Childhood food insecurity and incident asthma: A population-based cohort study of children in Ontario, Canada.

BACKGROUND: Childhood food insecurity has been associated with prevalent asthma in cross-sectional studies. Little is known about the relationship between food insecurity and incident asthma.
METHODS: We used administrative databases linked with the Canadian Community Health Survey, to conduct a retrospective cohort study of children <18 years in Ontario, Canada. Children without a previous diagnosis of asthma who had a household response to the Household Food Security Survey Module (HFSSM) were followed until March 31, 2018 for new asthma diagnoses using a validated administrative coding algorithm. We used multivariable Cox proportional hazard models to examine the association between food insecurity and incident asthma, and adjusted models sequentially for clinical and clinical/socioeconomic risk factors. As additional analyses, we examined associations by HFSSM respondent type, severity of food insecurity, and age of asthma diagnosis. Moreover, we assessed for interaction between food security and child's sex, household smoking status, and maternal asthma on the risk of incident asthma.
RESULTS: Among the 27,746 included children, 5.1% lived in food insecure households. Over a median of 8.34 years, the incidence of asthma was 7.33/1000 person-years (PY) among food insecure children and 5.91/1000 PY among food secure children (unadjusted hazard ratio [HR] 1.24, 95% CI 1.00 to 1.54, p = 0.051). In adjusted analyses associations were similar (HR 1.16, 95% CI 0.91 to 1.47, p = 0.24 adjusted for clinical risk factors, HR 1.24, 95% CI 0.97 to 1.60, p = 0.09 adjusted for clinical/socioeconomic factors). Associations did not qualitatively change by HFSSM respondent type, severity of food insecurity, and age of asthma diagnosis. There was no evidence of interaction in our models.
CONCLUSIONS: Food insecure children have numerous medical and social challenges. However, in this large population-based study, we did not observe that childhood food insecurity was associated with an increased risk of incident asthma when adjusted for important clinical and socioeconomic confounders.

Introduction

One in 6 Canadian children face hunger and dietary compromise from food insecurity (inadequate or insecure access to healthy food due to financial constraints) [1]. Where nutritious and sufficient food is crucial to the growth and development of children [2], it is essential to understand the long-term health impact of food insecurity on our youth.

Asthma is a common chronic health condition in both children and adolescents [3]. As a multifactorial disease, asthma is thought to have genetic and environmental influences [4]. Asthma has been linked with exposure to smoking, allergens [5], socioeconomic status, obesity [6], poor nutrition [7], and psychological stress [8]. An association between childhood food insecurity and prevalent asthma has also been proposed in previous observational studies. Prior studies however, have had several methodological limitations; they have been small in size, single-centered, cross-sectional in design, used unvalidated measures of food insecurity, and captured self- or parental reported health outcomes [9-11]. Moreover, studies did not fully consider important confounders that can impact food insecurity-asthma relationships including maternal health and smoking. There has additionally been limited investigation on whether childhood food insecurity is associated with new asthma diagnoses (i.e. incident disease).

In this study, we used linked health services databases, and household health survey data (Canadian Community Health Survey), to examine the association between childhood food insecurity and incident asthma in a large cohort of Canadian children. We hypothesized that there would be an independent association between food insecurity and incident asthma.

Materials and methods

Design and setting

We conducted a population-based cohort study of children <18 years living in Ontario, Canada. Ontario is Canada’s most populous province (>14 million residents). Residents have universal access to health services including hospital and physician care. Use of health services is captured by administrative codes held in secure databases available for access at ICES (formerly the Institute for Clinical Evaluative Sciences). 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. We followed the guidelines for the REporting of studies Conducted using Observational Routinely-collected Data (RECORD) (S1 Table) [12].

Data sources

We used linked ICES administrative databases and the Canadian Community Health Survey as our data sources. Datasets were linked using unique encoded identifiers and analyzed at ICES. Our project protocol (i.e. dataset creation plan) and analytic code are included in S1 and S2 Files.

A description of ICES databases is provided in S2 Table. We captured vital statistics and demographics from the Registered Persons Database of Ontario. We assessed immigration status using the Immigration, Refugees and Citizenship Canada’s Permanent Resident Database. We used the Ontario Marginalization Index (ON-MARG) database to present measures of marginalization. These measures included residential instability (e.g. living alone, living in multi-unit housing), material deprivation (e.g. low income, unemployment), dependency (e.g. age ≥65 years), and ethnic concentration (e.g. recent immigrant, visible minority) [13,14].

We also had access to a unique database (MOMBABY) which links mothers to biological children born in Ontario. MOMBABY allowed for capture of perinatal and maternal risk factors associated with asthma (e.g. maternal asthma). We abstracted additional child and maternal comorbidities using the Canadian Institute for Health Information’s Discharge Abstract Database (contains data acquired during hospital visits), and the National Ambulatory Care Reporting System Database (data on emergency department or ED visits). Finally, we used the Ontario Health Insurance Plan (OHIP) database to present additional comorbidities and health services utilization (e.g. visits to doctor’s offices, use of ED services).

The Canadian Community Health Survey (CCHS), is a national, cross-sectional survey that contains health status information, health services use, and determinants of health at the individual- and household- level [15]. We captured food security status using the CCHS’ Household Food Security Survey Module (HFSSM). The HFSSM was adapted from a long-standing food insecurity questionnaire used in the United States [16]. Details are provided in S3 Table. There are 18 questions related to household access to food over the previous 12 months; 10 of 18 questions target adult food experiences in the household (adult scale). Remaining questions focus on the experiences of children (child scale). Based upon the results of the HFSSM, children can be classified into several food security categories:

Food secure: No or one indication of difficulty with income-related food access

Moderately food insecure: Indication of some compromise in quality and/or quantity of food consumed (i.e. 2 to 4 affirmative responses)

Severely food insecure: Reduced food intake and disrupted eating patterns (i.e. ≥5 affirmative responses) [17].

Since the initial inclusion of the HFSSM in the CCHS in 2004, it has been used in 5 additional CCHS cycles that are available for use at ICES (2005, 2007–2008, 2009–2010, 2011–2012, 2013–2014). In addition to food security status, the CCHS also allowed for capture of important socioeconomic risk factors for disease including household smoking status, number of children in the home, and highest level of household education.

Cohort

Our child cohort was built using previously described methods [18]. In brief, we included children in three ways. We included: 1) children whose mothers completed the HFSSM, 2) children age 12–17 years who personally completed the HFSSM, and 3) children who were siblings of a child HFSSM respondent. We confirmed that children, siblings and mothers were alive and living in the same household when the HFSSM was completed (matched upon residential postal codes). After identifying eligible children, we excluded those with a previous diagnosis of asthma to focus on the risk of incident disease (diagnosis of asthma detailed below).

Exposure

Our primary exposure was food insecurity as defined by the child scale of the HFSSM. Given the small number of severely food insecure children in our cohort (n = 1079), we a-priori grouped both moderately and severely food insecure children into one “food insecure” category. A similar approach has been used in previous research studies on food insecurity, including our recently published report [18,19]. We defined unexposed children as those who were food secure as per the HFSSM.

Characteristics

We examined the age, racial belonging, perinatal history (caesarian section, prematurity and intrauterine growth restriction [IUGR]), comorbidities (e.g. respiratory syncytial virus), and baseline healthcare utilization (e.g. physician visits) of included children. We also captured important maternal characteristics including maternal age on child’s birthday, immigration status, and maternal comorbidities. Moreover, we present household socioeconomic characteristics including measures of marginalization (as defined by the ON-MARG database above), location of residence (urban vs. rural), household smoking status, income, number of children in the home, highest household level of education.

Primary outcome

We followed children in our cohort from the date of their first household HFSSM survey within the study period (i.e. our index date) until March 31, 2018. We captured new diagnoses of asthma using the Ontario Asthma Cohort (ASTHMA). The ASTHMA Cohort defines someone as having asthma if they have had at least one hospital admission with an asthma diagnosis, or two or more OHIP claims with an asthma diagnosis within two years. In those ≥18 years, the sensitivity of the algorithm is 80.6%, specificity 81.4%, positive predictive value 72.5%, negative predictive value 87.3% compared with medical chart review. In those <18 years, the algorithm’s sensitivity is 89% with a specificity of 72% compared with medical chart review [20]. The ASTHMA cohort has been widely used to understand the epidemiology and healthcare burden of asthma in our province [3,21,22].

As additional post-hoc outcomes, we examined the association between childhood food insecurity and asthma by HFSSM respondent type (i.e. self-responder vs. other, age of responder). We also examined the association between food insecurity and incident asthma using a 3-category exposure definition of food insecurity (i.e. secure, moderate and severely food insecure), and conducted a related trend analysis. Recognizing that asthma diagnoses at ages <3 years can be unreliable [23], we carried out an additional analysis where we restricted to asthma diagnoses made at age 3 years or older. Finally we examined for interactions between food security status and child’s sex, maternal asthma, and household smoking status on the risk of incident asthma.

Statistical analysis

We present the baseline characteristics of included children descriptively (means and standard deviations, medians and interquartile ranges, numbers and percentages). We contrast the characteristics between food secure and insecure children using t-tests and Chi-squared analyses.

The details of our time-to-event analysis have been reported previously [18]. In brief, we generated non-parametric Kaplan-Meier (KM) curves to compare the unadjusted incidence of asthma by exposure group. We then used marginal Cox proportional hazard models to adjust sequentially for clinical and clinical/socioeconomic confounders. We chose confounders based upon their known association with both food insecurity and asthma [24]. Our clinical confounders included race (white, black, other), history of prematurity and IUGR, health care utilization (visits to general practitioners/pediatricians, hospital or ED visits), maternal age at child’s birth, maternal immigration status, maternal asthma status, and household smoking status [25-29]. Socioeconomic confounders included marginalization (i.e. level of dependency, instability as defined above), location of residence, home ownership status, single parent household, income status, number of children in the home, highest level of household education [25,30-32]. Prior to including these covariates in our Cox models, we assessed for collinearity using Variance Inflation Factors and Pearson’s Correlation Coefficients.

Moreover, we examined for interactions between food security status and three variables (child’s sex, maternal asthma status, household smoking status) using log likelihood ratio tests. To preserve statistical power, we conducted post-hoc analyses using our model adjusted for clinical confounders only.

Results

Our study cohort inclusions and exclusions are provided in S5 Table. In total, there were 34,042 children with a household response to the HFSSM during the study period. After excluding 6,296 children with previously diagnosed asthma, there were 27,746 children left in our cohort (26,331 from food secure households and 1,415 from food insecure households). At the time of the index HFSSM survey, the mean age of included children was 8.7 years.

In Tables 1 and 2 we illustrate the characteristics of included children by food security status. Compared with food secure children, food insecure children were more often female, and were more commonly from ethnic minority groups. Food insecure children lived in lower income and marginalized neighbourhoods, rented housing, were from single parent families, and were more often exposed to smoking in the home. We also found that food insecure children more often had a history of IUGR, and that at baseline, they used more ED services than food secure children.

Table 1

Baseline characteristics of food insecure and secure children in Ontario, Canada.

	Food Secure	Food Insecure	P-value
	N = 26,331	N = 1,415
Demographics of children
Age at HFSSM survey date
Mean (SD)	8.66 ± 5.27	8.83 ± 5.02	0.244
Median (IQR)	9 (4–13)	9 (5–13)
0–3 years (pre-school)	5,992 (22.8%)	282 (19.9%)	0.051
4–5 years (kindergarten)	2,646 (10.0%)	144 (10.2%)
6–13 years (grade/middle school)	11,414 (43.3%)	657 (46.4%)
14–17 years (high school)	6,279 (23.8%)	332 (23.5%)
Female, N(%)	13,363 (50.8%)	759 (53.6%)	0.034
Income quintile, N(%) a
Quintile 1 (lowest)	3,763 (14.3%)	465 (32.9%)	< .001
Quintile 2	4,739 (18.0%)	322 (22.8%)
Quintile 3	5,658 (21.5%)	299 (21.1%)
Quintile 4	6,147 (23.3%)	191 (13.5%)
Quintile 5 (highest)	5,961 (22.6%)	137 (9.7%)
Measures of marginalization
Dependency, N(%)
Quintile 1 (least dependent)	5,449 (20.7%)	270 (19.1%)	0.019
Quintile 2	5,446 (20.7%)	305 (21.6%)
Quintile 3	5,250 (19.9%)	280 (19.8%)
Quintile 4	5,185 (19.7%)	248 (17.5%)
Quintile 5 (most dependent)	4,827 (18.3%)	305 (21.6%)
Deprivation, N(%)
Quintile 1 (least deprived)	6,014 (22.8%)	108 (7.6%)	< .001
Quintile 2	6,223 (23.6%)	224 (15.8%)
Quintile 3	5,483 (20.8%)	257 (18.2%)
Quintile 4	4,532 (17.2%)	293 (20.7%)
Quintile 5 (most deprived)	3,905 (14.8%)	526 (37.2%)
Ethnic Concentration, N(%)
Quintile 1 (least concentrated)	6,648 (25.2%)	355 (25.1%)	< .001
Quintile 2	6,331 (24.0%)	295 (20.8%)
Quintile 3	5,670 (21.5%)	225 (15.9%)
Quintile 4	4,256 (16.2%)	235 (16.6%)
Quintile 5 (most concentrated)	3,252 (12.4%)	298 (21.1%)
Instability, N(%)
Quintile 1 (less instability)	6,309 (24.0%)	217 (15.3%)	< .001
Quintile 2	6,231 (23.7%)	236 (16.7%)
Quintile 3	5,527 (21.0%)	291 (20.6%)
Quintile 4	5,021 (19.1%)	356 (25.2%)
Quintile 5 (most instability)	3,069 (11.7%)	308 (21.8%)
Rural location, N (%)	5,725 (21.7%)	290 (20.5%)	0.484
Ethnic Origin
European	21,291 (80.9%)	996 (70.4%)	< .001
Chinese	500 (1.9%)	22 (1.6%)	0.353
South Asian	830 (3.2%)	62 (4.4%)	0.011
Other	6,706 (25.5%)	566 (40.0%)	< .001
Racial belonging
White	20,392 (77.4%)	836 (59.1%)	< .001
Black	597 (2.3%)	124 (8.8%)	< .001
East/Southeast Asian	901 (3.4%)	46 (3.3%)	0.73
West Asian/Arab	276 (1.0%)	36 (2.5%)	< .001
South Asian	798 (3.0%)	61 (4.3%)	0.007
Latin American	201 (0.8%)	17 (1.2%)	0.069
Other	3,486 (13.2%)	322 (22.8%)	< .001
Comorbidities
Obesity	≤10	≤5	0.269
Prematurity	1,623 (6.2%)	100 (7.1%)	0.17
Intrauterine growth restriction	389 (1.5%)	33 (2.3%)	0.01
RSV	111 (0.4%)	6 (0.4%)	0.989
C-Section delivery	5,719 (21.7%)	316 (22.3%)	0.586
Health Services Utilization
Mean (SD) hospital encounters	0.08 ± 0.33	0.07 ± 0.29	0.172
Mean (SD) ED encounters	0.44 ± 0.95	0.56 ± 1.15	< .001
Mean (SD) GP visits	2.10 ± 2.77	1.97 ± 2.50	0.1
Mean (SD) Pediatrician visits	0.53 ± 1.67	0.59 ± 1.63	0.178
Mean (SD) Respirologist visits	0.00 ± 0.03	0.00 ± 0.06	0.042

Missing marginalization data was recorded as “3”.

Cell sizes <6 are not presented in accordance with ICES privacy regulations.

Abbreviations: GP, general practitioner; ED, emergency department; RSV, respiratory syncytial virus.

a Neighborhood income per person equivalent is a household size-adjusted measure of household income, based upon 2006 census summary data at the dissemination area level, using person-equivalents implied by low income cut-offs. Quintiles are defined within each area to reflect the relative nature of this measure, and to ensure that each area has about an equal percentage of the population in each income quintile [33].

Table 2

Household characteristics of food insecure and secure children in Ontario, Canada.

	Food Secure	Food Insecure	P-value
	N = 26,331	N = 1,415
Household Food Security Status
Food Secure	23,359 (88.7%)	0 (0.0%)	< .001
Moderate Food Insecurity	2,764 (10.5%)	826 (58.4%)
Severe Food Insecurity	200 (0.8%)	563 (39.8%)
Smoking in home N (%)
Yes	2,217 (8.4%)	335 (23.7%)	< .001
Unknown	7 (0.0%)	0 (0.0%)
Home ownership N (%)
Yes	22,630 (85.9%)	674 (47.6%)	< .001
Unknown	< = 70	< = 5
Single parent household N (%)
Yes	3,485 (13.2%)	604 (42.7%)	< .001
Unknown	2,025 (7.7%)	125 (8.8%)
Distribution of household income in deciles N (%) a
1 (lowest)	1,911 (7.3%)	618 (43.7%)	< .001
2	2,202 (8.4%)	260 (18.4%)
3	2,354 (8.9%)	213 (15.1%)
4	2,662 (10.1%)	119 (8.4%)
5	2,754 (10.5%)	70 (4.9%)
6	2,974 (11.3%)	49 (3.5%)
7	2,939 (11.2%)	12 (0.8%)
8	2,650 (10.1%)	13 (0.9%)
9	2,701 (10.3%)	8 (0.6%)
10 (highest)	< = 1,800	< = 5
Unknown	< = 1,390	< = 50
Number of children in household N (%)
1	5,829 (22.1%)	279 (19.7%)	< .001
2	12,711 (48.3%)	558 (39.4%)
3	5,627 (21.4%)	361 (25.5%)
4+	2,164 (8.2%)	217 (15.3%)
Highest level of household education
Less than secondary	735 (2.8%)	128 (9.0%)	< .001
Post-Secondary	3,244 (12.3%)	412 (29.1%)
Certificate	11,777 (44.7%)	644 (45.5%)
Bachelor’s Degree or higher	9,077 (34.5%)	132 (9.3%)
Unknown	1,498 (5.7%)	99 (7.0%)

Cell sizes <6 are not presented in accordance with ICES privacy regulations.

a Neighborhood income per person equivalent is a household size-adjusted measure of household income, based upon 2006 census summary data at the dissemination area level, using person-equivalents implied by low income cut-offs. Quintiles are defined within each area to reflect the relative nature of this measure, and to ensure that each area has about an equal percentage of the population in each income quintile [33].

The characteristics of the 18,270 unique mothers of the children in our cohort are presented in Table 3. Compared with the mothers of food secure children, mothers of food insecure children were younger at the time of their child’s birth, more often recent immigrants, and more frequently had asthma, obesity and diabetes than mothers of food secure children.

Table 3

Maternal characteristics of food insecure and secure children in Ontario, Canada.

	Food Secure	Food Insecure	P-value
	N = 17,349	N = 921
Age at child’s birth
Mean (SD) age (years)	29.47 ± 5.10	27.48 ± 5.86	< .001
Median (IQR) age (years)	30 (26–33)	27 (23–32)
Immigrant status a
Recent Immigrant	743 (2.8%)	69 (4.9%)	< .001
Longer-term immigrant	1,330 (5.1%)	112 (7.9%)
Long-term resident	24,258 (92.1%)	1,234 (87.2%)
Obesity	143 (0.5%)	21 (1.5%)	< .001
Diabetes	784 (3.0%)	73 (5.2%)	< .001
Charlson score b
0	25,736 (97.7%)	1,375 (97.2%)	0.374
1	304 (1.2%)	21 (1.5%)
2+	291 (1.1%)	19 (1.3%)
Asthma	3,125 (11.9%)	290 (20.5%)	< .001

a Recent immigrant is defined as a person who landed officially as permanent resident <10 years prior to the interview date. A longer-term immigrant is defined as a person who landed officially as permanent resident 10–19 years prior to interview date. A long-term resident is defined as a person who landed officially as permanent resident > = 20 years prior to interview date.

b Charlson score is a weighted measure ranging from 0–31, which captures the relative effects of 17 different health conditions and is based on ICD-10 diagnostic codes. Each disease is assigned a value, and the sum of the values produces an individual’s Charlson score. The Charlson score provides a measure of expected mortality, rather than quality-of-life related morbidity. [34].

Relationship between food insecurity and incident asthma

Children in our cohort were followed for a mean of 8.34 years (maximum duration of follow-up 13.24 year) for incident asthma. We found 1398 new diagnoses over follow-up. Children received diagnoses at a mean age of 8.90 years.

The rate of new asthma diagnoses/1000 person-years (PY) was higher in food insecure vs. secure children, but the association was not statistically significant (7.33/1000 PY vs. 5.91/1000 PY; HR 1.24, 95% CI 1.00 to 1.54, p = 0.051) (Table 4). A similar non-significant relationship was observed when adjusted for clinical (HR 1.09, 95% CI 0.85 to 1.4, p = 0.496) and clinical/socioeconomic confounders (HR 1.24, 95% CI 0.97 to 1.60, p = 0.09) (Tables 4 and S6). Of note there were no collinearity issues between any of the model covariates (Pearson’s r < |0.5| and variance inflation factor < 4).

Table 4

Association between food security status and new diagnoses of asthma.

	Secure	Insecure	P-value
Number of Children	26,331	1415	-
Median (IQR) follow-up (years)	8.34 (6.18–10.64)	8.34 (5.9–10.88)	-
New asthma diagnoses [N (%)]	1311 (4.98)	87 (6.15)	-
Rate per 1000 person-year	5.91	7.33	-
Unadjusted HR (95% CI)	Ref	1.24 (1.00 to 1.54)	0.051
Adjusted HR (95% CI) (clinical)a	Ref	1.16 (0.91 to 1.47)	0.235
Adjusted HR (95% CI) (clinical/socioeconomic)b	Ref	1.24 (0.97 to 1.60)	0.089

a Clinical confounders included race, history of prematurity, intrauterine growth restriction, visits to general practitioners/pediatricians, hospital or ED visits, maternal age at child’s birth, maternal immigration status, maternal asthma status, household smoking status.

b Clinical/socioeconomic confounders included the clinical confounders above, along with marginalization, location of residence, home ownership status, single parent household, household income, the number of children in the home, and their highest level of household education.

Abbreviations: IQR, interquartile range; HR, hazard ratio; PY, person-year.

When we examined the relationship between food insecurity and incident asthma by respondent type, there was also no statistically significant relationship observed (S7 Table). After excluding asthma diagnoses made in children <3 years, we noted similar non-significant results (HR 1.09, 95% CI 0.85 to 1.40, p = 0.479) (S8 Table). Although our analysis was limited by a small number of severely food insecure children, there did appear to be a statistically significant association between severe food insecurity and incident asthma (HR 2.32, 95% CI 1.1 to 4.9, p = 0.028) (S9 Table). However, when we conducted a related trend analysis, there was no significant association observed (HR 1.18, 95% CI 0.94 to 1.49, p = 0.132) (S10 Table). Finally, we did not find evidence of interaction between food security status and child’s sex, maternal asthma status, nor household smoking on the risk of incident asthma (p values 0.208, 0.329, 0.306 respectively).

Discussion

Food insecure children face social and health challenges including poor nutrition, poverty and obesity [35-37]. Previous reports (mainly cross-sectional) also suggest that food insecurity and prevalent asthma are related. For example, in a Canadian study of 9,142 children (10–15 years) and youth (16–21 years) who had measures of hunger ascertained (Canadian National Longitudinal Survey of Children and Youth survey), youth with repeated episodes of hunger had a higher odds of self-reported chronic conditions including asthma in adjusted analysis [9]. In a Brazilian study of 1,307 children aged 6–12 years from public elementary schools who had a response on the Brazilian food security scale, there was also a statistically significant association between asthma (defined by self-reported wheezing in the prior 12 months) and moderate and severe food insecurity (moderate food insecurity OR 1.71, 95% CI 1.01 to 2.89; severe food insecurity OR 2.51, 95% CI 1.28 to 4.93) [11]. A United States study of 11,099 3rd grade children who completed the United States Department of Agriculture (USDA) 18-item HFSSM, noted that children in food insecure households had a 4% higher adjusted odds of asthma (95% CI 1.02 to 1.06), and that the odds of asthma doubled (OR 2.00, 95% CI 1.97 to 2.03) in households that were both food-insecure and poor.(10) In a study of 6,731 children aged 13–14 who also completed the USDA 18-item HFSSM, household food insecurity in the year before kindergarten and in second grade was associated with a higher odds of parental-reported asthma (18% and 55% respectively) [38]. Food insecurity in the 2nd grade was also linked with a higher odds of asthma in the 5th and 8th grades (OR 1.55, 1.53 to 1.58).

There were several methodological limitations of these previous studies. In some studies >50% of children were lost to follow-up over time. Many did not adjust for maternal health or asthma status, perinatal risk factors, nor race, housing quality, smoking, income and education which have been linked with food insecurity [39]. It is known for example, that low income and low levels of education are linked with asthma in children [30] and that both factors are strongly associated with food insecurity.[28] There are racial disparities in asthma (black children are twice as likely as white children to have asthma),[40] and food security status. Poor housing quality has been linked with asthma, [25] and is more common in food insecure children. Maternal smoking and tobacco exposure in early life are amongst the strongest risk factors for childhood asthma, [31,32] and are also more common in food insecure children. Studies have also been cross-sectional and have not been able to illicit the association between food insecurity and incident asthma diagnoses.

In a large cohort of Ontarian children, we carefully adjusted for childhood, perinatal, maternal, and household confounders and did not observe a relationship between food insecurity and incident asthma diagnoses. To our knowledge, ours is the first large cohort study to investigate this association. Our study suggests that while food insecurity is an important determinant of health, it may not have an independent role in the pathogenesis of childhood asthma.

There are many strengths to our cohort study. We included a large number of children with universal access to healthcare and followed them for a median of 9 years. We used novel methods and unique data sources to increase the number of children in our cohort (linked children to their mothers and siblings). We focused on the risk of incident asthma, an outcome particularly important for chronic disease prevention efforts. Further, we used a validated coding algorithm for physician-diagnosed asthma, which is extremely well validated in both children and adults in our province. Moreover, we conducted this study in Canada, where children have full access to health services. This is extremely important as food insecure populations across other regions, might not have equal access to healthcare services. Additionally, although some analyses were limited by statistical power, we examined food insecurity-incident asthma relationships by HFSSM respondent type, severity of food insecurity, and age of asthma diagnoses.

Limitations of our study are that the CCHS does not include First Nations individuals, or members of the Canadian military [15]. We recognize that the CCHS is also subject to survey sampling bias, though we attempted to reduce bias by not only including children who completed the HFSSM, but those whose mothers and siblings completed the survey. Surveys are subject to response bias, and questions about food insecurity might be sensitive to answer. However, the HFSSM remains our gold standard measure of food security status in Canada [41].

Another limitation of our study is that we captured food security status at a single time point and it is possible that food security experiences may have changed over the course of follow-up. We used physician diagnosed asthma as our outcome, but recognize that we could not capture outcomes in children who did not use any health services over the study period. Although we examined for many confounders in our analyses, we could not include those that were not measurable with our data sources. These include environmental factors such as air pollution and home air quality, exposure to allergens, nutrient deficiency, breastfeeding during infancy, and psychological stress and well-being [5,7,8,42-45]. Finally our study results are only fully generalizable to children living in Ontario, Canada.

In conclusion, although children who live with food insecurity have numerous clinical and socioeconomic challenges, in this large cohort study we did not find an independent association between food insecurity and incident asthma.

RECORD checklist of recommendations for the reporting of studies conducted using routinely collected health data.

(DOCX)

Click here for additional data file.

Description of Ontario health administrative databases.

(DOCX)

Click here for additional data file.

Description of the household food security survey module.

(DOCX)

Click here for additional data file.

Study covariates.

(DOCX)

Click here for additional data file.

Participant inclusions and exclusions.

(DOCX)

Click here for additional data file.

Association between food insecurity and incident asthma, adjusted for clinical and socioeconomic confounders.

(DOCX)

Click here for additional data file.

Association between childhood food insecurity and incident asthma by respondent type, adjusted for clinical confounders.

(DOCX)

Click here for additional data file.

Association between food insecurity and incident asthma, excluding asthma diagnoses at ages <3 years, adjusted for clinical confounders.

(DOCX)

Click here for additional data file.

Association between childhood food insecurity and asthma using 3-level exposure status, adjusted for clinical confounders.

(DOCX)

Click here for additional data file.

Association between childhood food insecurity and incident asthma using 3-category exposure treated as continuous variable, adjusted for clinical confounders.

(DOCX)

Click here for additional data file.

Dataset creation plan.

(DOCX)

Click here for additional data file.

Analytic code.

(TXT)

Click here for additional data file.

7 Dec 2020

PONE-D-20-29752

Childhood food insecurity and incident asthma: a population-based cohort study of children in Ontario, Canada

PLOS ONE

Dear Dr. Clemens,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jan 16 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Maria Christine Magnus, MPH

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for submitting the above manuscript to PLOS ONE. During our internal evaluation of the manuscript, we found some minor occurrences of overlapping text with the following previous publication(s), some of which you are an author, which needs to be addressed:

- https://onlinelibrary.wiley.com/doi/abs/10.1111/dme.14396

We would like to make you aware that copying extracts from previous publications, especially outside the methods section, word-for-word is unacceptable. In addition, the reproduction of text from published reports has implications for the copyright that may apply to the publications.

Please revise the manuscript to quote or rephrase the duplicated text and cite your sources for text outside the methods section. Please note that further consideration is dependent on the submission of a manuscript that addresses these concerns about the overlap in text with published work

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. Thank you for stating the following in the Competing Interests section:

'Outside of this work, KC received a research award, sponsored in part by Astra Zeneca. She has attended Merck sponsored conferences. She has received honoraria for delivering Certified Medical Education Talks from Sutherland Global Services Canada ULC, the Toronto Ontario Knowledge Translation Working Group Inc and the Canadian Medical and Surgical Knowledge Translation Group. There are no other conflicts of interest to disclose.'

a. Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these.

Please note that we cannot proceed with consideration of your article until this information has been declared.

b. Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Study is interesting but I have some concerns related to covariates and findings.

1. It is not very clear which covariates were included in the adjusted models. How they were chosen for the final models? Not all covariates are real confounders and models might be “overfitted”. You should consider using some methods for choosing confounders, like causality graphs. For example, when comorbidities were defined and were they included in the final model? They might be consequences of food insecurity and should not be included as confounders. What are clinical confounders and socioeconomic confounders? They should be explained as footnotes in the main tables and also in the text. How covariates were grouped or removed (page 15)?

2. You report that the association between food insecurity and incidence of asthma is non-significant after adjustment for all the confounders. However, in S7 Table you report significant adjusted hazard ratio for severe food insecurity (vs. food secure) and a dose-response effect for 3-category definition of food insecurity. Why this finding is not reported as a main finding? P value for trend should be reported. It seems that the study conclusions are inconsistent with the results.

3. Please, combine T4 and T5 and explain clinical and socioeconomic confounders in the footnote.

4. In Table 1, study characteristics are presented against exposure variable and in S9 Table against outcome variable. They should be presented similar way in both tables.

5. You present results by respondent type but other possible modifying effects should also be investigated, such as effect of sex and maternal asthma. Also interactions should be investigated.

6. In the Discussion page 22, you write that your findings differ from previous studies. Please, rewrite this paragraph after considering the suggestions above.

Reviewer #2: The manuscript addresses an interesting topic, the relationship between food security and the incidence of childhood asthma. Several methodologic considerations and the overall lack of detail dampen my enthusiasm for the analysis as currently presented.

Major points:

I suggest presenting a biologic mechanism that justifies an induction period of 12 months, particularly since the majority of children had moderate food insecurity. Along those lines, I would have like to have the definition of secure, moderate and severe food insecurity.

Rationale for combining moderate and severe food insecurity in terms of asthma risk. The citations used to justify the exposure variable addressed mental health and health care costs, not asthma.

The age range includes children < 3 years old where a diagnosis of asthma is unreliable. This is a major limitation of the manuscript.

Not enough information is given on the data sources. For example – what are the coverage rates? Is it reasonable to think that families of low income status and recent immigrants would be less likely to have completed the CCHS. No information on coverage rates for any data source were provided.

Important risk factors are missing (environmental, psychosocial, housing quality, etc) and these limitations should be highlighted more. History of respiratory infection was mentioned as a covariate but it appears the variable was actually RSV specifically.

There wasn’t enough information in the manuscript describing the covariates. I find it awkward to have to go to a supplemental table to obtain variable definitions. I suggest defining the covariates in the manuscript and the data source. For example: social deprivation is very important but I can’t judge its validity due to lack of information. Many covariates that appear in the table are not mentioned in the text.

I believe there is a typo in the Results ‘26,336 food insecure’ - I think they mean secure. ‘1,414 secure’ = insecure

Table 1: for variables where quintiles are presented, at least present the data range so the reader has an idea of what the levels are.

Table 1: there’s no definition for many variables. For example – dependency, deprivation, instability, obesity.

Table 2. No need to include both yes and no. Same comment as above for income. Added category of moderate food insecure which wasn’t mentioned in the exposure section. Overall, there are too many categories for several of the variables.

Table 3. Present units: like age in years, etc. There is no definition for immigrant status, charlson score.

Table 5. Tables should stand alone. What are the models adjusted for?

The section of prevalent asthma lacks any detail – most of the information is in the supplemental table. This is difficult to evaluate. I suggest including the essential information or, if there is not room, to remove it.

The literature review lacks consistent detail on all studies cited: add age range and location of all studies. Comparing this study, with an age range of <18 to other studies is problematic and the authors should discuss these differences in depth. Editorial suggestion: include age ranges, not just the grade level as this may change between countries. Also there’s no mention of how the exposure variables are similar or different between studies making it hard to decide if these studies are at all comparable.

Given the weaknesses of the study, I believe the conclusions are overstated.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

12 Apr 2021

Editor Comments

Comment 1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response 1. Thank you. We have re-reviewed PLOS ONE’s style requirements and have updated our title page and manuscript as suggested.

Comment 2. Thank you for submitting the above manuscript to PLOS ONE. During our internal evaluation of the manuscript, we found some minor occurrences of overlapping text with the following previous publication(s), some of which you are an author, which needs to be addressed:

- https://onlinelibrary.wiley.com/doi/abs/10.1111/dme.14396

We would like to make you aware that copying extracts from previous publications, especially outside the methods section, word-for-word is unacceptable. In addition, the reproduction of text from published reports has implications for the copyright that may apply to the publications.

Please revise the manuscript to quote or rephrase the duplicated text and cite your sources for text outside the methods section. Please note that further consideration is dependent on the submission of a manuscript that addresses these concerns about the overlap in text with published work

Response 2. We apologize for this. Both the current paper and our recently published manuscript were written at the same time, and upon re-review we completely agree that there is overlap in some sections.

We have re-reviewed our manuscript in detail and have rephrased our prose throughout. In the Methods section, we now cite our recently published Diabetic Medicine paper, and have shorted our description of our cohort build and analytic methods. Please note that some of the phrasing used throughout our paper, does have to be written verbatim as per ICES policies.

Comment 3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

Response 3. There are legal and ethical restrictions that preclude us from sharing our data publicly.

In S1 Table, we report: “The dataset from this study is held securely in coded form at ICES. While data sharing agreements prohibit ICES from making the dataset publicly available, access can be granted to those who meet pre-specified criteria for confidential access, available at https://www.ices.on.ca/DAS.The full data set creation plan and underlying analytic code are available in Supporting Information 2 and 3.”

Comment 4: In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

Response 4: In our revised cover letter, we have detailed the ethical and legal restrictions upon sharing de-identified ICES data.

Comment 5. Thank you for stating the following in the Competing Interests section:

'Outside of this work, KC received a research award, sponsored in part by Astra Zeneca. She has attended Merck sponsored conferences. She has received honoraria for delivering Certified Medical Education Talks from Sutherland Global Services Canada ULC, the Toronto Ontario Knowledge Translation Working Group Inc and the Canadian Medical and Surgical Knowledge Translation Group. There are no other conflicts of interest to disclose.'

a. Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

b. Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Response 5: We have provided our competing interests statement in our cover letter and have reported that “this does not alter our adherence to PLOS ONE policies on sharing data and materials”.

Reviewer comments

Reviewer #1: Study is interesting but I have some concerns related to covariates and findings.

Comment 1. It is not very clear which covariates were included in the adjusted models. How they were chosen for the final models? Not all covariates are real confounders and models might be “overfitted”. You should consider using some methods for choosing confounders, like causality graphs. For example, when comorbidities were defined and were they included in the final model? They might be consequences of food insecurity and should not be included as confounders. What are clinical confounders and socioeconomic confounders? They should be explained as footnotes in the main tables and also in the text. How covariates were grouped or removed (page 15)?

Response 1. Thank you. We conducted a thorough literature review on food insecurity and asthma and chose our covariates based upon their association with both our exposure (food insecurity) and our outcome (i.e. confounders). The covariates chosen also align with the variables used in our previously published paper. Based upon our review, we do not believe that covariates are direct consequences of food security alone, without also being associated with our exposure.

In the Methods section on page 10, we now explicitly present the covariates used in each model. In S4 Table, more details are provided on how each confounder was treated and what our referent groups were. Prior to including covariates in our models, we examined for collinearity using Pearson r test and the VIF test. There were no collinearity issues found and thus no variables were removed. We have reported this in our Methods section on page 10, and in our Results section on page 15.

In our Tables, Supplementary Tables and footnotes, we have described how we grouped covariates

In our new Table 4, we have also listed all of the clinical and socioeconomic confounders we adjusted for as footnotes.

Comment 2. You report that the association between food insecurity and incidence of asthma is non-significant after adjustment for all the confounders. However, in S7 Table you report significant adjusted hazard ratio for severe food insecurity (vs. food secure) and a dose-response effect for 3-category definition of food insecurity. Why this finding is not reported as a main finding? P value for trend should be reported. It seems that the study conclusions are inconsistent with the results.

Response 2. This was a post-hoc exploratory analysis, and was unfortunately limited by a small number of severely food insecure children and events. However, in the revised manuscript, we have reported this result more explicitly on page 15. We have now also conducted a related trend analysis and added this to our results on page 15 and S10 Table of the Supplemental Material. The result of our trend analysis was consistent with or main results – that there was not a statistically significant relationship between childhood food insecurity and incident asthma.

Comment 3. Please, combine T4 and T5 and explain clinical and socioeconomic confounders in the footnote.

Response 3. We have now combined Tables 4 and 5 (new Table 4) and have included a description of all clinical and socioeconomic confounders in the footnotes.

Comment 4. In Table 1, study characteristics are presented against exposure variable and in S9 Table against outcome variable. They should be presented similar way in both tables.

Response 4. Our prevalent asthma analysis has been removed from this manuscript (per Reviewer 2’s comment). As such, S9 Table has been deleted from the resubmission.

Comment 5. You present results by respondent type but other possible modifying effects should also be investigated, such as effect of sex and maternal asthma. Also interactions should be investigated.

Response 5. We have now assessed for interaction between food insecurity and maternal asthma, child’s sex and smoking in the home. We have described our methods on page 10. We have presented results on page 15. The log-likelihood ratio test suggested no evidence of interaction. As such there were no interaction terms included in our models.

Comment 6. In the Discussion page 22, you write that your findings differ from previous studies. Please, rewrite this paragraph after considering the suggestions above.

Response 6. We have re-written our Discussion to focus upon included results and have discussed study implications, strengths and limitations.

Reviewer #2: The manuscript addresses an interesting topic, the relationship between food security and the incidence of childhood asthma. Several methodologic considerations and the overall lack of detail dampen my enthusiasm for the analysis as currently presented.

Comment 1. I suggest presenting a biologic mechanism that justifies an induction period of 12 months, particularly since the majority of children had moderate food insecurity. Along those lines, I would have like to have the definition of secure, moderate and severe food insecurity.

Response 1. In the revised draft, we have moved our definitions of food security status to the main manuscript (Methods page 7).

We used a 12 month look-back for food insecurity status as this is the look-back period captured by the HFSSM. Both the CCHS’ HFSSM and the USDA’s HFSSM assess food security status in this fashion. We recognize that capturing food security status cross-sectionally is a limitation, and that the food security status could have changed as we followed our cohort over time. We have described this limitation in our Discussion on page 18.

Comment 2. Rationale for combining moderate and severe food insecurity in terms of asthma risk. The citations used to justify the exposure variable addressed mental health and health care costs, not asthma.

Response 2. There were only 1079 severely food insecure children in our cohort. Because we anticipated a limited number of outcomes, we elected a-priori combine moderate and severely food insecure children into one category (food insecure). We provided the mental health/healthcare costs paper as a reference, as authors grouped exposure categories similarly.

Of note, we did conduct additional analyses in the current paper, where we examined relationships using a 3-category definition of food insecurity (Table S9-S10) and the results of our analyses were similar to our main findings.

Comment 3. The age range includes children < 3 years old where a diagnosis of asthma is unreliable. This is a major limitation of the manuscript.

Response 3. Thank you for this important comment. As a post-hoc sensitivity analyses, we examined associations where asthma diagnoses at ages <3 years were excluded. We found results similar to our primary analysis. We report our methods on page 10, and our results on page 15 and in Table S8.

Comment 4. Not enough information is given on the data sources. For example – what are the coverage rates? Is it reasonable to think that families of low income status and recent immigrants would be less likely to have completed the CCHS. No information on coverage rates for any data source were provided.

Response 4. Given our provinces universal healthcare system, all residents with an Ontario healthcard who sought healthcare are included in ICES databases. We do recognize that those who did not receive healthcare will not be captured in this study and have commented upon this in our Discussion on page 19.

We also recognize that we had to exclude children who did not have a household response to the HFSSM (numbers provided in S5 Table). We also comment in our discussion that there may be sampling bias in our study due to use of the CCHS survey. However, we attempted to reduce the risk of sampling by not only including children who completed the CCHS, but also children of mothers who completed the CCHS. These points have been discussed in detail on page 18-19 of the Discussion.

Comment 5. Important risk factors are missing (environmental, psychosocial, housing quality, etc.) and these limitations should be highlighted more. History of respiratory infection was mentioned as a covariate but it appears the variable was actually RSV specifically.

Response 5. Thank you. We recognize that asthma is a multifactorial disease with clinical, environmental, and socioeconomic risk factors. We captured as many as able (maternal comorbidities, perinatal history, child comorbidities, SES, smoking, household characteristics), but recognize that some were not “measurable” for us to adjust for. We have expanded upon your point in our Discussion on page 19.

You are indeed correct, we captured RSV infection rather than respiratory infection and have corrected this in our Methods section on page 8.

Comment 6. There wasn’t enough information in the manuscript describing the covariates. I find it awkward to have to go to a supplemental table to obtain variable definitions. I suggest defining the covariates in the manuscript and the data source. For example: social deprivation is very important but I can’t judge its validity due to lack of information. Many covariates that appear in the table are not mentioned in the text.

Response 6. We have now provided the details of all covariates early in the Methods section on page 8. Those included in our adjusted models are reported on page 9-10. We have also included covariates included in each model in the footnotes of Table 4. We still present S4 Table as it might help readers to understand how we treated each covariate and what our referent group was for each.

Comment 7. I believe there is a typo in the Results ‘26,336 food insecure’ - I think they mean secure. ‘1,414 secure’ = insecure

Response 7. We apologize for this typo and have corrected this.

Comment 8. Tables

Table 1: for variables where quintiles are presented, at least present the data range so the reader has an idea of what the levels are.

Table 1: there’s no definition for many variables. For example – dependency, deprivation, instability, obesity.

Table 2. No need to include both yes and no. Same comment as above for income. Added category of moderate food insecure which wasn’t mentioned in the exposure section. Overall, there are too many categories for several of the variables.

Table 3. Present units: like age in years, etc. There is no definition for immigrant status, charlson score.

Table 5. Tables should stand alone. What are the models adjusted for?

Response 8. We have updated all Tables based upon your helpful feedback. Unfortunately we are not able to provide a range for quintiles. This is because quintiles differ by neighborhood, and so a quintile of 5 can be very different in Toronto vs. Windsor. We have included an explanatory footnote in Tables 1 and 2.

In the Methods Section we provide more details on the Ontario Marginalization Database including what measures of dependency, deprivation etc. mean.

We have included yes rather than both yes and no.

We have removed the moderate food insecurity category in Table 2.

We have presented units like age in years.

We have also provided definitions for immigrant status and Charlson score as footnotes in Table 3.

We have merged Tables 4 and 5 as suggested and in the footnotes have reported what the models were adjusted for.

Comment 9. The section of prevalent asthma lacks any detail – most of the information is in the supplemental table. This is difficult to evaluate. I suggest including the essential information or, if there is not room, to remove it.

Response 9. We have elected to remove our prevalent asthma analysis from this paper.

Comment 10. The literature review lacks consistent detail on all studies cited: add age range and location of all studies. Comparing this study, with an age range of <18 to other studies is problematic and the authors should discuss these differences in depth. Editorial suggestion: include age ranges, not just the grade level as this may change between countries. Also there’s no mention of how the exposure variables are similar or different between studies making it hard to decide if these studies are at all comparable.

Response 10. We have re-written this section for consistency and have added more detail to help to illustrate the comparability of studies.

Submitted filename: Response to reviewers (2).docx

Click here for additional data file.

14 May 2021

Childhood food insecurity and incident asthma: a population-based cohort study of children in Ontario, Canada

PONE-D-20-29752R1

Dear Dr. Clemens,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Maria Christine Magnus, MPH

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for your response. I think you addressed all my comments and I have no further comments to add.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

20 May 2021

PONE-D-20-29752R1

Childhood food insecurity and incident asthma: a population-based cohort study of children in Ontario, Canada

Dear Dr. Clemens:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Maria Christine Magnus

Academic Editor

PLOS ONE