CARRFS eNews spoke with Dr. Valerie Tarasuk, Professor, University of Toronto about Measuring Food Insecurity and Health in Canada.
What inspired you to work on the issue about household food insecurity?
Back in the 1980s when I was a graduate student in Toronto, food banks had just started and my Master’s thesis was to look at food problems for low-income single-parent families in Regent Park, Toronto. My first research project as faculty member at the University of Toronto was to do a study on nutritional vulnerability and food insecurity among women in families who use food banks. In the beginning, all we knew was how many people were going to food banks. But in the mid-90’s, we started to see individual questions about food insecurity on national surveys – very basic questions like “is your child going hungry”. It wasn’t until 2004 that we finally started to use a standardized, validated instrument on a national survey. Statistics Canada incorporated the 18-item module – the Household Food Security Survey Module – into their Canadian Community Health Survey (CCHS). That module can be used to generate prevalence estimates, but it also provides a measure of the severity of food insecurity – a very treasured measure. The module is being administered regularly, so we can now look at trends over time.
What are the challenges you encounter using the CCHS?
The CCHS is a health survey; and if we want to understand the social and economic conditions that give rise to this problem, we need to know something about household characteristics. CCHS, like all health surveys, has only bare bones socio-demographic variables. We can get measures about income and income source, but they are very limited. For example, the variable that asks you what your main source of income is has an option for social assistance. To be on social assistance is to be at very high risk of food insecurity, so this is important. But there are at least two programs that run under that umbrella in most jurisdictions. One is the general welfare program and the other is a program that provides income support for people with disabilities. People who are considered to be disabled typically get better benefits and therefore are probably less at risk of food insecurity than people receiving welfare. These details don’t get captured in the CCHS. It is important to understand food insecurity within a broad social policy framework (including income and other socio-demographic variables) because any mitigation of the problem will require policy interventions in these areas.
“Increasingly cities are becoming leaders in implementing [social] policy.”
Our report on Household Food Insecurity is based on the 2011 CCHS data. CCHS has a census metric variable that provides us with estimates for the largest cities in Canada, but the household-level weights necessary to calculate population prevalence estimates for household food insecurity have not been calibrated at the level of census metropolitan areas. So, while we have calculated household food insecurity rates for cities, we have had some concerns about the stability of these estimates because of the need to use household weights. Despite those constraints, we have some very interesting findings from our examination of the data for cities. The beauty with this food security module is that it has now been run for multiple years of CCHS data. This allows us to look at changes over time, and we are seeing profound change in some census metropolitan areas as well as quite dramatic variation in prevalence rates from one city to the next. What is interesting is that lots of cities are setting up food policy councils and food strategies. Things are happening at the community level around food; and in the case of Calgary at least, we are seeing initiatives to reduce poverty. It would be very valuable for those councils and government and non-profit stakeholders generally to have a benchmark around food insecurity rates. Increasingly, cities are becoming leaders in implementing policy. When you think about it, it is really important for Calgary to know what its food insecurity rates are prior to looking at its local programming to find out how well they are meeting the needs of the population over time.
In your report, Newfoundland and Labrador comes out with the lowest rate of Food Insecurity in Canada, why is that?
In the report that we released, we looked at the prevalence rate from 2005 to 2011. Not every province and territory had data on food insecurity for every year, because some had opted out of the food security module on cycles of CCHS when it was optional; but still, we had quite a few years to work with. From 2007 onward, we see a steady decrease in food insecurity in Newfoundland and Labrador. In 2007 they looked very much the same as the other Atlantic Provinces. Then they started to look different. By 2011, their prevalence had dropped by one third from 15.7 percent in 2007 to 10.6 percent in 2011. In fact, in 2011, they had the lowest food insecurity rate in the country. What changed over time? Well, in 2006, Newfoundland launched a very aggressive poverty reduction strategy. But is that what led to the drop in food insecurity, or is it because of an upswing in the economy? We don’t know yet.
We are in the process of trying to figure out what drove the change in Newfoundland and Labrador, but I think we will quickly reach the limits of the demographic variables on the CCHS. Maybe what we need to do is to link CCHS to other data sources - income tax records, for example – in order to see the effect of the taxation policy. If we look at who is food insecure – from looking at the national data over time, we know a couple of things – we know that there are two groups that are particularly vulnerable. One is those on social assistance, and the other is people in the low-income end of the workforce. About 61 percent of food insecure households in 2011 were reliant on employment income. So when we look at the Newfoundland story, we have to look at two groups of people – the people on social assistance and the people in lower paying jobs. We already know from our preliminary work that the probability of being food insecure if you are on social assistance has fallen in this province since 2007, but that is only one part of the story. So there is a huge complexity here. It is not going to be one policy, but multiple changes that have taken place. What I love about the Newfoundland story though is that it shows that things can change, and change for the better.
Have you been able to link food insecurity with obesity?
We have looked at the relation between food insecurity and obesity. The relationship is not straightforward. Literature shows that at the cross-sectional level – when we look at body weight and food insecurity – we can see a slightly higher rate of obesity amongst women in food insecure households in Canada, but no similar association is seen for men or children. Does that mean that food insecurity predisposes people to obesity? There have been a few longitudinal studies in the US and they don’t show that. They don’t show that becoming food insecure leads to weight gain. Is there a cause and effect relationship here? In fact I doubt it!
We have done quite a lot of work to look at the relationship between food insecurity and a broad spectrum of health measures. Food insecure adults in Canada have systematically poorer health. That is true whether we look at self-rated health or we look at the physician-diagnosed conditions. But it doesn’t seem to fall into a particular pattern. It doesn’t matter if we are looking at back problems, or migraines or diabetes. That said, the relationship between food insecurity and mental illness is by far the strongest. Almost half of the women in severe food insecure household in Canada have been diagnosed with mental illness. This is very dramatic and obviously cause for concern and further investigation. The most common interpretation of observed associations between food insecurity and all of these health measures is that food insecurity causes people’s health to diminish. I am sure it does, but another scenario is that to be low income with a chronic disease is to be more vulnerable to food insecurity. Our research is pointing in this direction; and if we are right, the implications for both policy and practice are huge.
At the same time, there is definitely something about food insecurity that is creating a health disadvantage. There are some very important papers coming out from the National Longitudinal Survey of Children and Youth which started to track children in 1994. The lead investigator in this work is Dr. Lynn McIntyre from University of Calgary. If you look at children over time that have lived in a severely food insecure household, they are more likely to have been diagnosed with a whole range of health issues such as asthma and depression – significant chronic health problems that are carrying on into adolescence and early adulthood. What the studies tell us is that the circumstances for children in Canada living in food insecure households are bad enough to predispose them to some fairly significant health problems going forward.
What is your next project?
We have the good fortune to have funding from the Canadian Institutes for Health Research (CIHR) to try to identify effective policy interventions. Part of our work will be to tease apart what changed in Newfoundland and Labrador and what caused those changes. In our team, there is also a group at the University of Calgary (led by Herb Emery and Lynn McIntyre) who have been looking at pensions and guaranteed annual incomes. One of the observations they have made from the CCHS data is that people who are elderly appear to be at much lower risk of food insecurity. What is it about the policies that kick in at the age of 65 that seem to be so effective?
“Food insecure adults in Canada have systematically poorer health.”
Another direction that we are moving in, working in partnership with scientists from the Centre for Addiction and Mental Health (CAMH), is to look at CCHS data that have been linked to health services utilization data (OHIP) records. Our purpose of doing that is two-fold: one is to enable us to estimate the cost of food insecurity from a health care perspective. We know that people who are food insecure are more likely to be unhealthy. There are studies that suggest that if food insecure individuals have chronic disease(s), they will be less able to manage their disease(s). There is some literature that suggests that food insecurity could give rise to health problems as I referred to earlier in the work done with the National Longitudinal Survey of Children and Youth. We want to link CCHS food insecurity measures to individuals’ health records to document these trajectories and estimate the health service utilization costs of this problem. We think it is really important since the kind of policy recommendations that emerge to reduce food insecurity include interventions that will cost money. Raising social assistance rates, improving tax benefits for working poor, and providing drug benefits for low-income people etc., are all interventions with a price tag. It is important to look at the health care costs associated with the current problems. If we can get a cost estimate from the linkage of these data sets, we would then be able to say: “you can pay this way or you can pay that way!” – either an investment in prevention or paying the cost in health care.
Anything that an organization like CARRFS can help out with?
I think it is important for people at the local level to start using these data. If indeed it turns out that the household weight is a problem for local level/health region analysis – then I think an organization like CARRFS can go to battle for it. If we were able to generate data on the local level, it would be very important as an informative action. It could be as simple as requesting that Statistics Canada can do a little more work on survey weights, which is not a huge ask. <>
By Jostein Algroy
The Canadian Community Health Survey requires substantial work and effort. Dr. Valerie Tarasuk explains some of the challenges in an interview with CARRFS eNews.
When Statistics Canada runs a survey like the Canadian Community Health Survey (CCHS), they need to be cost efficient and minimize the time it takes, or what is called the “respondent burden.” Therefore they are always looking to do what is necessary and not to do anything more. This is a difficult task as they have multiple agendas that they are trying to achieve with a single survey. And so, it has been a very long process to get the 18-item Household Food Security Survey Module on this survey. For years, I have worked closely with the staff at Health Canada – the Office of Nutrition Policy and Promotion – which is involved in monitoring food insecurity, to try to get better questions asked on these surveys. The core question has been: how do we monitor the problem on a national level; and what is it that needs to be done? At the same time, my research group has been working to learn as much as we possibly could about food insecurity in this country from whatever questions did get included on national surveys. We published a paper on earlier questions of food insecurity based on the National Population Health Survey. In that survey, they used only three questions. Nevertheless, we managed to analyze the relationship between social demographic characteristics and health outcomes with some remarkable findings [Household Food Insufficiency Is Associated with Poorer Health, The Journal of Nutrition, 2003].
In 2001, I went on sabbatical and Health Canada contracted me to write a discussion paper on the issue of food security measurement [Discussion Paper on Household and Individual Food Insecurity, 2011]. The paper went through a peer review process and was translated into French. It became a working document for the government as they carried on the conversation on the measurement of food insecurity.
“They [Statistics Canada] are always looking to do what is necessary and not to do anything more.”
In 2004, there was a special cycle of the CCHS that focused on Nutrition – including detailed dietary intake measures. Because of the focus on nutrition we needed to do it right – meaning including food insecurity in the survey. That sets the precedent and there has been no looking back. The Household Food Security Survey Module is now mandatory content on the CCHS on alternate cycles, and I hope this will continue.
Statistics Canada is preparing to do another focus survey on Nutrition in 2015. It is a smaller sample than the regular annual cycle of the CCHS survey, but invaluable insofar as it gives us a chance to look at the effects of food insecurity on people’s diets. I am part of an expert advisory group regarding the 2015 survey and was recently at a meeting where again, on the table were the questions: “Must we ask all the 18 questions on this module and can we produce a shorter module? If we must, must we measure food security? What do we learn by having it on the 2015 survey? Do we need to take that much time?” The answer of course, is YES – we have to measure it and we need to measure it properly. <>