How did your research into body mass index (BMI) in African American adults and proximity to fast food restaurants originate?
This project was a secondary project that came out of a longitudinal cohort study that we've been conducting with the Windsor Village United Methodist Church in Houston Texas, which is a large, primarily African American mega-church with over fifteen hundred congregation members.
The original longitudinal cohort study was designed to better understand social, environmental, and behavioral factors that might influence cancer risk among African Americans.
The original study collected survey data on an annual basis among a convenience sample of about fifteen hundred church attendees, the majority of which were women. And the BMI project was a side project from data that were collected in the first year of that cohort study.
Part of the motivation for this project was that we realized there was a high proportion of overweight and obesity among the participants in the original study.
And so because the purpose of the original study was to better understand the various factors that might predispose African Americans to cancer risk, it seemed important to further investigate what might be associated with the overweight and obesity rates in this sample.
Why was your research conducted solely in African Americans?
It's certainly not to suggest that the things that we studied might operate differently among other racial and ethnic groups. We don't necessarily believe that.
However, one of the things that we know is that unfortunately African Americans are under-represented in many research studies. As a consequence, part of the reason why we developed the original longitudinal cohort study, which is nicknamed Project CHURCH, was because we wanted to devote effort to collecting data within this population and better understanding some of the things that might predispose them to cancer risk.
We know that African Americans unfortunately experience a number of different health disparities related to cancer as well as diabetes and heart disease and other health conditions that can be influenced by being overweight or obese.
What did your research involve?
We linked information about participants’ home addresses to a national database with the location of fast food restaurants in the Houston area.
Geocoding is when you associate a residential address, or any address, with a location in space to get latitude and longitude coordinates. In this study, we used geocoding to associate where each study participant lived to the fast food environment around them.
Then we calculated our two main predictor variables. One was the distance in miles along the road and street network to the closest fast food restaurant to each participant’s home. That was our proximity variable.
The other variable involved fast food density. We actually created a few density variables. What we did in that case was we created a buffer, a kind of a circle more or less, around each participant’s residential address using different buffer areas.
So we looked at first the closest buffer, which was basically a half a mile around each participants address using the road and street network. Then we counted the number of fast food restaurants that fell within that buffer and divided it by the area of the buffer.
We did the same thing for buffers at one mile around the address, two miles around the address, and five miles around the address, to look at density of fast food restaurants and how that might also be associated with BMI among our participants.
Participant’s BMI, as well as background information including age, income, educational level, the presence of children in the home, and so on were collected as part of the original cohort study.
We related the background information about participants to the fast food proximity and density variables to better understand how fast food restaurant availability might relate to BMI.
How did your research control for other factors that may influence a person’s BMI?
We had a good deal of information that was available to us that we could include in our statistical models to essentially control for the influence of these factors on the relationship between fast food availability and BMI. This isn't something that many other studies in the area were able to do because such detailed information is often not available in a single study.
So we knew information about participants’ self-reported physical activity levels, whether or not there were children in the home (which has been associated with people's physical activity rate), fruit and vegetable intake, and their age, sex, marital status, income, educational and occupational status. Because we knew where they lived, we also had an idea of the median neighborhood income and the percentage of African Americans in the neighborhood, which we also controlled for, because we know that the prevalence of fast food restaurants may vary based on these factors. We also knew how long they lived in their neighborhood and we included this as a covariate as well to try to better understand the unique influence of fast food restaurant availability on BMI.
What did your research find and were you surprised by these findings?
We found was that there was a significant association between proximity to the closest fast food restaurant and BMI, such that the closer participants lived to the closest fast food restaurant, the higher their BMI.
Next, we split the sample to look at whether or not people's personal income level moderated the relationship between fast food availability and BMI, in other words, whether or not the relationship between fast food availability and BMI differed depending on how much money people had available to them through their income.
We found that the relationship between close proximity to fast food and a higher BMI was actually stronger among individuals who had lower income. This makes sense because fast food is designed to be affordable, and so might be particularly appealing to those of lower income.
With regard to our density variables we found that for the sample as a whole, there was no significant relationship between the density of fast food restaurants around the home and BMI.
However, when looking at these associations based on participant’s income, we found that there was a significant relationship between fast food density and BMI, such that within two miles of the home, a higher density of fast food restaurants was associated with a higher BMI – but only among lower income participants.
You also asked whether or not I was surprised by the finding. There have been a number of studies that have been conducted looking at fast food availability and BMI or overweight and obesity.
In general, the findings have been mixed. That may be because it might depend on the area of the country in which the study is conducted. Or, perhaps because previous studies have not taken in to account how associations might vary by participant’s personal income levels.
There's only a single study we are aware of that looked at these relationships among an entirely African American sample. They only looked at the density of fast food restaurants. Similar to our study, for the sample of the whole they didn't find a significant relationship between the density of fast food restaurants and BMI. However, they didn't look at whether or not those relationships differed by people's income.
How do you think these results can be explained?
Well, I have to say that this is an association study - not a cause and effect study. It doesn't necessarily mean that everybody who lives close to a fast food restaurant is going to have a higher BMI. It just means that these factors were associated within our sample.
Unfortunately, we don't know if people in our sample ate fast food. We don't know how often they ate fast food. If they ate fast food, we don’t know which restaurants they patronized or what products they chose. So, there's a lot of missing information to be addressed in future studies.
We know typically people who frequently eat at fast food restaurants have a higher BMI. That's what we think our results might imply - that if you live closer to a fast food restaurant, you're probably more likely to eat there. If you're more likely to eat there, or eat there often because it's available and convenient to you, then that may influence your body mass. However, we can't say that based on our data alone. We can only think that our data may imply that.
Perhaps if a fast food restaurant is close to you, or if there are a lot of fast food restaurants nearby, it makes it really convenient to eat there. In addition, it also may cue you to want to eat there – the smell of the french fries or the chicken cooking - it's a very enticing smell. So perhaps if it's something that you're exposed to on a regular basis because it's really close to your home, or because there's a lot of fast food restaurants near your home when you're on your way in and out of the neighborhood, maybe it makes you more inclined to stop there because you begin to “crave” it.
How did your research compare to previous research carried out on BMI and neighbourhood environment?
I believe I already addressed this in response to a previous question. In short, our density results were consistent with another study conducted among African Americans, but the addition of the proximity factor and the moderation by personal income added to the literature in this area.
However, I will say that fast food is only one component of the food environment. There are many other things to consider in the neighborhood food landscape.
We didn't include information about the availability of local grocery stores; the affordability of local grocery stores; the availability of other restaurants that weren't fast food restaurants that perhaps offered healthier menu choices. We don't know those things and how those things might impact people's BMI or their choices to eat fast food.
Unfortunately, I think few studies really comprehensively examine the food environment in that way, to be able to speak to how the food environment as a whole affects BMI, because it's complicated to do so.
What impact do you think your research will have?
I hope that it will raise awareness about the potential for the environment to affect people's health. People may disregard the fact that our environment may be set up in ways to purposefully influence our behavior. Whether that's through the presentation of advertising, or the the availability and placement of certain consumable products, like fast food.
Of course people can choose to do whatever they want. People are free beings. However, there are also some things in the background at play that can have an influence on what people do, that's perhaps even a little bit outside of their awareness.
I hope that my research reminds people that there may be factors in the environment that influence our health behaviors.
What further research needs to be carried out into BMI and neighbourhood environment?
Well as I said, what we would like to do, and what I think would add to the literature, would be to better understand on an individual level what people are actually eating, how often they patronize fast food restaurants, what products they purchase at fast food restaurants, and the portion size of the items they purchase at fast food restaurants and integrate this into the study of fast food availability and BMI.
Perhaps future studies will look at the impact of people's income on these relations, which we hadn't seen previously, which ended up being an important part of our study, which again makes sense when you're considering that fast food is a particularly affordable option for individuals who have lower income and perhaps have reduced opportunity to purchase whatever type of food that they would ideally like to eat.
Another potential future direction, which would be great to do in theory but maybe difficult to implement in practice, would be to actually conduct a longitudinal study on this topic. For example, to examine changes in BMI over time based on naturally changing conditions within their neighborhood, such as increases in the availability of fast food, or decreases in the availability of fast food based on the opening or closing of restaurants around where people live.
If we had that information, and we collected that over time, we would really be getting a lot closer to better understanding a cause and effect relationship. And, as I said previously, future studies should seek to include other aspects of the food landscape as well.
Where can readers find more information?
For additional information about Project CHURCH (Creating a Higher Understanding of Cancer Research and Community Health), the longitudinal cohort parent study from which the data reported in our study came: http://www.mdanderson.org/publications/annual-report/issues/2009-2010/project-church.html
For more information on overweight/obesity rates by race and sex: http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6107a5.htm
About Dr Lorraine Reitzel
Lorraine R. Reitzel, Ph.D. is an Assistant Professor in the Department of Health Disparities Research at the University of Texas MD Anderson Cancer Center, and a licensed psychologist in Texas.
She earned her Ph.D. in clinical psychology from Florida State University and completed her internship at the Federal Medical Center in Fort Worth, Texas, in 2005.
Dr. Reitzel’s current research program focuses on increasing understanding of the influence of neighborhood context, socioeconomic status, and social status on cancer-related risk behaviors. She is also interested in the development and evaluation of theoretically-based interventions to address tobacco dependence among low socioeconomic status and other underserved groups.
Her research in these areas has been previously supported by the National Cancer Institute through a R25 award, the Centers for Disease Control and Prevention through a K01 award, and the Food and Drug Administration Center for Tobacco Products through a research subcontract. Dr. Reitzel is currently supported on two R01 grants from the National Institutes of Health and a project funded by the Cancer Prevention and Research Institute of Texas.
Dr. Reitzel has over 50 empirical peer-reviewed publications in journals including the American Journal of Public Health, Addiction, Social Science & Medicine, and Health & Place.
Dr. Reitzel’s recent work includes the integration of Ecological Momentary Assessment and geospatial methodologies to better understand how tobacco retail outlet density and proximity affects real-time smoking urges among quitting smokers, and how homeless smokers’ daily travels in relation to a shelter affect their mood and stress in-the-moment during a smoking quit attempt.
Dr. Reitzel also serves as a Consulting Editor of the American Psychological Association’s Clinician’s Research Digest.
Patient-derived organoids: Transforming cancer research and personalized medicine