Abstract
Objective
To examine the relationship between household food insecurity and healthcare costs in children living in Ontario, Canada.
Methods
We conducted a cross-sectional, population-based study using four cycles of the Canadian Community Health Survey (2007–2008, 2009–2010, 2011–2012, 2013–2014) linked with administrative health databases (ICES). We included Ontario children aged 1–17 years with a measure of household food insecurity (Household Food Security Survey Module) over the previous 12 months. Our primary outcome was the direct public-payer healthcare costs per child over the same time period (in Canadian dollars, standardized to year 2020). We used gamma-log–transformed generalized estimating equations accounting for the clustering of children to examine this relationship, and adjusted models for important sociodemographic covariates. As a secondary outcome, we examined healthcare usage of specific services and associated costs (e.g. visits to hospitals, surgeries).
Results
We found that adjusted healthcare costs were higher in children from food-insecure than from food-secure households ($676.79 [95% CI: $535.26, $855.74] vs. $563.98 [$457.00, $695.99], p = 0.047). Compared with children living in food-secure households, those in insecure households more often accessed hospitals, emergency departments, day surgeries, and home care, and used prescription medications. Children from food-secure households had higher usage of non-physician healthcare (e.g. optometry) and family physician rostering services.
Conclusion
Even after adjusting for measurable social determinants of health, household food insecurity was associated with higher public-payer health services costs and utilization among children and youth. Efforts to mitigate food insecurity could lessen child healthcare needs, as well as associated costs to our healthcare systems.
Supplementary Information
The online version contains supplementary material available at 10.17269/s41997-023-00812-2.
Keywords: Food insecurity; Children; Costs, healthcare
Résumé
Objectif
Examiner la relation entre l’insécurité alimentaire des ménages et les coûts des soins de santé chez les enfants vivant en Ontario, au Canada.
Méthode
Nous avons mené une étude populationnelle transversale en utilisant les quatre cycles de l’Enquête sur la santé dans les collectivités canadiennes (2007–2008, 2009–2010, 2011–2012, 2013–2014) liés à des bases de données administratives sur la santé (ICES). Nous avons inclus les enfants ontariens de 1 à 17 ans et un indicateur d’insécurité alimentaire des ménages (le Module d’enquête sur la sécurité alimentaire des ménages) au cours des 12 mois antérieurs. Les coûts directs des soins de santé publics par enfant au cours de cette période (en dollars canadiens de 2020) ont constitué notre résultat principal. Nous avons utilisé des équations d’estimation généralisées transformées par la fonction logarithme gamma tenant compte du regroupement des enfants pour analyser cette relation, et des modèles ajustés pour les covariables sociodémographiques importantes. Comme résultat secondaire, nous avons analysé l’utilisation de certains services de soins de santé (p. ex. les visites dans les hôpitaux, les chirurgies) et les coûts associés.
Résultats
Nous avons constaté que les coûts ajustés des soins de santé étaient plus élevés chez les enfants des ménages aux prises avec l’insécurité alimentaire que chez ceux des ménages à l’abri de l’insécurité alimentaire (676,79 $ [IC de 95%: 535,26 $, 855,74 $] contre 563,98 $ [457,00 $, 695,99 $], p = 0,047). Comparativement aux enfants des ménages à l’abri de l’insécurité alimentaire, ceux qui vivaient dans des ménages aux prises avec l’insécurité avaient plus souvent recours aux hôpitaux, aux services des urgences, aux chirurgies d’un jour et aux soins à domicile, et ils prenaient des médicaments sur ordonnance. Les enfants des ménages à l’abri de l’insécurité alimentaire avaient plus souvent recours aux soins de santé non médicaux (p. ex. l’optométrie) et aux services de leur médecin de famille attitré.
Conclusion
Même après l’apport d’ajustements pour tenir compte des déterminants sociaux de la santé mesurables, l’insécurité alimentaire des ménages était associée à des coûts de soins de santé publics plus élevés et à une plus grande utilisation de ces soins chez les enfants et les jeunes. Des efforts pour atténuer l’insécurité alimentaire pourraient réduire les besoins de soins de santé des enfants, ainsi que les coûts associés pour nos systèmes de soins de santé.
Mots-clés: Insécurité alimentaire, enfants, coûts, soins de santé
Introduction
In 2022, 6.9 million Canadians (including almost 1.8 million children) lived in a food-insecure household, an increase from 5.8 million individuals in 2021. The detrimental health and psychosocial consequences of food insecurity for North American children have been extensively documented; children from food-insecure households have more challenges with diet (Kirkpatrick & Tarasuk, 2008), have poorer mental health (Tarasuk et al., 2018), more often live with obesity (Kaur et al., 2015), and face increased rates of intentional and unintentional injuries (Men et al., 2021a).
In Canadian adults, the health and psychosocial consequences of food insecurity have translated into higher healthcare utilization and costs. Indeed in previous studies, both total healthcare costs and mean individual healthcare costs were much higher in food-insecure vs food-secure adults even in adjusted analysis (Tarasuk et al., 2015). The relationship between household food insecurity in childhood and healthcare costs has not been explored. Understanding this relationship might assist governments and policy makers with healthcare planning (e.g. anticipating service needs and costs), as well as the allocation of resources (i.e. investment in policies to address food insecurity).
With access to a comprehensive national survey (Canadian Community Health Survey, hereafter CCHS) and the ability to link the CCHS with administrative health databases from the province of Ontario, we examined the connection between household food insecurity and direct healthcare costs in a large cohort of Canadian children. We hypothesized that children living in food-insecure households would have more healthcare usage and higher public-payer healthcare costs than children from food-secure households.
Methods
We followed the guidelines for the REporting of studies Conducted using Observational Routinely-collected Data (RECORD) (Supplementary Materials Table 1) (Benchimol et al., 2015).
Design and setting
We conducted a cross-sectional, population-based study of children aged 1–17 years from Ontario, Canada. At the time of this study, all children had universal access to hospital and physician services, and diagnostic tests through Ontario’s Health Insurance Plan (OHIP). Prescription medications were only OHIP-covered for people on social assistance, those living with select health conditions (e.g. cancer), and those 65 years of age or older. Data on the use of hospital, physician, diagnostic services, and publicly funded prescription medications are held in administrative health databases at ICES. In 2018, the institute formerly known as the Institute for Clinical Evaluative Sciences formally adopted the initialism ICES as its official name. This change acknowledges the growth and evolution of the organization’s research since its inception in 1992, while retaining the familiarity of the former acronym within the scientific community and beyond. ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze healthcare and demographic data, without consent, for health system evaluation and improvement. Use of data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a Research Ethics Board.
Data sources
Health administrative data
Health administrative databases (ICES) were used to summarize the demographic characteristics, comorbidities, and healthcare usage of included children as well as their mothers. A detailed description of the health administrative databases used and the variables they contain are included in Supplementary Materials Table 2. All databases were linked using unique encoded identifiers and analyzed at ICES.
In brief, we used the Registered Persons Database of Ontario for vital statistics. To grow our cohort of eligible children for inclusion, we used the ICES-developed MOMBABY database, derived from the CIHI-DAD. MOMBABY allowed us to link mothers with their children born in an Ontario hospital (see description of cohort build below). We also used the IRCC-PR to determine the migrant status of mothers. The ON-MARG dataset provided levels of marginalization (e.g. dependency) by dissemination area or census tract (Matheson et al., 2012).
To quantify child-level healthcare usage and costs, we used the CIHI-DAD (hospitalizations), the National Ambulatory Care Reporting System Database (emergency department visits), the Same Day Surgery Database, and the Home Care database (services provided by or coordinated by community care). The OHIP database contains information about visits to physicians, and laboratory and diagnostic testing. For data on prescription medications, we used the Ontario Drug Benefits database (holds drugs listed on the provincial formulary). The Client Agency Program Enrollment Database was used to describe the enrolment of children with specific primary care groups (i.e. “rostering”).
We then used validated costing guidelines to quantify healthcare expenditures associated with health service utilization (Wodchis et al., 2013). To do this, we leveraged additional datasets including the Ontario Case Costing Information Database (cost data for acute inpatient, day surgery, and ambulatory care cases), EstSOBPrice (approximates cost of OHIP services), and the GAPP Decision Support System (assessment of physician payments). Costing guidelines and associated databases have been used across several impactful research studies in Ontario (Luca et al., 2018; Rosella et al., 2014; Tarasuk et al., 2015).
CCHS
The CCHS is a national cross-sectional survey of Canadians aged 12 years of age and older. It contains population-level data on both health and the determinants of health (Health Canada, 2006). Select cycles of the CCHS include the 18-item Household Food Security Survey Module (HFSSM), a gold standard measure of food insecurity status in Canada. HFSSM survey items ask respondents about income-related problems associated with inadequate or insecure access to food in their household in the previous 12 months (Government of Canada, 2020).
Population
Using the methods described in our previous research (Clemens et al., 2020, 2021), we identified all Ontarian children with a household response to the HFSSM from CCHS cycles 2005, 2007–2008, 2009–2010, 2011–2012, and 2013–2014. Respondents (Ontario sample) were linked to ICES data holdings with their provincial health card number using a combination of deterministic and probabilistic linkage methods; the provincial health card was then encoded to enable linkage with other data holdings at ICES.
We grew our cohort of eligible children not only by examining the HFSSM survey responses of children aged 12–17 years, but also by including the siblings of child respondents, as well as children’s mothers (via MOMBABY). We restricted inclusion to those aged 1–17 years to facilitate a full year of look-back for use of health services. We further restricted to children with complete costing data available (i.e. CCHS survey date on or after April 1, 2007) as outpatient care costs only became measurable after this time.
Exposure
Knowing the relatively small sample size of children from food-insecure households from prior studies, we a priori defined our primary exposure as a binary measure of household food insecurity (i.e. any affirmative response on the HFSSM which included marginally, moderately, and severely food-insecure categories). Prior work has suggested that those from marginally food-insecure households are more similar to those from food-insecure than from food-secure households (Cook et al., 2013). In sensitivity analysis, we used a four-level categorical variable which is based on the number of affirmative responses (Supplementary Materials Table 3).
Characteristics
We considered the following baseline characteristics of included children: child’s age, sex, survey cycle, rurality of residence, marginalization index, mother’s age at child’s birth, mother’s immigration status, ethnic origin (self-identified ethnic group to which the respondents’ ancestors belonged), and racial identity. We also include characteristics of the child’s household, including the household income decile, smoking status and level of education, single-parent status, and the number of children in the home. The details of presented covariates are available in Supplementary Materials Table 4.
Outcomes
Our primary outcome was direct per-child healthcare costs in the 12 months prior to administration of the HFSSM, adjusted to the year 2020 in Canadian dollars. Direct costs included the costs of hospitalizations, day surgeries, emergency department visits, home care, outpatient and ambulatory care, laboratory testing, and public-payer prescription medications. Since government-funded prescription medications were only available to select families including those on social assistance over the study period, in a sub-analysis, we provided total healthcare costs but excluded public-payer prescription medications. As secondary outcomes, we assessed the healthcare usage of children by food insecurity status as well as the cost of individual health services (e.g. hospitalizations).
Analysis
We summarized the demographic and socioeconomic characteristics of included children descriptively, and displayed the mean (SD) healthcare costs by food insecurity status. To assess the relationship between food insecurity and direct per-child healthcare costs, we used generalized estimating equations to fit gamma-log models, and accounted for the clustering of children who lived in the same household (Clemens et al., 2020, 2021). We adjusted our primary model for sociodemographic factors that have been linked with both food insecurity and healthcare utilization in children (Wong, 1999). These included the child’s age, sex, racial identity, mother’s immigrant status, household smoking status, single-parent status, income distribution, number of children in the household, highest level of household education, and survey cycle. We used similar analyses for our secondary outcomes. As a post hoc sensitivity analysis, we examined the unadjusted relationship between food insecurity and care costs using the 4-category food insecurity status exposure variable.
Results
Details of our cohort build are provided in Supplementary Materials Table 5. There were a total of 167,442 Ontarians who participated in a CCHS survey over the study period. We were able to link all respondents with ICES databases, apart from those who did not have a valid health card number or CCHS interview date (17,224 or ~ 10% of CCHS respondents were excluded). Additional exclusions at the child, sibling, and maternal levels are provided in Supplementary Materials Table 5 (exclusions based upon data cleaning, missing responses on the HFSSM, lack of children in the home).
A total of 25,222 children met inclusion for the study, of whom 3991 (15.8%) were living in food-insecure households (6.0% from marginally food-insecure, 7.1% from moderately food-insecure, and 2.7% from severely food-insecure households). Compared to children in food-secure households, those from food-insecure households more often identified as non-Caucasian, lived in marginalized neighbourhoods, and had a household income in the lower income deciles. These children also more often lived in non-owned homes, resided with smokers, had single parents, and had several other children in their home (Table 1).
Table 1.
Characteristics of the 25,222 children included in the study, by food insecurity status
| Characteristic | Total n = 25,222 | Food-secure n = 21,231 | Food-insecure n = 3991 | p value |
|---|---|---|---|---|
| Age at interview date, years, mean ± SD | 9.72 ± 4.86 | 9.75 ± 4.87 | 9.55 ± 4.81 | 0.013 |
| Female sex, N (%) | 12,287 (48.7%) | 10,338 (48.7%) | 1949 (48.8%) | 0.869 |
| Rural residence, yes, N (%) | 5174 (20.5%) | 4390 (20.7%) | 784 (19.6%) | 0.303 |
| Ethnic origin, N (%) | ||||
| European | 20,023 (79.4%) | 17,117 (80.6%) | 2906 (72.8%) | < 0.001 |
| Chinese | 519 (2.1%) | 465 (2.2%) | 54 (1.4%) | < 0.001 |
| South Asian | 891 (3.5%) | 707 (3.3%) | 184 (4.6%) | < 0.001 |
| Other | 6844 (27.1%) | 5390 (25.4%) | 1454 (36.4%) | < 0.001 |
| Racial belonging, N (%) | ||||
| White | 20,804 (82.5%) | 17,923 (84.4%) | 2881 (72.2%) | < 0.001 |
| Black | 798 (3.2%) | 508 (2.4%) | 290 (7.3%) | < 0.001 |
| East/Southeast Asian | 1009 (4.0%) | 883 (4.2%) | 126 (3.2%) | 0.003 |
| West Asian/Arab | 301 (1.2%) | 215 (1.0%) | 86 (2.2%) | < 0.001 |
| South Asian | 906 (3.6%) | 712 (3.4%) | 194 (4.9%) | < 0.001 |
| Latin American | 236 (0.9%) | 180 (0.8%) | 56 (1.4%) | < 0.001 |
| Other | 1556 (6.2%) | 1113 (5.2%) | 443 (11.1%) | < 0.001 |
| Residential deprivation, N (%) | ||||
| Quintile 1 (least deprived) | 5784 (22.9%) | 5352 (25.2%) | 432 (10.8%) | < 0.001 |
| Quintile 2 | 5914 (23.4%) | 5244 (24.7%) | 670 (16.8%) | |
| Quintile 3 | 5225 (20.7%) | 4410 (20.8%) | 815 (20.4%) | |
| Quintile 4 | 4305 (17.1%) | 3452 (16.3%) | 853 (21.4%) | |
| Quintile 5 (most deprived) | 3873 (15.4%) | 2671 (12.6%) | 1202 (30.1%) | |
| Residential instability, N (%) | ||||
| Quintile 1 (less instability) | 6136 (24.3%) | 5451 (25.7%) | 685 (17.2%) | < 0.001 |
| Quintile 2 | 5840 (23.2%) | 5080 (23.9%) | 760 (19.0%) | |
| Quintile 3 | 5338 (21.2%) | 4501 (21.2%) | 837 (21.0%) | |
| Quintile 4 | 4812 (19.1%) | 3869 (18.2%) | 943 (23.6%) | |
| Quintile 5 (most instability) | 2975 (11.8%) | 2228 (10.5%) | 747 (18.7%) | |
| Mother’s immigration status, N (%) | ||||
| Recent immigrant | 599 (2.4%) | 455 (2.1%) | 144 (3.6%) | < 0.001 |
| Longer-term immigrant | 1505 (6.0%) | 1194 (5.6%) | 311 (7.8%) | |
| Long-term resident | 23,118 (91.7%) | 19,582 (92.2%) | 3536 (88.6%) | |
| Smoking in home, yes, N (%) | 2146 (8.5%) | 1428 (6.7%) | 718 (18.0%) | < 0.001 |
| Home ownership, yes, N (%) | 21,487 (85.2%) | 19,088 (89.9%) | 2399 (60.1%) | < 0.001 |
| Single-parent household, yes, N (%) | 3921 (15.5%) | 2553 (12.0%) | 1368 (34.3%) | < 0.001 |
| Number of children living in household, N (%) | ||||
| 1 | 5346 (21.2%) | 4476 (21.1%) | 870 (21.8%) | < 0.001 |
| 2 | 12,283 (48.7%) | 10,635 (50.1%) | 1648 (41.3%) | |
| 3 | 5528 (21.9%) | 4555 (21.5%) | 973 (24.4%) | |
| 4 + | 2065 (8.2%) | 1565 (7.4%) | 500 (12.5%) | |
| Highest level of household education, N (%) | ||||
| Less than secondary school graduation | 742 (2.9%) | 448 (2.1%) | 294 (7.4%) | < 0.001 |
| Secondary school graduation | 3237 (12.8%) | 2321 (10.9%) | 916 (23.0%) | |
| Post-secondary school | 11,212 (44.5%) | 9306 (43.8%) | 1906 (47.8%) | |
| Post-secondary school graduation | 8508 (33.7%) | 7958 (37.5%) | 550 (13.8%) | |
| Distribution of household income, provincial deciles, N (%) | ||||
| 1 (lowest) | 2173 (8.6%) | 1014 (4.8%) | 1159 (29.0%) | < 0.001 |
| 2 | 2221 (8.8%) | 1524 (7.2%) | 697 (17.5%) | |
| 3 | 2234 (8.9%) | 1638 (7.7%) | 596 (14.9%) | |
| 4 | 2444 (9.7%) | 1997 (9.4%) | 447 (11.2%) | |
| 5 | 2618 (10.4%) | 2276 (10.7%) | 342 (8.6%) | |
| 6 | 2751 (10.9%) | 2527 (11.9%) | 224 (5.6%) | |
| 7 | 2709 (10.7%) | 2549 (12.0%) | 160 (4.0%) | |
| 8 | 2500 (9.9%) | 2408 (11.3%) | 92 (2.3%) | |
| 9 | 2605 (10.3%) | 2529 (11.9%) | 76 (1.9%) | |
| 10 (highest) | 1727 (6.8%) | 1704 (8.0%) | 23 (0.6%) | |
SD, standard deviation
Cell sizes ≤ 5 are suppressed to maintain patient privacy
Table 2 illustrates the number of children who used healthcare services in the 12 months prior to the CCHS interview by food insecurity status, alongside annual public-payer healthcare costs. Children from food-insecure households more often had hospitalizations, emergency department visits, and same-day surgeries than those from food-secure households. They also used more home care, and received more publicly funded prescription medications. In contrast, children from food-secure households were more likely to use non-physician services (e.g. optometry care) as well as services associated with primary care (family physician rostering).
Table 2.
Mean per-child healthcare utilization and costs in the 12 months prior to the HFSSM by food insecurity status
| Healthcare service | Food-secure (n = 21,231) | Food-insecure (n = 3991) | p value | |||
|---|---|---|---|---|---|---|
| N (%) who used service | Mean $ ± SD a | N (%) who used service | Mean $ ± SD | N b | Cost c | |
| Inpatient hospitalizations | 398 (1.9%) | 174.99 ± 3699.38 | 113 (2.8%) | 246.95 ± 2993.25 | < 0.001 | 0.246 |
| Emergency department visits | 5832 (27.5%) | 83.76 ± 203.60 | 1329 (33.3%) | 110.30 ± 242.29 | < 0.001 | < 0.001 |
| Same-day surgeries | 482 (2.3%) | 31.71 ± 234.53 | 118 (3.0%) | 41.25 ± 256.02 | 0.009 | 0.02 |
| Outpatient clinics | 2510 (11.8%) | 110.91 ± 483.73 | 502 (12.6%) | 122.81 ± 503.39 | 0.177 | 0.156 |
| Prescription medications | 621 (2.9%) | 14.07 ± 307.95 | 592 (14.8%) | 48.87 ± 500.25 | < 0.001 | < 0.001 |
| Home care services | 532 (2.5%) | 75.73 ± 1667.08 | 133 (3.3%) | 93.02 ± 1474.31 | 0.003 | 0.541 |
| Physician billings | 14,767 (69.6%) | 201.86 ± 458.06 | 2825 (70.8%) | 240.13 ± 525.56 | 0.121 | < 0.001 |
| Lab claims | 3902 (18.4%) | 11.87 ± 36.18 | 776 (19.4%) | 13.65 ± 39.62 | 0.112 | 0.005 |
| Non-physician servicesd | 6998 (33.0%) | 19.72 ± 55.23 | 1152 (28.9%) | 18.13 ± 92.53 | < 0.001 | 0.141 |
| Family physician capitation servicese | 15,365 (72.4%) | 55.17 ± 60.19 | 2767 (69.3%) | 50.66 ± 58.67 | < 0.001 | < 0.001 |
SD, standard deviation
aAdjusted to 2020 Canadian dollars
bComparing proportion of food-secure and food-insecure individuals who used the service
cComparing mean costs between food-secure and -insecure individuals who used the service
dIncludes optometry, physiotherapy, and ambulance services
ePre-defined basket of primary care services for patients enrolled (rostered) in Family Health Organizations and Family Health Teams
Total direct healthcare costs of children from food-insecure vs. secure households aligned with care utilization patterns. In unadjusted analysis, the total direct per-child healthcare costs were $1,111.75 (95% CI: 932.95, 1324.83) in children from food-insecure vs. $837.29 (769.92, 910.57) in those from secure households, p = 0.004 (Table 3). Care costs were marginally lower when public-payer prescription medications were removed from our model ($1062.97 food-insecure vs. $823.23 secure, p = 0.011). We noted similar findings when using the 4-level measure of food insecurity (Supplementary Materials Table 6). Interestingly, we also found that healthcare costs increased by level of food insecurity until the severely food-insecure category, where costs appeared to decline slightly.
Table 3.
Unadjusted and adjusted associations between food insecurity and total healthcare costs both with and without prescription medication use, in the 12 months prior to the HFSSM by food insecurity status
| Food insecurity status | Unadjustedb | Adjustedc | ||
|---|---|---|---|---|
|
β (95% CI), p value |
Mean healthcare costs $, (95% CI)a |
β (95% CI), p value |
Mean healthcare costs $, (95% CI) |
|
| Total costs with prescription medications | ||||
| Food-secure | 0.0 (ref) |
837.29 (769.92, 910.57) |
0.0 (ref) |
563.98 (457.00, 695.99) |
| Food-insecure |
0.28 (0.09, 0.48), p = 0.004 |
1111.75 (932.95, 1324.83) |
0.18 (0.002, 0.36), p = 0.047 |
676.79 (535.26, 855.74) |
| Total costs without prescription medications | ||||
| Food-secure | 0.0 (ref) |
823.23 (757.18, 895.04) |
0.0 (ref) |
545.89 (441.61, 674.79) |
| Food-insecure |
0.26 (0.06, 0.45), p = 0.011 |
1062.97 (888.75, 1271.35) |
0.17 (–0.006, 0.352), p = 0.059 |
649.03 (512.05, 822.65) |
aAdjusted to 2020 Canadian dollars
bGamma-log–transformed generalized estimating equations accounting for the clustering of children living in the same household
cGamma-log–transformed generalized estimating equations accounting for the clustering of children living in the same household along with adjustment for child’s age (linear), sex (binary), racial belonging (categorical), mother’s immigrant status (categorical), household smoking status, (binary), single-parent status (binary), income distribution (categorical), number of children in the household (categorical), highest level of household education (categorical), and survey cycle (categorical)
When examining the adjusted relationship between food insecurity and healthcare costs, we found that total mean per-child healthcare costs were attenuated, yet remained statistically significantly higher in children from food-insecure vs. secure households ($676.79 [535.26, 855.7]) vs. $563.98 [457.00, 695.99], p = 0.0471). Results were also qualitatively similar where the costs of prescription drugs were removed (mean per-child cost of $649.03 vs. $545.89, p = 0.059). Adjusted coefficients for all models are provided in Supplementary Materials Tables 7 and 8.
Discussion
In this large study of 25,222 children from Ontario, Canada, we found that total 12-month healthcare costs for children living in food-insecure households were significantly higher than the care costs of children from food-secure households. This relationship persisted after adjustment for measurable social determinants of health. This is the first comprehensive Canadian study to investigate this association.
Higher child healthcare costs associated with food insecurity could reflect its well-described physical consequences (e.g. obesity, poor nutrition) (Berkowitz et al., 2014; Seligman et al., 2007), and/or related psychological distress (depression, anxiety) (Arenas et al., 2019; Best Start Resource Centre, 2017). Indeed, among Canadian children, household food insecurity has been associated with poorer mental health (Men et al., 2021a, 2021b, 2021c), higher healthcare utilization for mental health reasons (Anderson et al., 2023), and intentional and unintentional injuries (Men et al., 2021a).
Uniquely, we observed that some healthcare costs were higher among children living in food-secure households than in food-insecure households. These included costs associated with non-physician services and primary care. In some studies, higher socioeconomic status has been associated with more use of preventive health services (e.g. eye examinations, primary care services), even in universal healthcare settings (McMaughan et al., 2020; Olah et al., 2013). Our finding that public-payer prescription medication costs were substantially higher among children from food-insecure households is not surprising, given prescription medications were only covered for families on social assistance over the study period.
Our results align with earlier research from the United States which showed that in a cohort of 29,341 children aged 2 to 17 years, household food insecurity was associated with 25.9% more emergency department visits (Thomas et al., 2019). In an American study of the caregivers of 11,539 children under the age of 3, the odds of being hospitalized were almost a third higher in children with household food insecurity than in those with food-secure households (Chilton et al., 2009). Moreover, our findings also echo a previous study of adults in Ontario which found that total healthcare costs and mean individual healthcare costs were much higher in food-insecure vs food-secure adults even in adjusted analysis (Tarasuk et al., 2015).
Implications
The findings of our study have implications for government and policy makers. While compared with annual provincial healthcare spending, the additional care costs associated with household food insecurity were relatively low, it is important to note that we have likely underestimated the full costs of food insecurity to the province. In this study, we were only able to measure direct healthcare costs, and the indirect costs of food insecurity (e.g. disability) can also have a significant impact on the economy (Boccuzzi, 2003). Also as children can live in food-insecure households for many years, their direct care costs are likely to grow. Moreover, while not addressed in the current study, food insecurity impacts healthcare costs in adults. Thus, supportive policy at the household level (e.g. increasing the income of low-income families) is likely to have dual benefit for adults and children.
Policy actions that have been found to reduce food insecurity in Canada include federal and provincial child and family benefits, a higher minimum wage, and social assistance benefits (Men et al., 2021b). Of note, Ontario has recently expanded prescription drug coverage for children and youth who do not have access to private plans (Government of Ontario, 2017), and research on how this has impacted food insecurity status and child healthcare utilization and costs would be of interest. It is also important to emphasize that targeted interventions may be particularly important for those at high risk of food insecurity (i.e. proportionate universalism) (Carey et al., 2015), as some groups require more health action than others. For example, special support might be helpful for refugees and some new immigrants (Carney & Krause, 2020) as well as for single female parents (Sarkar et al., 2020).
Strengths of our study include access to linked administrative databases (including those which link children and mothers) with several cycles of the CCHS, a large population-based health survey. We used a standardized and validated care utilization and costing algorithm to capture a comprehensive set of healthcare costs in a publicly funded healthcare system. Household food insecurity was also assessed using a well-validated questionnaire (HFSSM). Limitations are that the CCHS is a survey and subject to response bias. We did have to exclude those who did not participate in the CCHS. Also, the CCHS does not include First Nations individuals living on reserve, a population particularly susceptible to food insecurity (Skinner et al., 2013). We also did not have data on out-of-pocket expenditures or healthcare costs covered by private insurance plans or employee benefit programs, and we could not capture costs accrued in community health centres. Our study is also only generalizable to children in Ontario who lived in a food-insecure household between 2007 and 2014. The prevalence of household food insecurity in Ontario has increased substantially since 2014 (PROOF: Food Insecurity Policy Research, 2023). Insofar as the observed associations between food insecurity and care costs have persisted, the total healthcare costs associated with this problem must also be rising.
Conclusion
This study adds to a body of literature on the health and healthcare disparities of food-insecure children. Focused efforts to support food-insecure children and their families might drive improvement in child health, and reduce costs to our Canadian healthcare system.
Contributions to knowledge
What does this study add to existing knowledge?
Food insecurity is a growing issue, and an important determinant of health and healthcare utilization in adults. We conducted a cross-sectional, population-based study to determine the association between household food insecurity and healthcare costs in children.
In our large population-based cohort of > 20,000 children, we found that even after adjusting for measurable social determinants of health, household food insecurity was associated with higher healthcare costs and care utilization among children.
What are the key implications for public health interventions, practice, or policy?
Focused efforts to support food-insecure children and their families might drive improvement in child health, and reduce costs to our Canadian healthcare system.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under license from Canada Post Corporation and Statistics Canada. We thank the Toronto Community Health Profiles Partnership for providing access to the Ontario Marginalization Index. Parts or whole of this material are based on data and/or information compiled and provided by Immigration, Refugees and Citizenship Canada (IRCC). The CCHS is adapted from Statistics Canada. This does not constitute an endorsement by Statistics Canada of this product. Parts of this material are based on data and/or information compiled and provided by the MOH, MLTC, CIHI, and IRCC. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended nor should be inferred.
Author contributions
Clemens conceptualized the study, developed the protocol, interpreted results, and drafted the manuscript. Le helped to develop the protocol and analyzed results, and revised the manuscript’s methodology and results. Anderson helped to design the protocol, planned the analysis, interpreted results, and ensured the accuracy and quality of the manuscript. Comeau helped conceptualize the study, develop the protocol, and interpret the socioeconomic analysis, and she reviewed the manuscript critically for its content. Tarasuk designed the protocol, planned analysis, interpreted project outputs, and contributed meaningfully to all parts of the manuscript. Shariff conceptualized the study, planned database linkage, developed the protocol, interpreted findings, and developed the manuscript. All the authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Funding
This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). It was conducted at ICES Western which receives funding from Western University, the Schulich School of Medicine and Dentistry, Lawson Health Research Institute and the Academic Medical Organization of Southwestern Ontario. This study also received funding from the Children’s Health Research Institute Internal Research Fund.
Availability of data, material, and codes
The dataset from this study is held securely in coded form at ICES. While legal data sharing agreements between ICES and data providers (e.g., healthcare organizations and government) prohibit ICES from making the dataset publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at www.ices.on.ca/DAS (email: das@ices.on.ca). The full dataset creation plan and underlying analytic code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.
Declarations
Conflict of interest
The authors declare no competing interests.
Ethics approval
ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze healthcare and demographic data, without consent, for health system evaluation and improvement. Use of data in this project was authorized under Sect. 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a Research Ethics Board.
Consent to participate
Not applicable.
Consent for publication
All the authors have provided consent for publication.
Disclaimer
The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended nor should be inferred.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Anderson, K. K., Clemens, K. K., Le, B., Zhang, L., Comeau, J., Tarasuk, V., & Shariff, S. Z. (2023). Household food insecurity and health service use for mental and substance use disorders among children and adolescents in Ontario, Canada. Canadian Medical Association Journal, 195(28), E948–E955. [DOI] [PMC free article] [PubMed]
- Arenas DJ, Thomas A, Wang JC, DeLisser HM. A systematic review and meta-analysis of depression, anxiety, and sleep disorders in US adults with food insecurity. Journal of General Internal Medicine. 2019;34(12):2874–2882. doi: 10.1007/S11606-019-05202-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benchimol, E. I., Smeeth, L., Guttmann, A., Harron, K., Moher, D., Petersen, I., … Langan, S. M. (2015). The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Medicine, 12(10), e1001885. 10.1371/journal.pmed.1001885 [DOI] [PMC free article] [PubMed]
- Berkowitz, S. A., Seligman, H. K., & Choudhry, N. K. (2014). Treat or eat: food insecurity, cost-related medication underuse, and unmet needs. The American Journal of Medicine, 127(4). 10.1016/J.AMJMED.2014.01.002 [DOI] [PubMed]
- Best Start Resource Centre. (2017). Child and Family Poverty in Ontario. Toronto, Ontario, Canada.
- Boccuzzi, S. (2003). Indirect health care costs. In: W. X. Weintraub (Ed.), Cardiovascular Health Care Economics. Contemporary Cardiology. Totawa, NJ: Humana Press. 10.1007/978-1-59259-398-9_5
- Carey G, Crammond B, De Leeuw E. Towards health equity: A framework for the application of proportionate universalism. International Journal for Equity in Health. 2015;14(1):1–8. doi: 10.1186/S12939-015-0207-6/FIGURES/1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carney MA, Krause KC. Immigration/migration and healthy publics: the threat of food insecurity. Palgrave Communications 2020 6:1. 2020;6(1):1–12. doi: 10.1057/s41599-020-0461-0. [DOI] [Google Scholar]
- Chilton, M., Black, M. M., Berkowitz, C., Casey, P. H., Cook, J., Cutts, D., … Frank, D. A. (2009). Food insecurity and risk of poor health among US-born children of immigrants. American Journal of Public Health, 99(3), 556. 10.2105/AJPH.2008.144394 [DOI] [PMC free article] [PubMed]
- Clemens, K. K., Le, B., Anderson, K. K., & Shariff, S. Z. (2020). Childhood food insecurity and incident diabetes: a longitudinal cohort study of 34 042 children in Ontario, Canada. Diabetic Medicine : A Journal of the British Diabetic Association, e14396. 10.1111/dme.14396 [DOI] [PubMed]
- Clemens, K. K., Le, B., Ouédraogo, A. M., MacKenzie, C., Vinegar, M., & Shariff, S. Z. (2021). Childhood food insecurity and incident asthma: a population-based cohort study of children in Ontario, Canada. PloS One, 16(6). 10.1371/JOURNAL.PONE.0252301 [DOI] [PMC free article] [PubMed]
- Cook, J. T., Black, M., Chilton, M., Cutts, D., de Cuba, S. E., Heeren, T. C., … Frank, D. A. (2013). Are food insecurity’s health impacts underestimated in the U.S. population? Marginal food security also predicts adverse health outcomes in young U.S. children and mothers. Advances in Nutrition (Bethesda, Md.), 4(1), 51–61. 10.3945/AN.112.003228 [DOI] [PMC free article] [PubMed]
- Government of Canada. (2020). The Household Food Security Survey Module (HFSSM). Retrieved September 19, 2019, from https://www.canada.ca/en/health-canada/services/food-nutrition/food-nutrition-surveillance/health-nutrition-surveys/canadian-community-health-survey-cchs/household-food-insecurity-canada-overview/household-food-security-survey-module-hfssm-health-nutrition-surveys-health-canada.html
- Government of Ontario. (2017). OHIP+: Children and Youth Pharmacare - Drugs and Devices - Health Care Professionals - MOH. Retrieved September 12, 2022, from https://www.health.gov.on.ca/en/pro/programs/drugs/ohipplus/
- Health Canada. (2006). Canadian Community Health Survey, Cycle 2.2, Nutrition (2004): a guide to accessing and interpreting the data. Retrieved August 23, 2019, from https://www.canada.ca/en/health-canada/services/food-nutrition/food-nutrition-surveillance/health-nutrition-surveys/canadian-community-health-survey-cchs/canadian-community-health-survey-cycle-2-2-nutrition-2004-guide-accessing-interpreting-data-health-ca
- Kaur J, Lamb MM, Ogden CL. The association between food insecurity and obesity in children—The National Health and Nutrition Examination Survey. Journal of the Academy of Nutrition and Dietetics. 2015;115(5):751–758. doi: 10.1016/j.jand.2015.01.003. [DOI] [PubMed] [Google Scholar]
- Kirkpatrick SI, Tarasuk V. Food insecurity is associated with nutrient inadequacies among Canadian adults and adolescents. The Journal of Nutrition. 2008 doi: 10.1093/jn/138.3.604. [DOI] [PubMed] [Google Scholar]
- Luca DL, Kwong JC, Chu A, Sander B, O’Reilly R, McGeer AJ, Bloom DE. Impact of pneumococcal vaccination on pneumonia hospitalizations and related costs in Ontario: A population-based ecological study. Clinical Infectious Diseases. 2018;66(4):541–547. doi: 10.1093/CID/CIX850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Matheson FI, Dunn JR, Smith KLW, Moineddin R, Glazier RH. Development of the Canadian Marginalization index: a new tool for the study of inequality. Canadian Journal of Public Health. 2012;103(SUPPL.2):3–5. doi: 10.17269/cjph.103.3096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McMaughan, D. J., Oloruntoba, O., & Smith, M. L. (2020). Socioeconomic status and access to healthcare: interrelated drivers for healthy aging. Frontiers in Public Health, 8. 10.3389/FPUBH.2020.00231 [DOI] [PMC free article] [PubMed]
- Men F, Elgar FJ, Tarasuk V. Food insecurity is associated with mental health problems among Canadian youth. Journal of Epidemiology and Community Health. 2021;75(8):741–748. doi: 10.1136/JECH-2020-216149. [DOI] [PubMed] [Google Scholar]
- Men, F., Urquia, M. L., & Tarasuk, V. (2021b). Examining the relationship between food insecurity and causes of injury in Canadian adults and adolescents. BMC Public Health, 21, 1557. 10.1186/S12889-021-11610-1 [DOI] [PMC free article] [PubMed]
- Men F, Urquia ML, Tarasuk V. The role of provincial social policies and economic environments in shaping food insecurity among Canadian families with children. Preventive Medicine. 2021;148:106558. doi: 10.1016/J.YPMED.2021.106558. [DOI] [PubMed] [Google Scholar]
- Olah ME, Gaisano G, Hwang SW. The effect of socioeconomic status on access to primary care: An audit study. CMAJ. 2013;185(6):E263–E269. doi: 10.1503/CMAJ.121383/-/DC1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- PROOF: Food Insecurity Policy Research. (2023). New data on household food insecurity in 2022. Retrieved May 12, 2023, from https://proof.utoronto.ca/2023/new-data-on-household-food-insecurity-in-2022/
- Rosella, L. C., Fitzpatrick, T., Wodchis, W. P., Calzavara, A., Manson, H., & Goel, V. (2014). High-cost health care users in Ontario, Canada: demographic, socio-economic, and health status characteristics. BMC Health Services Research, 14(1). 10.1186/S12913-014-0532-2 [DOI] [PMC free article] [PubMed]
- Sarkar A, Traverso-Yepez M, Gadag V. Food insecurity among single parents and seniors: A case study in an urban population in Canada. Canadian Studies in Population. 2020;47(4):263–277. doi: 10.1007/S42650-020-00037-5/TABLES/5. [DOI] [Google Scholar]
- Seligman HK, Bindman AB, Vittinghoff E, Kanaya AM, Kushel MB. Food insecurity is associated with diabetes mellitus: Results from the National Health Examination and Nutrition Examination Survey (NHANES) 1999–2002. Journal of General Internal Medicine. 2007 doi: 10.1007/s11606-007-0192-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Skinner, K., Hanning, R. M., Desjardins, E., & Tsuji, L. J. S. (2013). Giving voice to food insecurity in a remote indigenous community in subarctic Ontario, Canada: Traditional ways, ways to cope, ways forward. BMC Public Health,13(1), 1–13. https://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-13-427 [DOI] [PMC free article] [PubMed]
- Tarasuk V, Cheng J, de Oliveira C, Dachner N, Gundersen C, Kurdyak P. Association between household food insecurity and annual health care costs. CMAJ. 2015;187(14):E429–E436. doi: 10.1503/cmaj.150234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tarasuk V, Cheng J, Gundersen C, de Oliveira C, Kurdyak P. The relation between food insecurity and mental health care service utilization in Ontario. Canadian Journal of Psychiatry. 2018;63(8):557–569. doi: 10.1177/0706743717752879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thomas MMC, Miller DP, Morrissey TW. Food insecurity and child health. Pediatrics. 2019;144(4):e20190397. doi: 10.1542/peds.2019-0397. [DOI] [PubMed] [Google Scholar]
- Wodchis, W. P., Bushmeneva, K., Nikitovic, M., & Mckillop, I. (2013). Guidelines on Person-Level Costing Using Administrative Databases in Ontario, Working Paper Series, Volume 1, May 2013. Retrieved July 15, 2022 from http://hsprn.ca.
- Wong, S. (1999). Factors influencing children’s health care utilization (UC San Francisco). Retrieved February 13, 2023 from https://escholarship.org/uc/item/016130td
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset from this study is held securely in coded form at ICES. While legal data sharing agreements between ICES and data providers (e.g., healthcare organizations and government) prohibit ICES from making the dataset publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at www.ices.on.ca/DAS (email: das@ices.on.ca). The full dataset creation plan and underlying analytic code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.