Prevalence and socio-demographic predictors of food insecurity among regional and remote Western Australian children Stephanie Godrich,1 Johnny Lo,2 Christina Davies,3,4 Jill Darby,1 Amanda Devine1
T
he Social Determinants of Health (SDH) are “the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life”.1 The SDH include factors such as education, employment, income and availability of social support and related policies.1-4 Differences in these SDH exist between societies and as such, avoidable, unequal spread of SDH results in health inequities.3,4 People living in regional and remote areas are particularly susceptible to health inequities given the social and economic disadvantage associated with living in these areas.5-7 A key issue resulting from health inequities, particularly among vulnerable populations such as children, is food insecurity (FI).8,9
Abstract
FI results from reduced, restricted or uncertain physical and economic access to sufficient, safe, nutritious and appropriate food.10 Despite access to food being a basic human right,11 there is still no consistent method to measure FI prevalence among populations. Measurement tools range from individual questions in self-completed surveys12 to multi-item questionnaires delivered via trained interviewers.13 As a nation, Australia lags behind in the accurate measurement of FI among its population, especially children. Previous FI studies in Australia8,14-16 measured FI using a single question, or a few broad questions, which may not adequately establish the extent of the issue. Furthermore,
Implications for public health: One in five children were FI, demonstrating that FI is an issue in Western Australia.
Objective: Inequities can negatively impact the health outcomes of children. The aims of this study were to: i) ascertain the prevalence of food insecurity (FI) among regional and remote Western Australian (WA) children; and ii) determine which socio-demographic factors predicted child FI. Methods: Caregiver-child dyads (n=219) completed cross-sectional surveys. Descriptive statistics and logistic regression analyses were conducted using IBM SPSS version 23. Results: Overall, 20.1% of children were classified as FI. Children whose family received government financial assistance were more likely to be FI (OR 2.60; CI 1.15, 5.91; p=0.022), as were children living in a Medium disadvantage area (OR 2.60; CI 1.18, 5.72; p=0.017), compared to High or Low SEIFA ratings. Conclusions: Study findings are suggestive of the impact low income has on capacity to be food secure. The higher FI prevalence among children from families receiving financial assistance and living in medium disadvantage areas indicates more support for these families is required. Recommendations include: ensuring government plans and policies adequately support disadvantaged families; increasing employment opportunities; establishing evidence on the causes and the potential impact of FI on children’s health.
Key words: food security, regional, remote, children previous Australian studies measured FI among adults and child residents, but as reported by adults, and by and large on the urban population.8,14-16 The measurement of child-reported FI data is important, as it would facilitate a greater understanding of the issue and its determinants. Child-reported data would aid efforts to improve negative developmental, mental and physical health outcomes of FI.17 Existing evidence suggests that FI individuals may reduce the quality of food consumed,9
meal size or omit meals altogether, leading to severe hunger.18 The consequent reduced diet diversity, low consumption of core food groups19 and poor nutrient intake has been associated with negative health and educational outcomes among children, such as poor scholastic achievement,20 behavioural and emotional issues and absenteeism,16 nutrient deficiencies,20 poor general health16 and impaired development.19 The resultant diet-related health impacts include risk of unhealthy weight gain resulting in obesity,18
1. School of Medical and Health Sciences, Edith Cowan University, Western Australia 2. School of Science, Edith Cowan University, Western Australia 3. School of Population Health, The University of Western Australia 4. Public Health Advocacy Institute of Western Australia, Curtin University, Western Australia Correspondence to: Dr Stephanie Godrich, School of Medical and Health Sciences, Edith Cowan University, 270 Joondalup Drive, Joondalup, WA 6027; e-mail: [email protected] The authors have stated the following conflict of interest: SG is a consultant of Foodbank WA, a food relief organisation that provides nutrition education and cooking sessions to WA schools. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Aust NZ J Public Health. 2017; Online; doi: 10.1111/1753-6405.12716
2017 Online
Australian and New Zealand Journal of Public Health © 2017 The Authors
1
Godrich et al.
poor physical and mental health19 and increased risk of long-term chronic health conditions21 and non-communicable diseases.22 The health inequities underpinning these adverse health outcomes need to be mediated through points of intervention, such as the implementation of government policy3 that focus on the SDH in regional and remote areas. Despite government reports and strategic frameworks emphasising the importance of improving health disparities,7 particularly FI,9,23-27 there is a lack of evidence regarding child FI prevalence, as reported by the child, and the underlying SDH that are instrumental in predisposing children to FI. An additional current evidence gap includes a focus on regional and remote areas, given children living in these areas face higher levels of disadvantage.5,7 There are clear gaps in the availability of health data from these areas that need to be filled, in order to improve health inequity.7 The aims of this research were therefore to: i) ascertain the prevalence of FI among regional and remote Western Australian (WA) children; and ii) determine which socio-demographic factors increased the likelihood of child FI.
Methods Sampling and consent The study samples included school children (9–13 years) and their caregivers living in regional and remote WA. The age range was selected given fruit and vegetable consumption decreases with increasing age28 and to facilitate comparison of children’s dietary intake with the Australian Dietary Guidelines (a focus of the wider study). A master database including all non-metropolitan schools was compiled, using websites from each school governance structure.29-31 Schools were categorised by WA region, as defined by their listed location in each of the Department of Regional Development’s Region in Profile.32 The Australian Statistical Geography Standard (ASGS)33 was used to categorise schools by Remoteness Area (RA). Schools were also based in towns ranging in Socio-Economic Indexes for Areas (SEIFA) Index of Relative Socio-economic Disadvantage (IRSD),42 which classifies locations into index scores from one (most disadvantaged) to 10 (least disadvantaged).34 The concept of ‘disadvantage’ is defined by the Australian Bureau of Statistics in relation to “people’s access to material and social resources,
2
and their ability to participate in society”.36 This variable includes elements such as low household income, and low household rent.34 Non-probability sampling was undertaken; schools that were due to participate in the Foodbank WA Food Sensations® nutrition program were invited to participate in this study.
A total of 347 caregivers and 340 children indicated their consent to participate by returning signed caregiver and child consent forms. A caregiver–child dyad was the chosen method and, as matched caregiver and child surveys were a requirement for inclusion, 256 dyads were included in the sample.35
School principals (n=32) were telephoned and provided with a study overview to determine interest in receiving a follow up email containing Department of Education approval letter, principal invitation letter and consent form. Twenty-three principals provided written consent for their school to participate (71.8% participation rate). Schools declining to participate cited a lack of time or interest by their teachers or the existence of FI in their community, which they did not want highlighted. Seventy-six classes (71 teachers) were invited to participate by provision of a teacher information letter and consent form. Class teachers were required to provide written consent for their class’ involvement; a total of 97.2% of teachers did so (n=69 teachers, 74 classes). Classes were offered an on-site teacher and class briefing session (n=51), where possible, as an engagement strategy and to facilitate a clear understanding of the research processes.37 The session was used to explain the study, discuss the timeline and explain consent required. The remaining classes (n=23) either declined a briefing session or budgetary constraints prevented delivery. These classes received mailed study packs, which included an identical teacher information letter and detailed study procedures. Participating teachers and classes resulted in a total of 1,814 children and 1,814 of their caregivers being invited to participate in the study.38
Instrument development
As a requirement of study approval, and in addition to written principal and teacher consent, written informed consent was required from: 1) caregiver participants for their own participation; 2) caregiver participants for their child’s participation; and 3) child participants for their own participation. The take-home sealable caregiver/child consent form envelope was either disseminated in the teacher/class briefing session, or via class teachers. The form included a three-part consent form attached for caregivers and children to indicate consent. A tailored list of mental health, food relief and support services available in the family’s town/ region were also included, as a precautionary measure in case any of the survey questions caused distress among caregivers.35
Child and caregiver surveys were selfadministered, paper-based and pictorial to account for varying literacy levels. The child survey included the Child Food Security Survey Module (CFSSM),39 used with permission, (Table 1). The CFSSM is the most widely used tool to assess FI in children and has been assessed for validity and reliability in the US.39 The CFSSM is considered an excellent indicator of FI experienced by children, particularly among children ≥12 years of age, where responses were found to be more consistent among the small samples used in validation tests, compared to younger age groups.19,39 However, other research has demonstrated this tool was appropriate for children as young as six years of age40 and studies have verified children’s capacity as reliable reporters.41 Further, research by Connell et al. revealed children’s descriptions of FI strongly resembled those of adults.17 This suggests children are capable of understanding and reporting on their own FI experiences. The nine-item CFSSM tool included child-directed questions using a one-month reference period to assist with recall. For example, “In the last month... Did you worry that food at home would run out before your family got money to buy more?” Response options included “A lot”, “Sometimes” or “Never”. In addition, sociodemographic questions comprising selfreported gender, age, number of household residents were included. The caregiver survey comprised socio-demographic questions including gender, age, number of adults and children living in the household, caregiver educational attainment and caregiver employment status. A number of variables were used to indicate financial resources, including SEIFA IRSD42 and household receipt of government financial assistance. Table 2 contains a list of demographic variables included.
Data collection Data collection occurred between March 2013 and December 2015. This timeframe included pilot testing with 27 caregiverchild dyads. Reliability and face validity of
Australian and New Zealand Journal of Public Health © 2017 The Authors
2017 Online
Food insecurity prevalence among regional and remote WA children
the survey instruments was assessed and confirmed through test–retest during the pilot testing period. The latter was conducted through the provision of written feedback from teachers and caregivers regarding study processes and question understanding among themselves and their students/ children. The process for the overall sample included completion of caregiver surveys at home and returned to the class teacher via their child in the sealable consent form envelope. Class teachers disseminated surveys to consenting students, who used a privacy sticker to secure their survey upon completion. All child and caregiver surveys and consent form envelopes were collected by class teachers and posted to the study centre in a pre-paid envelope.35
Table 1: Child Food Security Survey Module Results, Adapted from Connell, Nord et al., J Nutr (2004; 134; 10; 2566-2572). Child Food Security Survey Module Questions Did you worry that food at home would run out before your family got money to buy more? Did the food that your family had run out and you didn’t have money to get more? How often were you not able to eat a balanced meal because your family didn’t have enough money? Did your meals only include a few kinds of cheap foods because your family was running out of money to buy food? Has the size of your meals been cut because your family didn’t have enough money for food? Did you have to eat less because your family didn’t have enough money to buy food? Did you have to skip a meal because your family didn’t have enough money for food? Were you hungry but didn’t eat because your family didn’t have enough food? Did you ever not eat for the whole day because your family didn’t have enough money for food?
Statistical analysis Paper-based child and caregiver surveys were entered into separate Microsoft Excel spreadsheets and imported into IBM SPSS version 23 (IBM, Armonk, NY, USA) for analyses. Any cases with incomplete data were excluded in analyses; the final sample included n=219 cases. Analyses of the overall sample data included descriptive statistics to ascertain the prevalence of child FI, followed by simple and multivariable logistic regression analyses to determine the sociodemographic predictors of child FI. Child survey data included in analyses comprised gender, age and number of household residents and the CFSSM. Caregiver survey data included in analyses encompassed caregiver gender, caregiver age, number of child residents, caregiver educational attainment, caregiver employment status, SEIFA IRSD (based on town name) and receipt of government financial assistance.
219
14.6%
219 219 218 219 219
5.0% 9.6% 3.3% 6.8% 1.8%
Total n 219
Total % 100
33 186
15.1 84.9
1.33 (0.55–3.19) 1.00 (ref)
Unemployed/volunteer Part time Full time
34 87 98
15.5 39.7 44.7
1.73 (0.71–4.22) 0.86 (0.41–1.83) 1.00 (ref)
Primary school/Secondary school Diploma/Apprenticeship Undergraduate University degree/ Post-graduate University degree Yes No
96 67 56
43.8 30.6 25.6
2.11 (0.88–5.07) 1.17 (0.43–3.16) 1.00 (ref)
54 165
24.7 75.3
2.34 (1.15–4.76) 1.00 (ref)
Low score, high disadvantage Medium score, medium disadvantage High score, low disadvantage Regional Remote
127 64 28 133 86
58.0 29.2 12.8 60.7 39.3
1.00 (ref) 2.01 (0.99–4.08) 0.57 (0.15–2.06) 1.00 (ref) 1.37 (0.70–2.68)
28 94 65 22 10
12.8 42.9 29.7 10.0 4.6
1.00 (ref) 0.94 (0.31–2.85) 1.04 (0.32–3.29) 1.72 (0.44–6.63) 4.60 (0.95–22.16)
22 106 63 24 4 71 148
10.0 48.4 28.8 11.0 1.8 32.4 67.6
1.00 (ref) 0.50 (0.17–1.48) 0.62 (0.20–1.94) 1.09 (0.30–3.97) 2.66 (0.30–23.42) 1.41 (0.71–2.8) 1.00 (ref)
20 60 78 46 15
9.1 27.4 35.6 21.0 6.8
1.00 (ref) 1.09 (0.34–3.48) 0.77 (0.24–2.44) 0.36 (0.09–1.44) 0.46 (0.07–2.79)
26-66 years
Caregiver gender
Male Female
OR (95% CI) p-value 0.98 (0.93–1.03) 0.490
Caregiver Employment Status
Caregiver Educational Attainment
Family Receipt of Financial Assistance SEIFA IRSD decile range
Remote Location
2-3 4 5 6 7 or more Number of resident children
Child gender
0.519 0.323 0.227 0.707 0.151+ 0.094 0.745
0.018+ 0.061+
Number of total residents
Variable recoding Due to low cell counts, the response categories for a number of variables included in the overall sample were recoded using IBM SPSS Statistics (version 23). Variables included caregiver educational attainment, number of household residents, caregiver employment status, SEIFA IRSD decile. For example, caregiver educational attainment was recoded to ‘Primary School or Secondary School’, ‘Diploma or Apprenticeship’ or ‘Undergraduate/Postgraduate University Degree’ (Table 2). All recoded variables were manually checked to ensure correct recoding.
217 219 218
Child Food Security Survey Module Responses of “A lot” or “Sometimes” Combined 21.2% 6.4% 7.8%
Table 2: Simple logistic regression models for socio-demographic factors and child food insecurity (n=219). Variable Caregiver Age
Data analysis
n
1 2 3 4 5 Male Female
0.052 0.394 0.348 0.186+ 0.918 0.945 0.428 0.057 0.298 0.217 0.419 0.887 0.376 0.326 0.333
Child Age 9 10 11 12 13
0.883 0.663 0.151 0.400
1.00 (ref) = reference category. SEIFA indicates Socio-Economic Indexes for Areas Index of Relative Socio-Economic Disadvantage. +Significant at p
T
he Social Determinants of Health (SDH) are “the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life”.1 The SDH include factors such as education, employment, income and availability of social support and related policies.1-4 Differences in these SDH exist between societies and as such, avoidable, unequal spread of SDH results in health inequities.3,4 People living in regional and remote areas are particularly susceptible to health inequities given the social and economic disadvantage associated with living in these areas.5-7 A key issue resulting from health inequities, particularly among vulnerable populations such as children, is food insecurity (FI).8,9
Abstract
FI results from reduced, restricted or uncertain physical and economic access to sufficient, safe, nutritious and appropriate food.10 Despite access to food being a basic human right,11 there is still no consistent method to measure FI prevalence among populations. Measurement tools range from individual questions in self-completed surveys12 to multi-item questionnaires delivered via trained interviewers.13 As a nation, Australia lags behind in the accurate measurement of FI among its population, especially children. Previous FI studies in Australia8,14-16 measured FI using a single question, or a few broad questions, which may not adequately establish the extent of the issue. Furthermore,
Implications for public health: One in five children were FI, demonstrating that FI is an issue in Western Australia.
Objective: Inequities can negatively impact the health outcomes of children. The aims of this study were to: i) ascertain the prevalence of food insecurity (FI) among regional and remote Western Australian (WA) children; and ii) determine which socio-demographic factors predicted child FI. Methods: Caregiver-child dyads (n=219) completed cross-sectional surveys. Descriptive statistics and logistic regression analyses were conducted using IBM SPSS version 23. Results: Overall, 20.1% of children were classified as FI. Children whose family received government financial assistance were more likely to be FI (OR 2.60; CI 1.15, 5.91; p=0.022), as were children living in a Medium disadvantage area (OR 2.60; CI 1.18, 5.72; p=0.017), compared to High or Low SEIFA ratings. Conclusions: Study findings are suggestive of the impact low income has on capacity to be food secure. The higher FI prevalence among children from families receiving financial assistance and living in medium disadvantage areas indicates more support for these families is required. Recommendations include: ensuring government plans and policies adequately support disadvantaged families; increasing employment opportunities; establishing evidence on the causes and the potential impact of FI on children’s health.
Key words: food security, regional, remote, children previous Australian studies measured FI among adults and child residents, but as reported by adults, and by and large on the urban population.8,14-16 The measurement of child-reported FI data is important, as it would facilitate a greater understanding of the issue and its determinants. Child-reported data would aid efforts to improve negative developmental, mental and physical health outcomes of FI.17 Existing evidence suggests that FI individuals may reduce the quality of food consumed,9
meal size or omit meals altogether, leading to severe hunger.18 The consequent reduced diet diversity, low consumption of core food groups19 and poor nutrient intake has been associated with negative health and educational outcomes among children, such as poor scholastic achievement,20 behavioural and emotional issues and absenteeism,16 nutrient deficiencies,20 poor general health16 and impaired development.19 The resultant diet-related health impacts include risk of unhealthy weight gain resulting in obesity,18
1. School of Medical and Health Sciences, Edith Cowan University, Western Australia 2. School of Science, Edith Cowan University, Western Australia 3. School of Population Health, The University of Western Australia 4. Public Health Advocacy Institute of Western Australia, Curtin University, Western Australia Correspondence to: Dr Stephanie Godrich, School of Medical and Health Sciences, Edith Cowan University, 270 Joondalup Drive, Joondalup, WA 6027; e-mail: [email protected] The authors have stated the following conflict of interest: SG is a consultant of Foodbank WA, a food relief organisation that provides nutrition education and cooking sessions to WA schools. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Aust NZ J Public Health. 2017; Online; doi: 10.1111/1753-6405.12716
2017 Online
Australian and New Zealand Journal of Public Health © 2017 The Authors
1
Godrich et al.
poor physical and mental health19 and increased risk of long-term chronic health conditions21 and non-communicable diseases.22 The health inequities underpinning these adverse health outcomes need to be mediated through points of intervention, such as the implementation of government policy3 that focus on the SDH in regional and remote areas. Despite government reports and strategic frameworks emphasising the importance of improving health disparities,7 particularly FI,9,23-27 there is a lack of evidence regarding child FI prevalence, as reported by the child, and the underlying SDH that are instrumental in predisposing children to FI. An additional current evidence gap includes a focus on regional and remote areas, given children living in these areas face higher levels of disadvantage.5,7 There are clear gaps in the availability of health data from these areas that need to be filled, in order to improve health inequity.7 The aims of this research were therefore to: i) ascertain the prevalence of FI among regional and remote Western Australian (WA) children; and ii) determine which socio-demographic factors increased the likelihood of child FI.
Methods Sampling and consent The study samples included school children (9–13 years) and their caregivers living in regional and remote WA. The age range was selected given fruit and vegetable consumption decreases with increasing age28 and to facilitate comparison of children’s dietary intake with the Australian Dietary Guidelines (a focus of the wider study). A master database including all non-metropolitan schools was compiled, using websites from each school governance structure.29-31 Schools were categorised by WA region, as defined by their listed location in each of the Department of Regional Development’s Region in Profile.32 The Australian Statistical Geography Standard (ASGS)33 was used to categorise schools by Remoteness Area (RA). Schools were also based in towns ranging in Socio-Economic Indexes for Areas (SEIFA) Index of Relative Socio-economic Disadvantage (IRSD),42 which classifies locations into index scores from one (most disadvantaged) to 10 (least disadvantaged).34 The concept of ‘disadvantage’ is defined by the Australian Bureau of Statistics in relation to “people’s access to material and social resources,
2
and their ability to participate in society”.36 This variable includes elements such as low household income, and low household rent.34 Non-probability sampling was undertaken; schools that were due to participate in the Foodbank WA Food Sensations® nutrition program were invited to participate in this study.
A total of 347 caregivers and 340 children indicated their consent to participate by returning signed caregiver and child consent forms. A caregiver–child dyad was the chosen method and, as matched caregiver and child surveys were a requirement for inclusion, 256 dyads were included in the sample.35
School principals (n=32) were telephoned and provided with a study overview to determine interest in receiving a follow up email containing Department of Education approval letter, principal invitation letter and consent form. Twenty-three principals provided written consent for their school to participate (71.8% participation rate). Schools declining to participate cited a lack of time or interest by their teachers or the existence of FI in their community, which they did not want highlighted. Seventy-six classes (71 teachers) were invited to participate by provision of a teacher information letter and consent form. Class teachers were required to provide written consent for their class’ involvement; a total of 97.2% of teachers did so (n=69 teachers, 74 classes). Classes were offered an on-site teacher and class briefing session (n=51), where possible, as an engagement strategy and to facilitate a clear understanding of the research processes.37 The session was used to explain the study, discuss the timeline and explain consent required. The remaining classes (n=23) either declined a briefing session or budgetary constraints prevented delivery. These classes received mailed study packs, which included an identical teacher information letter and detailed study procedures. Participating teachers and classes resulted in a total of 1,814 children and 1,814 of their caregivers being invited to participate in the study.38
Instrument development
As a requirement of study approval, and in addition to written principal and teacher consent, written informed consent was required from: 1) caregiver participants for their own participation; 2) caregiver participants for their child’s participation; and 3) child participants for their own participation. The take-home sealable caregiver/child consent form envelope was either disseminated in the teacher/class briefing session, or via class teachers. The form included a three-part consent form attached for caregivers and children to indicate consent. A tailored list of mental health, food relief and support services available in the family’s town/ region were also included, as a precautionary measure in case any of the survey questions caused distress among caregivers.35
Child and caregiver surveys were selfadministered, paper-based and pictorial to account for varying literacy levels. The child survey included the Child Food Security Survey Module (CFSSM),39 used with permission, (Table 1). The CFSSM is the most widely used tool to assess FI in children and has been assessed for validity and reliability in the US.39 The CFSSM is considered an excellent indicator of FI experienced by children, particularly among children ≥12 years of age, where responses were found to be more consistent among the small samples used in validation tests, compared to younger age groups.19,39 However, other research has demonstrated this tool was appropriate for children as young as six years of age40 and studies have verified children’s capacity as reliable reporters.41 Further, research by Connell et al. revealed children’s descriptions of FI strongly resembled those of adults.17 This suggests children are capable of understanding and reporting on their own FI experiences. The nine-item CFSSM tool included child-directed questions using a one-month reference period to assist with recall. For example, “In the last month... Did you worry that food at home would run out before your family got money to buy more?” Response options included “A lot”, “Sometimes” or “Never”. In addition, sociodemographic questions comprising selfreported gender, age, number of household residents were included. The caregiver survey comprised socio-demographic questions including gender, age, number of adults and children living in the household, caregiver educational attainment and caregiver employment status. A number of variables were used to indicate financial resources, including SEIFA IRSD42 and household receipt of government financial assistance. Table 2 contains a list of demographic variables included.
Data collection Data collection occurred between March 2013 and December 2015. This timeframe included pilot testing with 27 caregiverchild dyads. Reliability and face validity of
Australian and New Zealand Journal of Public Health © 2017 The Authors
2017 Online
Food insecurity prevalence among regional and remote WA children
the survey instruments was assessed and confirmed through test–retest during the pilot testing period. The latter was conducted through the provision of written feedback from teachers and caregivers regarding study processes and question understanding among themselves and their students/ children. The process for the overall sample included completion of caregiver surveys at home and returned to the class teacher via their child in the sealable consent form envelope. Class teachers disseminated surveys to consenting students, who used a privacy sticker to secure their survey upon completion. All child and caregiver surveys and consent form envelopes were collected by class teachers and posted to the study centre in a pre-paid envelope.35
Table 1: Child Food Security Survey Module Results, Adapted from Connell, Nord et al., J Nutr (2004; 134; 10; 2566-2572). Child Food Security Survey Module Questions Did you worry that food at home would run out before your family got money to buy more? Did the food that your family had run out and you didn’t have money to get more? How often were you not able to eat a balanced meal because your family didn’t have enough money? Did your meals only include a few kinds of cheap foods because your family was running out of money to buy food? Has the size of your meals been cut because your family didn’t have enough money for food? Did you have to eat less because your family didn’t have enough money to buy food? Did you have to skip a meal because your family didn’t have enough money for food? Were you hungry but didn’t eat because your family didn’t have enough food? Did you ever not eat for the whole day because your family didn’t have enough money for food?
Statistical analysis Paper-based child and caregiver surveys were entered into separate Microsoft Excel spreadsheets and imported into IBM SPSS version 23 (IBM, Armonk, NY, USA) for analyses. Any cases with incomplete data were excluded in analyses; the final sample included n=219 cases. Analyses of the overall sample data included descriptive statistics to ascertain the prevalence of child FI, followed by simple and multivariable logistic regression analyses to determine the sociodemographic predictors of child FI. Child survey data included in analyses comprised gender, age and number of household residents and the CFSSM. Caregiver survey data included in analyses encompassed caregiver gender, caregiver age, number of child residents, caregiver educational attainment, caregiver employment status, SEIFA IRSD (based on town name) and receipt of government financial assistance.
219
14.6%
219 219 218 219 219
5.0% 9.6% 3.3% 6.8% 1.8%
Total n 219
Total % 100
33 186
15.1 84.9
1.33 (0.55–3.19) 1.00 (ref)
Unemployed/volunteer Part time Full time
34 87 98
15.5 39.7 44.7
1.73 (0.71–4.22) 0.86 (0.41–1.83) 1.00 (ref)
Primary school/Secondary school Diploma/Apprenticeship Undergraduate University degree/ Post-graduate University degree Yes No
96 67 56
43.8 30.6 25.6
2.11 (0.88–5.07) 1.17 (0.43–3.16) 1.00 (ref)
54 165
24.7 75.3
2.34 (1.15–4.76) 1.00 (ref)
Low score, high disadvantage Medium score, medium disadvantage High score, low disadvantage Regional Remote
127 64 28 133 86
58.0 29.2 12.8 60.7 39.3
1.00 (ref) 2.01 (0.99–4.08) 0.57 (0.15–2.06) 1.00 (ref) 1.37 (0.70–2.68)
28 94 65 22 10
12.8 42.9 29.7 10.0 4.6
1.00 (ref) 0.94 (0.31–2.85) 1.04 (0.32–3.29) 1.72 (0.44–6.63) 4.60 (0.95–22.16)
22 106 63 24 4 71 148
10.0 48.4 28.8 11.0 1.8 32.4 67.6
1.00 (ref) 0.50 (0.17–1.48) 0.62 (0.20–1.94) 1.09 (0.30–3.97) 2.66 (0.30–23.42) 1.41 (0.71–2.8) 1.00 (ref)
20 60 78 46 15
9.1 27.4 35.6 21.0 6.8
1.00 (ref) 1.09 (0.34–3.48) 0.77 (0.24–2.44) 0.36 (0.09–1.44) 0.46 (0.07–2.79)
26-66 years
Caregiver gender
Male Female
OR (95% CI) p-value 0.98 (0.93–1.03) 0.490
Caregiver Employment Status
Caregiver Educational Attainment
Family Receipt of Financial Assistance SEIFA IRSD decile range
Remote Location
2-3 4 5 6 7 or more Number of resident children
Child gender
0.519 0.323 0.227 0.707 0.151+ 0.094 0.745
0.018+ 0.061+
Number of total residents
Variable recoding Due to low cell counts, the response categories for a number of variables included in the overall sample were recoded using IBM SPSS Statistics (version 23). Variables included caregiver educational attainment, number of household residents, caregiver employment status, SEIFA IRSD decile. For example, caregiver educational attainment was recoded to ‘Primary School or Secondary School’, ‘Diploma or Apprenticeship’ or ‘Undergraduate/Postgraduate University Degree’ (Table 2). All recoded variables were manually checked to ensure correct recoding.
217 219 218
Child Food Security Survey Module Responses of “A lot” or “Sometimes” Combined 21.2% 6.4% 7.8%
Table 2: Simple logistic regression models for socio-demographic factors and child food insecurity (n=219). Variable Caregiver Age
Data analysis
n
1 2 3 4 5 Male Female
0.052 0.394 0.348 0.186+ 0.918 0.945 0.428 0.057 0.298 0.217 0.419 0.887 0.376 0.326 0.333
Child Age 9 10 11 12 13
0.883 0.663 0.151 0.400
1.00 (ref) = reference category. SEIFA indicates Socio-Economic Indexes for Areas Index of Relative Socio-Economic Disadvantage. +Significant at p
