Prevalence of food insecurity among caregivers of young children during the COVID-19 pandemic in Ireland
Abstract
Our objectives were to explore the prevalence of food insecurity in primary caregivers of young children in Ireland and to compare two tools for measuring food insecurity during the COVID-19 crisis. A cross-sectional online survey was conducted among caregivers of children aged <2 years in Ireland in May/June 2020. Relevant survey questions were closed-ended using two established tools for measuring food poverty/insecurity; the Irish Food Poverty Indicator and the Food Insecurity Experience Scale, developed by the Food and Agriculture Organization. Descriptive statistics were used to calculate the prevalence of food poverty/insecurity. To explore agreement in the classification of food insecurity by the Food Poverty Indicator and the Food Insecurity Experience Scale, Cohen's κ was used. Analyses are based on 716 participants; most had a 3rd-level education and were married or in a partnership. Per the food poverty index, Ireland's national measure of food poverty, 3.9% (n = 28) of our sample were experiencing food poverty. This rose to 10.5% (n = 75) experiencing food insecurity when using the Food Insecurity Experience Scale, which also measures worry/anxiety around access to food. There was low agreement between the tools, with 11.3% of the sample classified as food secure by one tool and food insecure by the other. Our current measure of food poverty in Ireland may not be sufficient to describe the food-access struggles or worry/anxiety about food access, experienced by the population, particularly during an emergency like COVID-19.
INTRODUCTION
Food security has been defined as a situation that exists when “all people, at all times, have physical and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for an active and healthy life” (Food and Agriculture Organization, 1996). This definition was expanded in 2002 to incorporate social access to food (Food and Agriculture Organization, 2002), acknowledging its important role within the concept of food security. However, other definitions and terms are used globally and there is a lack of consensus on a standard measurement tool for food insecurity, creating challenges in academia and public health policies (O'Connor et al., 2016). Some measurement tools focus on the absolute lack of access to safe or nutritious food, while others also consider concerns or fears around access to safe or nutritious food. The most severe manifestation of food insecurity is when someone has insufficient resources to meet all basic needs and thus skips meals and feels hungry for extended periods (Hendriks, 2015; Radimer et al., 1992; Saint Ville et al., 2019). The term food insecurity is often used interchangeably with food poverty, particularly in Ireland, although they are not exactly the same.
There is no direct measurement tool for determining the prevalence of food insecurity in Ireland. However, the percentage of the population experiencing food poverty has been reported annually since 2014 (Department of Social Protection, 2022b). This Irish-based food poverty measure was proposed by Carney and Maître in 2012 (Carney & Maître, 2012). They developed a composite indicator using questions relating to food on the Survey on Income and Living Conditions (SILC), carried out annually by the Irish Central Statistics Office (Central Statistics Office 2023). Questions related to food deprivation have been included in the SILC since it began in 2004, but a composite food poverty indicator had not been developed until Carney and Maître's work in 2012. They developed a 3-item indicator, with each question being dichotomous in response. This indicator includes statements relating to food quality and quantity but does not measure anxiety around food procurement or social acceptability of the sources of food, which some consider a limitation (Healy, 2019). SILC data are currently collected from one household member aged 16 years or over from a sampling frame of all private households. Combining the relevant three food deprivation indicators in this survey into the Food Poverty Indicator (FPI) allows for the classification of respondents in a binary manner as either experiencing food poverty or not.
The proportion of the Irish population experiencing food poverty, as measured by the FPI, has gone from 13.1% in 2014 to 8.7% in 2017 then to a low of 7% in 2018, rising again to 12% in 2020 with the latest figure being 8.9% in 2021 (Department of Social Protection, 2022b). Research conducted in 2022 by an independent research and insight agency exploring the impact of the cost-of-living crisis gives us some insight into the present situation (Brennan & Williams, 2022). Among the 1005 respondents to a survey in September 2022, 30% reported that they had less money available to spend on food (Brennan & Williams, 2022), and of the 920 who reported their cost of living had increased, over 90% reported that the cost of their food shop had increased. This suggests that the issue of food poverty could be getting worse.
Experience-based food insecurity scales are useful as they can capture social aspects of food insecurity that other measures cannot (Frongillo, 1999). In 2013, the FAO developed a global instrument for measuring food insecurity, the Food Insecurity Experience Scale (FIES), as part of their Voices of the Hungry project (Ballard et al., 2013). The FIES indicator is a household experience-based, 8-item questionnaire. It is largely based upon Radimer and colleagues' 1992 research, focusing on three levels of food insecurity: mild food insecurity where there is anxiety around decreasing choices with expected lack of food in the future, followed by deteriorating quality and reduced diversity of foods eaten (moderate food insecurity) and finally severe food insecurity with reduced food quantities consumed (Radimer et al., 1992). The eight dichotomous (Yes/No) questions are listed in ascending order of extremity of food insecurity; the first asks about uncertainty about food obtainment, the following two about inadequate food quality and the final five about insufficient food quantity.
A cross-sectional study in 2018 by Ahmadi and Melgar-Quiñonez measured food security in Ireland using the FIES and compared yearly results from 2014 to 2017. Food security status improved over this time, with 77.6% of respondents reportedly food secure in 2014 and 89.7% in 2017 (Ahmadi & Melgar-Quiñonez, 2018). Although it is positive that food security status increased over this time, concern remains over the 10.3% of the population who were food insecure in 2017, a figure higher than the 8.7% reported by the FPI for the same year (Department of Social Protection, 2022b).
There is currently no national measure to assess food poverty or food insecurity specifically among children. This is problematic because living in a food-insecure household has been associated with poor health-related quality of life among children aged 3–17 years (Casey et al., 2005) and poor language development at 18 months (Saha et al., 2010). Children are particularly vulnerable to food poverty and two recent Irish surveys completed on behalf of Barnardos (a children's charity) suggest that food poverty may be increasing among families with children in the household (Amárach Research, 2022; Coyne Research, 2023), with parents worried about having enough to feed their children and skipping meals to ensure food is available for their children. These issues have been highlighted by the Children's Rights Alliance in their 2023 Child Poverty Monitor (Children's Rights Alliance, 2023). Unfortunately, national data specifically on child food poverty are not currently being reported, as highlighted by Vazquez Mendoza and McDonagh (Vazquez Mendoza & McDonagh, 2022). A need for a reliable measure of food insecurity among children in the United Kingdom, and the periodic monitoring of the same, has also been highlighted (Aceves Martins et al., 2018).
Food insecurity is complex in its definition and assessment. The public health effects of food insecurity are significant, and this is particularly so for parents who may compromise their nutritional intake for the benefit of their children. Here, we describe the prevalence of food insecurity among a sample of caregivers of infants under the age of 2 years during the COVID-19 restrictions in Ireland. We also compare two tools for measuring food insecurity to explore whether there are differences in the prevalence of food insecurity depending on which tool is used.
METHODS
Study design, participants and data collection
Participants in this study were invited to complete a cross-sectional, self-administered questionnaire online. The primary focus of the questionnaire was to understand the experiences of caregivers of infants and young children (under 2 years of age) with feeding their young children during the COVID-19 pandemic in Ireland (i.e., infant and young child feeding during a national emergency). The focus was on parents with infants under 2 years of age as infants and young children are a particularly vulnerable subgroup due to their specific nutrient needs, and their need to eat food – potentially of modified consistency – more frequently. There were questions about demographic characteristics, breastfeeding, formula feeding and food security. A description of families' experiences of breastfeeding and formula feeding during the pandemic has been published elsewhere (O'Sullivan & Kennedy, 2023). Food security, which is the focus of this manuscript, was the last topic on the questionnaire.
The questionnaire was available online from May 11, 2020, for 1 month, on the SurveyMonkey platform. The questionnaire link was largely distributed through social media, particularly Twitter and Facebook, making this a convenience sample. The admin teams of various parenting support groups’ social media pages were contacted and asked to advertise the survey to their members. Those interested were eligible to participate if they were adults (at least 18 years old) living in Ireland and were the primary caregivers of a child under the age of 2 years. Additionally, it was necessary that they had sufficient English to give valid consent and fully complete the questionnaire.
Tools for measuring food poverty and food insecurity
Two tools were used in this research: the Irish deprivation-based measure of food poverty, the Food Poverty Index (FPI) developed in 2012 (Carney & Maître, 2012) and the Food Insecurity Experience Scale (FIES) developed by the Food and Agriculture Organization (Ballard et al., 2014).
Questions from the FPI appeared first on the present questionnaire; it is a 3-item questionnaire, dichotomous in response and the respondent is asked if the three statements apply to them. This tool includes indicators of food quality and quantity but does not measure anxiety or uncertainty around food procurement or the social acceptability of the sources of food. A participant is classified as food-deprived if they respond that any of the following apply to them: “I cannot afford a roast dinner or vegetarian equivalent once a week,” “I cannot afford a meal with meat or vegetarian equivalent every second day,” and “I missed a meal in the last two weeks due to a lack of money.” Participants are considered not food deprived if none of those statements apply to them.
The SILC, which includes the FPI questions, is typically conducted either in person or over the phone, so an investigator asks respondents the questions. For the present online, self-administered questionnaire, participants were asked to “Please indicate which, if any, of the following statements apply to you at the present time” (i.e., at the time they were completing the questionnaire). Response options provided were “This does apply to me” and “This does not apply to me.” “This does apply to me” (i.e., affirmative) was the left radio button and “This does not apply to me” (i.e., negative) was the right radio button (Box 1). These response options were chosen so that the original statement would not have to be modified for online administration.
BOX 1. Food Poverty Indicator, as administered on the present survey.
Please indicate which, if any, of the following statements apply to you at the present time:
| This does apply to me | This does not apply to me | |
| I cannot afford a roast dinner or vegetarian equivalent once a week | ◎ | ◎ |
| I missed a meal in the last 2 weeks due to a lack of money | ◎ | ◎ |
| I cannot afford a meal with meat or vegetarian equivalent every second day | ◎ | ◎ |
The FIES questionnaire is an experience-based, 8-item questionnaire about the experiences of those in the index household. It categorises respondents/households into food secure or to three levels of food insecurity: mild, moderate, and severe. Participants are presented with a “stem” for the questions, which, in the published tool is “During the last 12 months, was there a time when, because of lack of money or other resources:” To ascertain experiences of food insecurity related to COVID-19, we altered this question stem to “During the coronavirus crisis, was there a time when:” The eight dichotomous (Yes/No) questions that followed the stem ask about household's experiences of food insecurity and are listed in ascending order of severity; the first asks about uncertainty about food obtainment, the following two about inadequate food quality and the final five about insufficient food quantity (Box 2). All specify that the problem is caused by a lack of money or other resources. A negative response to all questions indicates a household is food secure, responding “yes” to 1–3 questions leads to a classification of mild food insecurity, 4–6 is considered moderate and 7–8 is considered severe food insecurity. Response options were “yes,” “no,” “don't know,” and “don't want to answer” in that order, from left to right. A dichotomous variable was later created using the FIES data to define households as food secure (sum score of 0 across all 8 items) or food insecure (sum score of 1–8 across all 8 items) to compare with the FPI.
BOX 2. The Food Insecurity Experience Scale, as administered in the present survey.
During the coronavirus crisis, was there a time when:
| Yes | No | Don't know | Don't want to answer | |
| You or others in your household worried about not having enough food to eat because of lack of money or other resources. | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household were unable to eat healthy and nutritious food because of lack of money or other resources? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household ate only a few kinds of foods because of lack of money or other resources to get food? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household had to skip a meal because there was not enough money or other resources to get food? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household ate less than you thought you should because of a lack of money or other resources? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when your household ran out of food because of a lack of money or other resources? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household were hungry but did not eat because there was not enough money or other resources for food? | ◎ | ◎ | ◎ | ◎ |
| Was there a time when you or others in your household went without eating for a whole day because of a lack of money or other resources? | ◎ | ◎ | ◎ | ◎ |
Data analysis
The prevalence of food insecurity, using both tools, was determined using descriptive statistics with prevalence expressed as n (%). To explore agreement in the classification of food insecurity by the FPI and FIES, Cohen's κ was used. Continuous data are expressed as mean ± standard deviation (SD), and categorical data are expressed as n (valid percent). Statistical analyses were performed using IBM SPSS Statistics, version 26.0.
Demographic data collected
Age was measured continuously in years. All other sociodemographic characteristics were categorical. Geographical location was categorised as “inner city,” “suburban, “regional town,” “country village/countryside.” The question relating to education contained the following categories: primary, secondary, post-leaving certificate course, third-level, other. Education status was later dichotomised to “secondary/post-leaving certificate” and “third level and above” due to the small number of participants with lower levels of education. Marital status was dichotomised to “married or have a partner” and “single/separated/divorced” due to small numbers of participants in individual categories. Current work status was categorised as “working from home,” “working outside the home,” “caring for children at home,” “maternity leave,” “not working,” and “other.” Whether the respondent had a medical card was answered with options of “yes,” “no,” and “don't want to answer.” In Ireland, a medical card allows you to get certain health services free of charge. People must apply for a medical card, and it is means-tested, with only those below a certain income threshold being eligible. Questions about whether respondents had lost their jobs due to the COVID-19 pandemic and whether they were receiving the emergency pandemic unemployment payment were answered with options of “yes,” “no,” and “not applicable.” Responses of “don't know” and “don't want to answer” were set to missing for analysis. There were very little missing demographic data, with ≤7 responses missing for some characteristics.
Data cleaning
Data were interrogated for outliers through the running of descriptive statistics and familiarisation with the quantitative data. During this process, it was discovered that there were 20 respondents who responded in the affirmative to all three statements on the FPI and in the negative to all eight statements on the FIES. Similarly, there was one participant who responded in the affirmative to all eight statements on the FIES and in the negative to all three statements on the FPI. This was considered by the authors to be a potential error. Some of these participants also provided qualitative data suggesting that they were not experiencing food insecurity. As such, these respondents were excluded from the quantitative analysis exploring the prevalence of food insecurity and the comparison of the two food insecurity measurement tools.
Thirty participants were excluded due to missing FIES information and three were excluded for not fully completing the FPI. One other participant was excluded as they indicated early in the questionnaire that they were not the primary caregiver of a child under 2 years of age. Finally, the 21 participants described in the data cleaning phase above were excluded for quantitative analysis.
RESULTS
There were 771 responses to the questionnaire. After the exclusions described above, 716 participants were included in the statistical analysis. The mean age of the sample was 35 years (Table 1). The majority (n = 712) were mothers to the child under 2 years. Most of the respondents had a minimum of a third-level education and were married or had a partner. A small number of participants (n = 32) had lost their jobs due to the COVID-19 pandemic and 29 were receiving an emergency pandemic unemployment payment from the Irish government. Fewer than 5% of the respondents reported having a medical card.
| Variable | Frequency, n (valid %), except where indicated |
|---|---|
| Age, years | 35 ± 4a |
| Relationship to index child | |
| Mother | 712 (99.4) |
| Father | 4 (0.6) |
| Geographic location | |
| Inner city | 52 (7.3) |
| Suburban | 306 (42.7) |
| Regional town | 160 (22.4) |
| Country village/countryside | 198 (27.6) |
| Educationb | |
| Secondary/post-leaving certificate | 45 (6.3) |
| Third-level or greater | 668 (93.7) |
| Marital statusb | |
| Married or have a partner | 704 (98.3) |
| Single/separated/divorced | 11 (1.6) |
| Current work status | |
| Working from home | 157 (21.9) |
| Working outside the home | 53 (7.4) |
| Caring for children at home | 68 (9.5) |
| Maternity/parental leave | 367 (51.2) |
| Not working | 62 (8.7) |
| Other | 9 (1.3) |
| Medical card holder | |
| Yes | 35 (4.9) |
| No | 677 (94.6) |
| Don't want to answer | 4 (0.6) |
| Lost job due to COVID-19 | |
| Yes | 32 (4.5) |
| No | 154 (21.5) |
| N/A | 519 (72.5) |
| Other | 11 (1.5) |
| Receiving emergency pandemic payment | |
| Yes | 29 (4.1) |
| No | 150 (20.9) |
| N/A | 537 (75) |
- a Continuous data are expressed as mean ± standard deviation (SD).
- b Some missing data so valid % reported.
Prevalence of food insecurity
Almost 90% of households were considered food secure by the FIES, with a raw score of 0 (Table 2). The proportion of households categorised as mildly, moderately or severely food insecure was 7.9%, 2.5% and 0.1%, respectively (total n = 75). A higher proportion of participants (96.1%) were classified as not experiencing food poverty by the FPI. A total of 28 (3.9%) participants responded positively to one, two or three of the FPI statements, indicating that they were experiencing food poverty (Table 3). Thus, depending on the tool used, the prevalence of food insecurity in the sample was between 3.9% and 10.5%.
| Raw score | Frequency of raw scores in the population, n (%) | Classification of food insecurity | Prevalence of food insecurity, n (%) |
|---|---|---|---|
| 0 | 641 (89.5) | Secure | 641 (89.5) |
| 1 | 39 (5.4) | Mild | 57 (7.9) |
| 2 | 11 (1.5) | ||
| 3 | 7 (1) | ||
| 4 | 9 (1.3) | Moderate | 17 (2.5) |
| 5 | 4 (0.6) | ||
| 6 | 4 (0.6) | ||
| 7 | 0 (0) | Severe | 1 (0.1) |
| 8 | 1 (0.1) |
| Sum of ‘yes’ responses to SILC FPI | Frequency in the population, n (%) | Classification of food poverty | Prevalence of food poverty, n (%) |
|---|---|---|---|
| 0 | 688 (96.1) | Not experiencing food poverty | 688 (96.1) |
| 1 | 22 (3) | Experiencing food poverty | 28 (3.9) |
| 2 | 4 (0.6) | ||
| 3 | 2 (0.3) |
Differences between the two indicators of food insecurity
For the purposes of comparison between the dichotomous FPI and categorical FIES, all levels of food insecurity within the FIES were combined. Thus, we compared those who were food secure or insecure as measured by the FIES – which includes a dimension focusing on anxiety or worry about access to food – and those who were not experiencing food poverty and those who were experiencing food poverty, as measured by the FPI. For the sake of comparison between the tools, the terms food security and insecurity are used throughout the remainder of the manuscript.
There was low agreement between the FIES and FPI; both tools identified 87.2% of the sample (n = 624) as food secure and 1.5% (n = 11) of the sample as food insecure, but there was disagreement for 11.3% of the sample (n = 81) with one tool classifying a participant as food secure and the other classifying them as food insecure (Table 4). The measure of agreement was low (Cohen's κ = 0.166, p < 0.001).
| FIES; insecure, n (%) | FIES; secure, n (%) | Row total, n (%) | |
|---|---|---|---|
| FPI; insecure, n (%) | 11 (1.5) | 17 (2.4) | 28 (3.9) |
| FPI; secure, n (%) | 64 (8.9) | 624 (87.2) | 689 (96.1) |
| Column total, n (%) | 75 (10.5) | 641 (89.5) |
Further interrogation of the data was conducted to understand the reason for the disagreement between the FPI and FIES. Seventeen participants who were classified as food insecure by the FPI, but their households were classified as food secure by the FIES (Table 4); of these, 15 responded in the affirmative to the statement “I cannot afford a roast dinner or vegetarian equivalent once a week” on the FPI.
Of the 64 participants whose households were classified as food insecure by the FIES but classified as food secure by the FPI (Table 4), 39 responded “yes” to just one of the FIES statements, 11 responded “yes” to two, and six responded “yes” to three; so, 56 were considered mildly food insecure by the FIES but food secure by the FPI. Of those who responded “yes” to just one FIES statement, 18 indicated that they were “worried [they] would not have enough food to eat,” 10 indicated that they had “eaten only a few kinds of foods” due to lack of money or other resources, seven “ate less than [they] thought [they] should,” two were “unable to eat healthy and nutritious foods,” and two reported that their “household ran out of food.”
DISCUSSION
When using the FPI questions, which is what is currently used to track food poverty in Ireland, the prevalence of food insecurity in our sample (at 3.9%) is lower than recently published national estimates, though it needs to be noted that our sample is not generalisable to the whole population. Official government statistics estimate that in 2020 – when our survey was conducted – 12% of the population was experiencing food poverty (Department of Social Protection, 2022b), per the FPI. That the prevalence of food insecurity is so low in our sample is not too surprising given that our sample is not representative of the general population as it reflects only those with very young children and a vast majority of the sample reportedly have a 3rd-level education and are either married or in a partnership.
The government's estimate of 12% of the population experiencing food poverty in 2020 was higher than both 2019 and 2021, which reported 7.4% and 8.9%, respectively. This is despite the Irish government's attempt to mitigate some of the financial losses experienced by families during the COVID-19 pandemic. Measures included the Pandemic Unemployment Payment, a social welfare payment for people who had lost their employment due to COVID-19 or had left work to care for a child because of school or childcare closures. This payment was linked to a person's prior income, with a person receiving up to €350 per week (Department of Social Protection, 2020). There was also an enhanced illness benefit for those individuals who had COVID-19-related work absences. In addition, due to the school closures during this period, those families who availed of the school meals programme could receive weekly boxes of food through the programme instead. However, an evaluation of the school meal programme has indicated that the scheme's cost fell for the 2020/2021 academic year, as a result of reduced uptake during school closures (Department of Social Protection, 2022a). Thus, it is likely that many vulnerable children were not accessing this alternative support.
A particularly interesting finding of the present study is the proportion of households classified as food insecure using the FIES measure (10.5%), which measures both worries about access to food and the objective experience of going without food. This suggests that if the FIES was used and we measured food insecurity in Ireland on a scale from mild to severe, we would likely see a much higher prevalence of food insecurity in the general population. As it stands, we are not estimating the proportion of the population who are experiencing anxiety or fear around their access to food, which potentially has a considerable impact on quality of life. It is also important to note that the FPI questions are asked of the person who is answering the questions, whereas the FIES questions ask about “you or others in your household.” Thus, part of the discrepancy between the two measurements could also be because the individual responding to the FPI is not experiencing food access challenges, but someone else in their household may be.
Problems with the FPI have been reported previously (Healy, 2019); using data from 2004 to 2015, Healy compared the prevalence of food poverty using the questions on the SILC survey (i.e., the FPI) and then with the addition of an indicator that measures an element of social exclusion related to food poverty – the inability to afford to have family or friends for a drink or a meal once a month. It was noted that, when the concept of social exclusion was included in the calculation, the prevalence of food poverty increased significantly, from 9% to 16% in 2004 and from 11.5% to 21.1% in 2015 (Healy, 2019). That alternative measures result in a larger proportion of the population classified as living in food poverty may be explained by the fact that the current indicator for measuring food poverty in Ireland is solely an economic measure; it is entirely focused on the ability to afford food. It does not capture the social elements of eating, nor does it capture any underlying concern around food access.
According to safefood, food poverty “encompasses both the lack of access to a nutritionally adequate diet, and the consequential impact on health and social participation” (safefood, 2021). Thus, the present study – and others – highlight multiple issues with the way food poverty/insecurity is measured and reported in Ireland. There is no specific indicator for food insecurity among children, anxiety about access to nutritious food is not considered, and the social dimension of food intake is also not reflected in our national measure. In addition, the current food poverty indicator in Ireland is based on data obtained from a sample of private households, which does not represent those living in public institutions, communal accommodation, those who are homeless, and some members of the Traveller community, an issue that has previously been highlighted (Healy, 2019; Vazquez Mendoza & McDonagh, 2022).
The challenges associated with measuring food poverty are not unique to Ireland; food poverty is a multi-dimensional concept and is difficult to measure. A universal definition would be useful to enable cross-country comparison, and this is particularly the case for measuring food insecurity experienced by children (Fram et al., 2015). Recognising food poverty as an issue, Ireland's current Programme for Government has committed to ‘work across government to address food poverty in children and ensure no child goes hungry’ (Department of the Taoiseach, 2020). Progress in meeting this commitment is being monitored by the Children's Rights Alliance (Children's Rights Alliance, 2022). However, we cannot truly understand the impact of any measures introduced to alleviate childhood food poverty until we develop an appropriate indicator of childhood food poverty that captures the experience of all children and ensures the data are reported consistently. Neither the FPI nor the FIES measure food insecurity specifically among children in a household.
Despite the FPI not including a measure for anxiety about access to nutritious food, the misclassification of food insecurity between the FPI and the FIES is not as simple as the FPI missing cases of mild food insecurity, though this was observed in the present study. The largest other source of misclassification was the people who were classified as food secure by the FIES but responded in the affirmative to the statement “I cannot afford a roast dinner or vegetarian equivalent once a week” on the FPI (n = 15). It is possible that the traditional “roast dinner” in Ireland is culturally outdated, and this question does not serve its intended purpose, as previously highlighted (Healy, 2019).
Strengths and limitations
Given the public health restrictions in place during the data collection period, the use of an online survey was a strength of this study as it was an opportunistic way to collect data. However, typically the questions that make up the FPI are part of the SILC, which is a telephone-administered survey, and it is possible that these questions are not as well suited to online administration. Given that we believe 21 of our respondents made errors in their responses (see data cleaning, above), these questions may be more suitable when an interviewer is available to provide clarification in the event of confusion. In addition to the change in the mode of administration of the FPI questions, we changed the recall period for the FIES from the standard, validated, 12-month period to a shorter time frame (i.e., “[d]uring the coronavirus crisis”). This was to capture people's experiences during the pandemic. However, it has been reported that shortening the recall period for some food insecurity questionnaires may lead to a lower estimated prevalence of food insecurity (Poblacion et al., 2021), so this needs to be borne in mind when interpreting our data. Finally, questions about food insecurity were placed at the end of this questionnaire and it is possible that there was some participant fatigue that led to errors in reporting.
CONCLUSION
Up to 1 in 10 of our sample were food insecure based on the FIES. The disparities observed in the results of the FPI and FIES highlight the need for a standardised definition of food insecurity in Ireland as we may not be currently fully capturing the lived experience of the population, particularly those who may be experiencing anxiety and worry related to food access. This present work is insufficient for determining the appropriate or best tool for the measurement of food security in Ireland, but it does highlight deficiencies in our current measurement tool. Future research should focus on optimising a tool for measuring food insecurity in Ireland, and food insecurity among children so that assistance programmes can be developed to target those most vulnerable.
ACKNOWLEDGEMENTS
The authors are grateful to the study participants for the time taken to share their experiences. The authors also thank the Irish Nutrition and Dietetic Institute for facilitating the survey on their SurveyMonkey platform. Open access funding provided by IReL.
FUNDING INFORMATION
The authors received no specific funding to carry out this research.
CONFLICT OF INTEREST STATEMENT
The authors have no conflicts of interest to declare.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
