with enhanced participant engagement, improved response accuracy and reduced rates of incomplete survey submissions. The presence of research staff also supported health literacy by providing immediate clarification and reassurance, contributing to a more comfortable and informed survey experience. Several families reported forming meaningful and trusting relationships with the research team, particularly Indigenous research staff, which positively influenced their willingness to remain engaged throughout the study and attend follow-up visits. Nonetheless, gradual participant attrition was observed over time. In this context, feasibility was defined as the ability to recruit Indigenous families, implement a culturally safe protocol and identify barriers to sustained engagement. Attrition may be attributed in part to the burden of repeated data collection in the I-QFC study design. Retention was highest when follow-up visits were conducted by a non-identified research midwife, highlighting the value of tailored engagement strategies in longitudinal studies. In comparison, the broader QFC study reported a withdrawal rate of 10.6% at the 6-week postpartum visit.¹⁸

Retention patterns observed in I-QFC should also be considered in the context of other Indigenous birth cohort studies in Australia. For example, the Aboriginal Birth Cohort Study in the NT has achieved long-term follow-up through sustained community engagement, strong Indigenous governance and repeated face-to-face contact over decades, largely within a geographically defined population.¹⁰ In contrast, shorter-term cohorts such as the Gudaga Study¹¹ and the Gomeroi gaaynngal study,¹² which followed mother–child dyads over pregnancy and early childhood, have reported variable retention across follow-up periods, reflecting the practical challenges of longitudinal engagement in urban, rural and regional contexts. Compared with these studies, I-QFC was embedded within a hospital-based cohort, involved repeated data and biological sample collection across multiple time points, and included both mothers and partners during a period of significant family transition. The observed attrition in I-QFC is therefore not unexpected and highlights the logistical, cultural, and temporal barriers associated with intensive, family-centred cohort designs. Rather than undermining feasibility, these retention patterns provide critical methodological insights that inform refinement of recruitment strategies, follow-up timing and models of engagement for future Indigenous-led cohort studies.

According to the national Australian data,⁴ the proportion of Indigenous mothers who smoked during the first 20 weeks of pregnancy declined from 50% in 2011 to 42% in 2020. In this context, pregnant individuals in I-QFC reported a lower smoking rate at enrolment (28.6%), which decreased further at subsequent visits. However, these findings should be interpreted cautiously given the feasibility nature of the study and variable data completeness across time points. At enrolment, over half of the I-QFC participants reported alcohol consumption, although most of this likely occurred before pregnancy recognition, a common and well-documented occurrence. According to the National Drug Strategy Household Survey,⁴ 64% of women who were pregnant but unaware of their pregnancy reported alcohol consumption. Broader trends nonetheless suggest declining alcohol use during pregnancy among Indigenous women, with Symons et al²³ reporting a significant decline from 61% to 32% over 5 years. Together, these patterns are consistent with growing awareness and positive shifts in health behaviours among Indigenous women and underscore the importance of culturally safe, strengths-based approaches to maternal health research. As a feasibility study, I-QFC findings should therefore be interpreted as descriptive and exploratory, reflecting patterns of engagement and data availability rather than representative estimates of health behaviours or outcomes. Importantly, differential completion across time points and participant groups also highlights which study components were more or less acceptable or feasible, informing methodological refinement for future cohort studies.

Regarding biological samples, the main reasons participants did not provide their samples involved various logistical, cultural and procedural issues. Later stages of the study required participants to be physically present at the study site for biological samples collection, which likely contributed to the drop-off rate, particularly as late pregnancy is often associated with increased fatigue.²⁴ Additional insights were obtained through informal focus groups conducted with some participating families following their 6-week postpartum visit. Within this discussion, some male partners expressed reluctance to provide samples, noting that they were ‘not the ones having the baby’. In addition, a proportion of participants declined sample collection due to personal or cultural beliefs, most notably in relation to placental tissue (n=12, 18.5% collected). Unlike the broader QFC study, the I-QFC study did not have access to 24-hour placenta collection services and sample collection was limited to standard weekday working hours, while most births occurred outside these periods.

The placenta, membranes and blood hold a deep spiritual significance for many Indigenous communities, symbolising a sacred connection between the woman, baby and Country.^{25 26} Traditional practices such as placental burial are integral to maintaining cultural continuity, fostering a spiritual relationship with country and reinforcing cultural identity and belonging.²⁶ Historical research practices characterised by culturally unsafe consent processes, exploitation and misuse of biological data, and disregard for cultural protocols, have contributed to ongoing concerns among Indigenous communities regarding biological data collection.^{27 28} In the I-QFC study, participants retained full ownership of their data and biological samples, with the right to request their destruction or repatriation at any time. Telehealth-based appointments and disruptions related to the COVID-19 pandemic further limited opportunities for in-person sample collection. These multifaceted considerations underscore the complexities of collecting biological samples in diverse study settings during unprecedented times. Rather than representing methodological limitations alone, these observed patterns of biological sample participation and attrition provide important feasibility insights into how Indigenous families engage with longitudinal, family-centred birth cohort studies, particularly those involving biological sample collection from multiple family members. Earlier community consultation may have enabled shared decision-making around whether placental sampling was appropriate, how it should be undertaken or whether alternative biological measures aligned more closely with community priorities.

Strengths and limitations

A strength of this study is the inclusion of Indigenous fathers, addressing a critical gap in birth cohort research that has traditionally focused on maternal and child health. As discussed earlier, paternal health, lifestyle and socioeconomic factors play a vital role in shaping early childhood development and well-being yet remain underrepresented in existing literature. Although the exact mechanisms are not completely understood, there is growing evidence that paternal health significantly influences offspring health and development, and provides a broader overview of family well-being.^{29 30} A systematic review found that poor mental health in fathers is associated with increased risk of maternal postpartum depression as well as poorer mental, social, emotional and behavioural development in children later in life.³¹ Furthermore, recent findings from the QFC study, 180 couples, revealed that mothers whose partners reported low antenatal social support subsequently reported lower postnatal social support and higher postnatal depression in mothers.³² Therefore, the inclusion of Indigenous fathers provides a more comprehensive understanding of family dynamics, paternal influences on child development and the broader social and cultural contexts that shape parenting practices within Indigenous communities.^{15 33} This approach enhances the study’s relevance and impact, offering unique insights that may inform culturally responsive research, policy development and support strategies for Indigenous families.

Also, the study encompassed participants with pre-existing health conditions, including those who identify as single (pregnant individuals only), as well as surrogates (n=1). Another strength of the I-QFC study is the wide range of data collected within Indigenous families across the perinatal period, highlighting the feasibility of this study design for Indigenous Australian populations. Unlike samples from non-pregnant cohorts, these data capture critical developmental windows, offering insights into maternal and fetal health interactions that shape long-term outcomes.^{1 2}

A key limitation of this study related to the variable data completeness and participant attrition, which constrain longitudinal analyses and require cautious interpretation of the findings. As a feasibility study, IQFC was designed to explore engagement, retention and acceptability, rather than produce generalisable or hypothesis-driven outcomes. Engagement of both parents declined over time (table 1), likely due to competing demands on families, particularly mothers, of whom 64% were not first-time parents (table 2) and therefore balancing care responsibilities for multiple children. Reduced time availability and lower perceived urgency as pregnancy or the postpartum period approached may have further contributed to declining participation.³⁴ Several mothers also perceived the surveys as overly clinical and preferred a yarning approach. This is an informal conversational method that fosters a relaxed and culturally appropriate environment for data collection.³⁵ In addition, the cohort was recruited from a metropolitan setting, which may limit applicability to Indigenous families in rural and remote contexts.^{36 37} A further key limitation of I-QFC was the absence of male research staff, particularly Indigenous male health workers, which may have reduced engagement and data completeness among male partners. Evidence indicates that gender-sensitive approaches improve male participation in research,^{38 39} and that Indigenous men are more likely to engage when research is conducted in culturally safe and respectful ways.¹⁵ During study implementation, several pragmatic adaptations were undertaken to support engagement, including the modification of participant information statement, inclusion of Indigenous research staff, the addition of incentives (eg, Jarjum dolls) and inclusion of culturally-specific questions. In addition, unlike QFC, partner participation was optional and completion of surveys was not mandatory. However, opportunities for more substantial protocol modification were limited by the study’s embedding within the broader QFC protocol, underscoring the importance of adaptive methodologies in the early design phase for Indigenous cohort studies.

Unfortunately, community consultations with Indigenous peoples were not incorporated into the study design process for the QFC study. This absence at the design stage likely contributed to challenges observed later in the study, including variable acceptability of biological sample collection and differential engagement across study components. Engaging with the community is a fundamental aspect of conducting research with Indigenous populations, as it builds trust, ensures cultural relevance and delivers mutual benefits for both researchers and participants.⁴⁰ It enables the integration of Indigenous knowledge and perspectives, promotes culturally appropriate methodologies, and increases community ownership and engagement in the research process.⁴¹ As such, we recognise that the limitations identified in I-QFC require a more fundamental shift in research design, implementation and governance. Therefore, the findings from the I-QFC study, together with insights from the Indigenous Health Research Priorities (I-Priorities) Study,⁴² are now directly informing the co-design and methodological planning of a forthcoming Indigenous-led prospective birth cohort study across Queensland, Australia, the Strong Families Study,⁴³ which is grounded in community-identified health research priorities.

Data availability statement

Data are available on reasonable request.