No low hanging fruits: salient factors for good nutrition by socioeconomic status

In recent years increasing obesity rates have become a major concern in the United States. Statistics show socioeconomic status plays a big role in obesity outcomes, with those of lower SES more likely to be obese. Although exercise and stress have also been shown as important contributors to obesity outcomes, this paper will focus on how nutrition differs by SES through structural factors, discussed here as the food environment. The food environment includes healthy food access, cost of food, education about food, time required to prepare food and social influences, any of which may be more or less serious moderating factors in determining nutrition and obesity outcomes in conjunction with SES. This paper examines which of these factors have the most impact on nutrition outcomes, and consequently, obesity outcomes, between SES groups since changes in the food environment may prove more effective than targeting individual behaviors toward good nutrition. To accomplish this goal, research articles and USDA data were analyzed and compared to determine which factors appear most salient between SES and the food environment and, ultimately, good nutrition and obesity rates.

NO LOW HANGING FRUITS: SALIENT FACTORS FOR GOOD NUTRITION BY SOCIOECONOMIC STATUS by Donovon Jenson A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Health, Society, and Policy Approved: ____ Norman Waitzman, SupervisoK Norm Chair, Departme ,0 ' Dr. Sylvia D. Torti Dean, Honors College Normaii W aitzman Department Honors^ dvisor April 2014 ABSTRACT In recent years increasing obesity rates have become a major concern in the United States. Statistics show socioeconomic status plays a big role in obesity outcomes, with those of lower SES more likely to be obese. Although exercise and stress have also been shown as important contributors to obesity outcomes, this paper will focus on how nutrition differs by SES through structural factors, discussed here as the food environment. The food environment includes healthy food access, cost of food, education about food, time required to prepare food and social influences, any of which may be more or less serious moderating factors in determining nutrition and obesity outcomes in conjunction with SES. This paper examines which of these factors have the most impact on nutrition outcomes, and consequently, obesity outcomes, between SES groups since changes in the food environment may prove more effective than targeting individual behaviors toward good nutrition. To accomplish this goal, research articles and USDA data were analyzed and compared to determine which factors appear most salient between SES and the food environment and, ultimately, good nutrition and obesity rates. TABLE OF CONTENTS ABSTRACT ii INTRODUCTION 1 METHODS 2 THE FOOD ENVIRONMENT 2 DISTANCE AND ACCESS 3 COOKING 6 EATING OUT 8 OVEREATING 11 EARLY HABITS 12 EDUCATION/FOOD LABELING 14 COST OF GROCERY FOODS 17 CONCLUSION 20 DISCUSSION 20 APPENDIX 23 REFERENCES 27 iii 1 INTRODUCTION Over the past several years, increasing obesity rates have become a major concern for the U.S. population. From 1960 to at least 2012, and almost certainly beyond, obesity rates have been spiraling upward at an alarming speed across all demographic groups, including age, gender and ethnicity. (1) In addition, obesity has been increasing at all levels of income and education. The pervasiveness of this trend has become a compelling issue with growing interest from the health community as officials look for viable solutions. (2) One misconception is that low socioeconomic status (further references as SES) populations compose the majority of the obese population and that their rates are growing the fastest. Contrary to this belief, evidence based on the NHANES, NHIS, and BRJFSS generally suggest those in the upper classes are actually gaining shares of obesity and income-related obesity disparities have weakened with time. Even when considering a threshold effect of obesity, this data suggests BMI has been increasing disproportionately toward higher SES populations, especially those classified as middle-income. (44) However even with these trajectories, disparities still exist between groups, with low SES individuals having the highest rates of obesity (3). While obesity has been increasing regardless of SES, it would be fallacious to believe individuals from differing levels of SES have the same structural factors, discussed here as food environments, available to them. Evidence shows that low SES individuals are more likely to have less nutritious diets, both through actual consumption and through purchasing patterns as a proxy. (14)(42)(43) Unfortunately, researchers have difficulty pinning exact causes to these disparities. This paper hopes to examine which factors in the food environment are most important for explaining the differences between different SES populations by analyzing and comparing various research articles and USDA data. METHODS To discover which factors were believed to contribute to this disparity obesity and nutrition literature was scanned to see what structural factors previous researchers found or postulated were relevant between SES groups. Then, using various research articles, USDA data and several other sources, arguments were built to show how well each factor might explain differences in nutrition and, ultimately, obesity, by SES. THE FOOD ENVIRONMENT To better understand risk factors one must have an understanding of what the food environment is and how it is affected by SES. An individual’s food environment, as used in this paper, refers to a number of variables including amount of time to prepare food, access to healthy food, distance from food, cost of foods, food-related education level, and social influences. Previous research has indicated that, taken together, these variables coagulate to form potential opportunities and mediating influences on food choice for an individual. Research tends to show that those with lower SES are more likely to have restrictions and outcomes that lead to obesity across these variables, however there is not much clarity about how much each of these factors accounts for the disparity. (14)(42)(43) By breaking these variables apart then comparing relevant research articles and datasets this paper hopes to prioritize which components are most important for explaining the difference between obesity rates in high and low SES populations. DISTANCE AND ACCESS When evaluating the food environment it’s logical to start with distance to food purchased to eat at home. Individuals are likely to purchase foods from locations that are within traveling range of their residence, whether it be a convenience store, supermarket, or restaurant. Supermarkets, defined as selling all categories of food, are much more likely to carry healthy food choices and alternatives. There is a significant amount of health research concerned with access and distance to supermarkets as an essential component of obtaining a nutritious diet. When discussing healthy food access, the term food desert, a geographic area where affordable and healthy food is difficult to obtain, inevitably arises. Some may argue that food deserts are a significant barrier for those with low SES and is a major component of the difference in nutrition and obesity rates between high and low SES populations, but that conclusion is hard to draw after developing the full picture. On the one side, there is certainly research showing distance and access can have an effect on consumption patterns. For example, there is research to indicate that being closer to a supermarket results in increased purchase and consumption of fruits and vegetables for those with low income and on food stamps. (4) There is also research which found that households closer to fresh vegetables are more likely to increase vegetable consumption. (5) Based on this information it does appear individuals whom are closer to healthy foods are more likely to purchase healthier diets. However, this seems unlikely to be the major reason for the difference between obesity in high and low SES populations. Furthering this argument, using the aforementioned data regarding distance and increased vegetable consumption as a base, it should follow that consumption patterns are changing to reflect increased healthy food access across the population. Yet, between 1998 and 2006 we see little change in food-type purchase distributions. (10) (Al) This is cause for skepticism about the generalizability of those earlier studies on distance and consumption. It is also difficult to connect lack of healthy food access to poor nutrition and obesity. One study comparing health food availability and obesity in Baltimore neighborhoods actually found a higher BMI for neighborhoods with more access to healthy food, compared with those living in low healthy food-availability neighborhoods. (11) Of course this is only one example, but other evidence points to access as insufficient to change diets. In addition, a qualitative study in which women of various SES levels were asked about factors regarding social and environmental influences on diet found self-reported access did not differ between SES groups. Neither high nor low SES individuals perceived access to be one of the issues in creating a healthy diet for themselves and their families. (12) These two studies are not concrete evidence, but do provide an idea of how food access may not be as much of a barrier as originally perceived. One major detractor from the idea that healthy food access drives the difference between low and high SES populations is that access has actually been increasing. There is evidence to show that at least into the late 1990’s access to fruits and vegetables actually increased. (6) That figure may be a little dated, but the USDA released a 2009 report which stated only about 4% percent of the population or 11.5 million people live in low income areas that are also more than 1 mile from a supermarket. This report also mentions an additional 12 million individuals that are not low income living in the same areas, facing the same difficulties with access, which we would not expect to find if food deserts were primarily a problem for low income individuals. It appears the problem is determined by neighborhood SES, not necessarily individuals themselves. Higher income individuals may have better access to transportation, but the report concluded by saying there is insufficient data to determine whether low-access is actually inadequate access for these populations. If low access does not equate to inadequate access doubt is shed on food deserts a huge driving factor between low and high SES populations. (7) In another USDA report, using 2010 data, it was estimated only 1.8 percent of households live more than 1 mile from a supermarket and do not own a vehicle, with supermarkets defined as selling all categories of food. They also estimated half of the U.S. population lives within 2 miles of 3 supermarkets. Access to multiple supermarkets is assumed to provide a wide variety of choices through competition, including healthy foods. This report also states that, in general, more households living greater than one mile from a supermarket had vehicles in 2010 when compared to 2006. This data shows healthy foods access has been increasing for the population as a whole. (8) Another promising set of data from the USDA shows that the proportion of poor in food deserts is only roughly 20-25% and may be decreasing over time. (9) Although it may have made a difference in the past or for certain populations, it seems as though access is a weak factor the difference in the diets of low SES populations seen currently. In summary, although food desserts are likely a problem contributing to nationwide obesity rates, they do not strongly explain a significant difference by SES. If there are roughly the same number, or less, low-income and non-low-income individuals struggling to access grocery stores, there is little reason to believe access improvements which would affect diet are differentiated by SES. There is even reason to doubt that low access actually means insufficient access. In addition, with both increasing healthy food access and obesity rates, there is skepticism toward whether access has a significant impact on dietary changes and obesity rates at any level of SES, including low SES populations. Given these issues, access and distance appear to be weak factors in the nutrition differences between low and high SES populations. COOKING Another part of the food environment for an individual is the time needed to cook healthy food. Cooking healthy meals takes a significant amount of time compared to other sources of food such as ready-to-eat, convenience, and fast food meals. Differences in time dedicated to cooking is one idea that has been used to explain nutrition and obesity differences between high and low SES populations. According to this theory low SES individuals lack time to cook healthy food while their high SES counterparts spend more time devoted to cooking, resulting in healthier meals and lower obesity rates, however, this seems like a weak explanation given the evidence. Before assigning an SES role to cooking time, it is important to note that since 1900 the amount of time spent cooking food in the home has been decreasing all around, from 44 hours a week down to 10 hours or less a week in 1999. It is likely this trend has continued. The lower amount of cooking time has been offset by increasing use of prepackaged ready-to-eat foods as well as dine-in restaurants, fast food, and other forms of eating out. These convenience foods are likely to be lower in nutritional value than traditional meals, which require longer cooking times. Widely available and purchased convenience foods also generally have higher fat content than home-made counterparts. This can certainly affect diet quality and obesity, however, the author is not aware of any data which suggests the decrease in time spent cooking at home is moderated by SES. (6) What the data shows as more likely is that most individuals no longer have time to prepare healthy food, for different reasons, but regardless of SES. The same qualitative study mentioned earlier found low SES women have work commitments and high SES women have family commitments, both of which limit time available to cook healthy foods in similar ways.(12) This was a qualitative study, but there is evidence to suggest middle and upper class populations actually consume more convenient and ready-to-eat foods than low SES populations by total volume. (13) (14) With this information in mind, it seems unlikely that increased obesity rates in low SES populations are due to decreased commitments to cooking. One other issue worth mentioning is that knowledge of how to cook does not seem to differ by SES. This argument has not been seen much recently, but it is still worth dispelling. Although the research is from 1999 and may be somewhat dated, no significant difference was found in cooking skills between SES groups, especially in a way that would result in varied nutrition levels. (15) Based on the data it does not appear as though time available for cooking differs between SES in a way that would result in worse nutrition for low SES populations, If anything, the data suggests higher SES individuals should be more likely to be obese based on time taken to cook. Due to this time to cook differences are a poor explanation and likely a weak factor in differing rates of obesity and nutrition between high and low SES populations. EATING OUT A common stereotype explaining the difference between these populations is low SES populations eat food away from home more often than their high SES counterparts. As mentioned before, meals away from home are generally higher in fat and calories and have less nutrients. The more meals you have outside the home, the more likely you are to be obese. One might believe low SES individuals eat out more often, but this can be dispelled upon further examination. One reason for the idea low SES populations eat out more may be because of the high volume of fast food locations in poor areas. Evidence clearly shows fast-food locations are more densely packed in low SES. Multiple studies have found high income neighborhoods had less restaurants and fast food locations than lower income neighborhoods. (16) (17) However, while there are differences in the proportions of types of out-to-eat locations based on neighborhood characteristics, in a study using census zip code data and restaurant outlet data little difference was found in availability of fast-food compared to full service restaurants by SES. To clarify, this means both low and high SES groups were able to access these kinds of out-to-eat locations, regardless of area density. (16) While evidence shows a difference by density, the density of out-to-eat locations has not been shown to signify that low SES individuals eat out of home more often. Directly contradicting the argument that low SES populations eat out more, the 2012 USDA consumption statistics actually suggest that high SES individuals eat more food away from home by volume than their low SES counterparts, so this is hard to use as an explanation for differences. (18) (A2) One idea to explain the differences has been that quality of food is different between SES levels, with low SES populations gravitating toward fast-food and higher SES populations gravitating toward dine-in locations which they believe has healthier food. In fact, there is qualitative evidence, and the assumption that can be drawn based on fast-food location density in low SES neighborhoods, to suggest type of out-to-eat food does differ by SES in this way. However, this is not enough to suggest major differences in nutritional quality because the idea that dine-in locations offer healthier food appears untrue. (12) To better flesh out the quality issue let’s compare two popular out-to-eat locations, differentiated by price, McDonald’s and Applcbec's. We will assume that Applebee’s targets at least medium SES populations and McDonald’s targets lower SES populations because of differences in average meal price. We will start by looking at the health values of a meal from McDonalds. Then we’ll pick a menu item from McDonalds and compare it to a similar item from Applebee’s. According to McDonald’s website, a single Big Mac has 29 grams of fat, 45% of the daily recommended value for a 2,000 calorie diet. A normal meal purchase would also include a medium fry, 19 grams of fat, and a drink, which could also potentially contain fat as well. That is at least 74% of the daily recommended value of fat, in a single meal. Other values, which should be limited, such as sodium and caloric intake follow similar patterns, while vitamins and minerals are much sparser. With such high values for things like fat and sodium, it’s easy to understand how individuals who eat out often reach values much higher than what is needed, which can lead to adding weight in the form of fat. Clearly this type of food is unhealthy. (19) (6) Now let’s compare the Big Mac to a cheeseburger at Applebee’s. According to Applebee’s website a cheeseburger there has 61 grams of fat, astonishingly more than a McDonald’s Big Mac. A cheeseburger at Applebee’s also has 1700 milligrams of sodium, compared to the big mac’s 970. Even two Big Macs don’t match the fat or the sodium in a single cheeseburger from Applebee’s. The meal from Applebee’s would also come with a side item, and possibly a drink, driving those numbers for the entire meal even higher. This is not a good start to proving dine-in restaurants are any healthier than fast-food locations. (Comparison table in (A3) appendix) (20) This is only one example, so one might believe there are other options within chains which are potentially confounding, but USDA data shows that on average fullservice restaurants are more likely to provide meals higher in fat, cholesterol and sodium than fast-food alternatives. (45) In fact, there is even more additional data to support that both styles of out-to-eat locations are unhealthy and lead to increased obesity. (22) One study in Atlanta found it nearly impossible to order a healthy entree based on information easily accessible for both fast food and dine-in restaurant locations. (21) Although Atlanta is a relatively obese area, in conjunction with other resources, this helps paint a picture of the difficulty in consuming healthy foods outside the home regardless of SES. It appears that when going out to eat individuals may not be capable of selecting a healthy choice in either dine-in or fast-food locations. After reviewing the data, it appears volume of away- from-home consumption is not a strong factor for explaining nutrition differences by SES. If anything, high SES populations should be experiencing higher obesity rates because of their higher consumption rates. In addition, style of out-to-eat location and quality of food do not seem to make a difference either as both groups consume generally unhealthy choices. Considering these things, differences in food consumed away-from-home seems a weak driving factor between high and low SES populations. OVEREATING One theory to explain SES differences in nutrition is that low SES individuals lack self-control and typically engage in overeating at a rate higher than other groups. The author found no evidence to support this position well. This explanation may be a symptom of stigmatization and stereotyping of those who are poor and obese. That being said, this argument will be briefly addressed. Over the last several decades portion sizes have increased in a way that is clearly unhealthy for individuals. Both restaurants and pre-packaged foods in grocery stores provide serving sizes that are too high in either calories or fat content. In addition, out-toeat locations also have expanded portion sizes. Increased sizes tend to push consumers to consume more food. This is across all types of food and does not appear specifically targeted toward low income consumers. If this was a primary factor for the disparity we would expect to see differences in portion sizes by SES, but the author found no evidence supporting this position. (6) Another big piece in dispelling this argument is that, according to USDA consumption rates, higher income individuals actually consume more food, in grams, of nearly every food type category. If those from high SES populations are actually eating more food, it makes little sense to use overeating as an explanation for why low SES individuals have higher rates of obesity. (18) (A2) Differences in self-control and overeating is a poorly supported explanation of 12 why lower SES individuals tend to have lower nutrition. The author found no data to show that self-control in eating is any more a problem for low SES individuals than it is for their higher SES counterparts. Due to this, poor self-control and overeating appear to be weak causes for differences seen in nutrition and obesity. EARLY HABITS One argument has been that low SES children are targeted by unique social pressures which encourage them to eat less healthy and these habits are transferred into adulthood. It has been postulated that perhaps these children have higher perceived barriers for eating healthy and consume less healthy diets at school than their higher SES counterparts, which eventually become habitual. Even from a young age, unhealthy eating is a norm for all SES groups and it shows in school lunches. Although efforts have been made to improve quality through federal nutrition standards, public schools have a reputation for low quality and low nutrition lunches, compounded by budget issues. Commonly seen foods include french fries, pizza, and chicken nuggets, usually fried in unhealthy oils with low overall nutrient levels. Another factor lowering nutritional intake from school lunches are that some students discard or give away their fruits and vegetables due to flavor and preference. Children who do this have already developed the perception healthy foods taste unpleasant by the time they reach elementary school, which hints that parental eating habits also play a large role in what are considered food norms. This vegetable discarding behavior is seen amongst all SES groups. Unsurprisingly, students who eat school lunch are more likely to be overweight than peers who brought lunch from home. However, this may not be moderated as much 13 by SES as one might believe. Over half of children enrolled in school with a lunch program are eating school provided lunches, usually because of either low cost or convenience. As discussed earlier, convenience is something that is important for all SES groups, so not only low SES populations are utilizing these services, although they do represent bigger proportions. (23) (24) Even though school lunch participation is skewed toward lower SES populations, there is still the issue of unhealthy snacks. Many schools have vending machines or other items available for purchase that are not nutritious. One study comparing food choices within schools found overall the number of unhealthy snacks and a la carte items far outweighed healthy options. There is reason to believe that differences in school lunch use by SES are mediated by increased a la carte options, providing children not eating school with equally or more unhealthy alternatives. (24) There is some evidence to suggest SES plays a role in whether healthy options are available in a la carte menus or not, with lower SES areas having less access. (25) However, as discussed previously, access may not be enough to change consumption habits. Based on this, it is difficult to assign a strong value to outcomes based only on availability. Fast food and junk food advertising also play a role for children’s food perceptions and preferences. Studies show that children are exposed to a huge amount of advertising marketing them unhealthy foods. One study found not even a single source marketing healthy foods on television ads on children’s networks. Out of these unhealthy foods, it was found fast food is advertised much more than all other kinds of food. However, there was little to indicate that these ads particularly target low income children Perhaps low SES children watch more television, but no data shows larger behavioral 14 effects and diet choices for these populations. (26) (6) If differential fast food consumption could be shown, there is some data to show perceptions might be different. One study found that frequency of fast food usage is associated with poor beliefs about healthy eating among adolescents. They found perceived barriers to healthy eating consistently increased with increased fast food use. The authors did not distinguish between fast-food and out-to-eat food. However, as discussed previously, higher SES individuals actually consume more foods away from home by volume. Though it could be possible there are different beliefs based on type of out-to-eat style, it seems unlikely due to previous discussion of the health merits of different out-to-eat styles. (27) Considering the data, it seems hard to discern how or why early beliefs and perceptions vary by SES, if at all. All groups are highly targeted for unhealthy foods and out-to-eat habits are similar. There is some evidence to show that low SES individuals may feel more barriers toward healthy eating, but perhaps that perception is justified by something else in the food environment. If these barriers are real, these kinds of perceptions are only a weak driving force between various SES levels. EDUCATION/FOOD LABELING One common argument is that low SES individuals are not educated enough to create a healthy and nutritious diet. There is certainly some merit to this position, but after further review it appears this is not the strongest driver between low and high SES groups. Even with the correct types of knowledge, individuals with low income struggle to create a healthy diet. So, while education may be a piece of the problem, price may also moderate differences between low and higher SES groups, which will be expanded 15 upon in the following section. Contrary to the belief that education levels can help bridge the gap in healthy food choices, there is quite a bit of information to suggest most individuals have a difficult time navigating the food environment for healthy choices, especially when eating out. Referring back to the Atlanta study mentioned earlier, researchers found it impossible to order a healthy entree based on information easily accessible in the restaurants, for anyone. (21) This position is supported by USDA data and other research studies. (22)(45) When going out to eat individuals may not be capable of selecting a healthy choice even if they understand basic nutrition. Without specifically asking restaurant patrons may never know what’s in their food, including fatty sauces and oils. Furthering this point, it appears education levels may not convince people to start eating healthy in out-to-eat locations anyway. One study found increased out of home food consumption is associated with increased body fat, regardless of education level. The study did not specifically look at education about food and nutrition, but it is worthwhile to note that this type of education is not well translated to most individuals as they progress through the education system. Those who finished high school are likely to have the same level of formal nutrition education as those who complete a bachelor’s degree and so on unless the degree is in nutrition or a related field. (29) One suggestion has been to increase information available in out-to-eat locations. However, evidence shows at least calorie labeling may not help change outcomes when eating out. Many individuals use away-from-home meals as a time to be more lax with their diets and consider nutrition in a fundamentally different way than they do when purchasing from a grocery store. Something worth mentioning is that forcing increased 16 health labeling may actually have an impact on the types of food out-to-eat locations provide, as opposed to affecting consumers themselves, as early data suggests having to show these values creates pressure for changes toward healthier foods. (28) It is also plausible that knowing what foods are healthy is not sufficient for individuals to make habit changes. According to the USDA, between the late 1990’s and mid 2000’s the American population has increased realistic perceptions of diet quality. (30) As can be seen by still climbing obesity rates and little change in food type consumption distributions, this awareness has done little to affect dietary change. As mentioned before, from 1998 to 2006 there was only marginal changes in food patterns; a small increase in whole grains but a small decrease in both fruit and vegetable consumption over this period. These have not been especially encouraging findings to suggest huge differences in choices by education and awareness. (10) It has also been argued that low SES individuals simply do not understand how to create a healthy diet with their income levels. While it is true that some individuals may simply not be aware of how to create a healthy diet with their income levels, low SES individuals also face real financial barriers toward doing so, which is discussed further in the next section. The same qualitative study mentioned earlier found that women with low SES perceive a high tradeoff in cost when purchasing healthy food. Regardless of the actual cost of putting to gether a healthy diet, these women may be unfamiliar with how to do so, and afraid that costs are too high. Food placement in grocery stores makes items which are unhealthy and inexpensive most visible. The idea is that perhaps if these women had better strategies for locating healthy food this perception would change, pushing them toward healthier and more nutritious options. The authors suggested 17 educating these women about how to create a low cost, healthy diet may be beneficial, but this does not appear to be the case if costs are too high. (12) There is quite a bit of evidence suggesting low income individuals cannot afford to buy healthy diets even with highly efficient strategies for doing so. This is the biggest reason to discredit educational differences as the strongest driver of outcome differences. Education is not sufficient when many low SES individuals cannot access the financial resources to create a healthy diet. High cost of healthy eating is both a perceived and a real barrier to improving diets for these individuals. Knowing about the cost barriers illuminates what appears to be the strongest cause for nutrition differences between SES status, cost of foods consumed at home. (14) (31) Although education appears to be an important factor it does not appear to be the most important factor at the moment. Even with an education around nutrition, if low SES individuals cannot access the minimum financial resources to purchase healthy foods then it is impossible for them to create a healthy diet relative to their higher SES counterparts . Having a better idea of where to turn, let us now examine cost of foods consumed at home as a potential difference between low and higher SES populations. COST OF GROCERY FOODS After having looked at education, it appears cost of at-home foods may play the most significant role in nutrition and obesity differences between low SES and their higher SES counterparts. Low income individuals are having a difficult time affording quality foods at the grocery store. Upon further examination there is quite a bit of evidence to give this idea support. One big point referred to earlier is that low income individuals cannot afford to 18 buy healthy diets even with efficient strategies for doing so. (14) This means that cost is overriding education factors in allowing individuals to create healthy diets. When costs hold individuals back from purchasing healthy foods, what are they getting instead? Studies have found the cost of foods low in nutrition and high in calories and fat content to be significantly less expensive than healthier foods. These foods have also been found to be less satiating, encouraging individuals to eat beyond what is required calorically. In addition, many of these foods come in large portion sizes, which encourages individuals to eat more. Due to the low cost per serving and high calorie and fat content, individuals can quickly find themselves gaining weight without getting proper nutrition because it is what’s affordable. (6) To better frame the next discussion, let us now examine the USDA’s Thrifty Food Plan. The USDA puts out monthly reports of roughly how much individuals or families are spending to make a nutritious diet, divided into four levels. The plan with the lowest dollar amount is called the ‘Thrifty Food Plan’, which should be sufficient to make a healthy diet, however studies have shown this to be untrue. (32) One study compared two market baskets, one using the Department of Agriculture’s Healthy Food Plan and one with healthier substitutes, the average cost of the healthy basket was $36 higher. That difference is equal to roughly 35% of lowincome consumer’s food budget (based on $2410) in a year. That difference is certainly too much of a burden for a low income individual to bear when considering other financial needs. (33) In fact, many individuals are actually spending even more than the minimums suggested by the USDA thrifty food plan and are still unable to create a healthy diet. (34) These cost barriers show how other factors may be important, but not 19 currently the strongest driving cause of the disparity. These populations cannot be blamed for constructing unhealthy diets when they cannot possibly finance them. There is one one final piece of especially strong evidence for explaining cost of grocery store foods as the most significant drivers of nutrition disparity by SES. There was an experimental study which was run to find how cost of food changes the nature of the diet. They found that as incomes become more constrained individuals create less and less healthy diets, regardless of individual characteristics. (35) In essence, when income drops individuals become priced out of buying healthy food. This is a significant finding for explaining a strong difference between low and high SES populations. One the other end of the scale, there is also evidence which shows making healthy foods more affordable for low SES populations can have a positive impact on diet patterns. Some research shows price reductions can increase the consumption of fruits and vegetables. (36) There is also research showing price changes, via taxes or subsidies, affect eating habits for certain populations, namely low SES populations, children, and adolescents. (37) This may become pivotal for the future as the United States attempts to reduce obesity outcomes throughout the country. For other populations, the data suggested price changes were not a huge factor in dietary intake, furthering the argument that this is a large driving factor between SES levels. To contrast this, one study from the USDA found that subsidies only marginally affect purchasing rates for fruits and vegetables among SNAP participants. Oddly, the same study mentioned ‘expectations of how long the price change will last’ as an important variable for price responsiveness. If these SNAP participants were aware they were participating in a study, it is likely they had low belief in these subsidies becoming a 20 long-term decrease in the cost of fruits and vegetables. As mentioned before, high costs of healthy eating are a strong perceptual barrier for those with low income. (38) After analysis, it appears cost of healthy groceries is the strongest component driving nutrition differences between SES groups. Low income populations are struggling financially to purchase healthy foods. There are probably other barriers and issues for this population but the cost of healthy foods should be a high priority if progress is to be made toward a more equitable distribution. CONCLUSION Cost of food consumed at-home seems to be the strongest factor in differentiating nutrition differences between low and higher SES populations. Low SES populations struggle to afford healthy diets regardless of what strategies they use. Education is likely to be the next strongest factor in explaining the differences between these groups. Even if healthy foods were within the budget of low SES populations, they may still not understand how to create healthy diets based on their knowledge level. In addition, perceived barriers are real at the moment, but are likely to be a problem even if those barriers are reduced or removed, as individuals have low belief in their ability to create healthy diets. The quality of out-to-eat meals may also play a role, however there is little data to show low SES nutrition differs from higher SES populations in this realm. Other factors such as access, cooking time, quantity of out-to-eat meals, overeating, and early perception disadvantages do not appear to moderate SES differences in nutrition nearly as heavily as the other factors. DISCUSSION Unfortunately, it appears the obesity rate is still climbing across all SES groups. Whatever interventions have occurred up to this point have not been enough to reduce obesity within the population. (39) However, In the last few years, there has been some hope for the future. A 2010 USDA report stated that individuals are using grocery store labels more often, though that can’t necessarily be tied to better health choices. (40) Health misinformation is an issue not really discussed here, but it appears as though focus on the obesity crisis has helped parse out better sources of information and clearer definitions of constitutes healthy food or, at the very least, stronger attempts to locate healthy options. Data, published 2014, gives hope individuals are eating out less and making better food choices. Although out-to-eat foods are likely to rebound parallel to economic recovery, there is some hopefulness that healthy changes are around to stay. (41) If this trajectory continues, much of the problems associated with eating out and obesity may be reduced. As for obesity differences between SES groups based on cost, it does not appear much is being done at the moment to make healthy foods more affordable or to improve food education for low SES populations. These other improvements will likely have an impact on overall obesity rates, but until the cost issue is addressed, the SES differential is likely to remain a huge issue. There may also be other factors in the food environment which were not examined here but are also important for SES differentials. Certainly not all possible factors have been considered here. Two major factors which need to be addressed are exercise and stress and how those factors interact with SES and nutrition and, ultimately, obesity outcomes. By including these, other factors which seem less salient from a purely nutritional standpoint may prove more important than concluded here. In addition, future studies may want to address more nuanced issues of the food environment, even within the categories that have been set out here. Some issues which may need attention are those of food security, a more thorough investigation of beliefs’ effect on nutrition outcomes, further analysis of subsidies on specific healthy foods, and addictive properties of high calorie and fat foods, among other things. 23 APPENDIX A1 - Taken from a USDA report, citation (10). At-home whole grain purchases by U.S. households increased between 1998 and 2006, but fruit and vegetable purchases fell Expenditure shares, percent Who!® grains Refined grains Fruits Vegetables Processed*' paefaipi Beverages Note: Packaged and processed foods includes frozen or refrigerated entrees, soups, candies, and prepared foods not included in the whole and refined grains categories. Source: USDA, Economic Research Service calculations using Nielsen Homescan data. A2 - Created using USDA consumption data, citation (18). Total fats and oils Retail weight grams per day Total fruits Retail weight grams per day 400.00 350.C0 30000 250.00 200.00 i5D.ee imm mm 0.00 S/.r “vO“' nK " ¥ * Total Vegetable Retail weight grams per day 400. DD 350. DO 30000 250.00 mmrn 150.00 100-oo SQJOQ 0.00 ,s*JQ4 ky Total Dairy fC’ .jrf> JSP Retail weight grams per day Total meats, poultry and fish Retail weight grams per day im m 160.00 140100 123.00 Total grains Retail weight grams per day 140.00 Total Nuts Retail weight grams per day 26 Total Sweeteners retail weight grams per day 120.00 - - 100.CD A3 - Made using data from data availab e on company websites, citation (19) and (20). Applebee's Cheeseburger 940 61 g 1700 mg McDonald’s Big Mac 550 29 g 970 mg Recommended Daily Value (on 2000 calorie diet) 2000 65 g 2400 mg 27 REFERENCES 1) Flegal, KM., Carroll, MD., Kuczmarski, RJ., and Johnson, CL. (1998). Overweight and obesity in the United States : prevalence and trends, 1960-1994. International Journal o f Obesity, Volume 22. Retrieved from: http://www.nature.eonx/iio/ioi]mal/v22/nl/pdf/0800541a.pdf 2) Cynthia L. Ogden, Ph.D.; Molly M. Lamb, Ph.D.; Margaret D. Carroll, M.S.P.H.; and Katherine M. Flegal, Ph.D. (2010). Obesity and Socioeconomic Status in Adults: United States, 2005-2008. NCHS Data Brief, No. 50 Retrieved from: http://www.cdc.gov/nchs/data/databriefs/db50.pdf 3) No authors given. (2014) Relationship Between Poverty and Overweight or Obesity Retrieved from: http://frac.org/initiatiyes/hunger-and-obesity/are-low-income-people-atgreater-risk-for-overweight-or-obesity/ 4) Rose, D. and Richard, R. (2004). Food Store Access and household fruit and vegetable use among participants in the US food stamp program. Public Health Nutrition 7(8), 1081-8. Retrieved From: http ://www.ncbi .nlm.nih. gov/pubmed/155483 47 5) Bodor, JN., Rose, D. and Farley, TA., Swalm, C., Scott, SK. (2008). Neighborhood fruit and vegetable availability and consumption: the role of small food stores in an urban environment. Public Health Nutrition, 11(4), 413-20. Retrieved from: http://www.ncbi.nlm.nih.gov/pubmed/17617930 6) French, S., Story, M. and Jeffery, R. (2001). Environmental Influences on Eating and Phsyical Activity. Annual Review Public Health, 22, 309-35. Retrieved From: http://www.annualreviews.Org/doi/pdf/l 0.1146/annurev.publhealth,22.1.309 7) Tuckermanty Et. al. (2009). Access to Affordable and Nutritious Food - Measuring and Understanding Food Deserts and Their Consequences: Report to Congress. United States Department o f Agriculture, Administrative Publication No. (AP-036), 160pp. Retrieved From: http://ers.usda.gov/publications/ap-administrative-pubIication/ap036.aspx#.Uzi9I-Ko4vt 8) Kaufman. P., et al. (2012). Access to Affordable and Nutritious Food: Updated distance to supermarkets using 2010 data. United States Department o f Agriculture, Economic Research Report No. (ERR-143), 54 pp. Retrieved From: http://www.ers.usda.gov/publications/err-economic-researchreport/err 143. aspx#.Uzi AROKo4vs 9) Dutko, P., Ploeg, M. and Farrigan, T. (2012). Characteristics and Influential Factors of Food Deserts. United States Department o f Agriculture, Economic Research Report Number 140. Retrieved from: http://www.ers.usda.gov/media/883903/errl40.pdf 28 10) Guthrie, J., Lin, B., Okrent, A. and Volpe, R. (2013). Americans’ Food Choices at Home and Away: How do they compare with recommendations? United States Department o f Agriculture, Food Choices and Health Retrieved from: http://ers.usda.gov/amber-waves/2013-februarv/americans-food-choices-at-home-andaway. aspx#.Uz8kHuKo4vu 11) Casagrande, S. (2010). Walkability, healthy food availability, and the association with obesity and diabetes is Baltimore city, Maryland. Dissertation Abstracts International: Section B The Sciences and Engineering, Volume 71 (1-B'), pp. 255. Retrieved from: http://web,b.ebscohost.com/ehost/detail?sid=40241d3f~91da-4aad-a75982b651c2d9e4%40sessionmgrl 11 &vid=l &hid=l 13&bdata=JnNpdGU9ZWhvc30tbG12 ZO%3d%3d#db=psvh&AN=2Ql 0-99140-254 12) Inglis, V., Ball, K. and Crawford, D. (2005). Why do women of low socioeconomic status have poorer dietary behaviors than women of higher socioeconomic status? A qualitative exploration. Appetite, Volume 45 (3), pp. 334-343. Retrieved from: http://web.b.ebscohost.com/ehost/detail?sid=deldlec2-bblf-45d7-83d91ce21ee82976%40sessionmgrl 12&vid=l &hid=l 13&bdata=JnNpdGU9ZWhvc30tbG12 ZQ%3d%3d#db=psyh&AN=2005-l 5989-016 13) Kirkpatrick, S. and Tarasuk, V. (2003). The Relationship between low income and household food expenditures in Canada. Public Health Nutrition, Volume 6(6), pp. 58997. Retrieved from: http://www.ncbi.nlm.nih.gov/pubmed/14690040 14) Darmon, N. and Drewnowski, A. (2008). Does Social Class Predict Diet Quality? The American Journal o f Clinical Nutrition, Volume 87(5), 1107-1117. Retrieved from: http://aicn.nutrition.Org/content/87/5/1107.full 15) Caraher, M. and Lang, T. (1999). Can’t cook, won’t cook: A review of cooking skills and their relevance to health promotion. International Journal o f Health Promotion and Education, Volume 37(3), pp. 89-100. Retrieved from: http://openaccess.city.ac.uk/502/ 16) Powell, L., Chaloupka, F. and Bao, Y. (2007). The Availability of Fast-Food and Full Service Restaurants in the United States: Associations with neighborhood characteristics. American Journal o f Preventative Medicine, Volume 33(4), pp. S240-S245. Retrieved from: http://www.sciencedirect.com/science/art.icle/pii/S074937970700431X 17) Riedpath, D., et al. (2002). An ecological study of the relationship between social and environmental determinants of obesity. Health and Place, Volume 8(2), pp. 141-145. Retrieved from: http://www.sciencedirect.eom/science/article/pii/S 1353829201000284 18) No author given. (2012). Commodity Consumption by Population Statistics. United States Department o f Agriculture. Retrieved from: http://ers.usda.gov/dataproducts/commoditv-consumptiop-bv-population-characteristics.aspx#.Uz81d-Ko4vv 19) McDonald’s USA Nutrition Facts for Popular Menu Items. McDonald’s Website. 29 Retrieved April 2, 2014 from: http://nutrition.mcdonalds.com/getnutrition/nutritionfacts.pdf 20) Applebee’s Nutritional Information. Applebee’s Website. Retrieved April 2, 2014 from: http://www.applebees.eom/~/media/docs/Applebees Nutritional Info.pdf 21) Story, M., et al. (2007). Creating Healthy Food and Eating Environments: Policy and Environmental Approaches. Annual Review Public Health, Volume 29, pp 253-72. Retrieved from: http://www.annu.alreviews.org/doi/pdf710.1146/annurev.publhealth.29.020907.090926 22) Bowman, S., and Vinyard, B. (2004). Fast Food Consumption of U.S. Adults: Impact on Energy and Nutrient Intakes and Overweight Status. Journal o f the American College o f Nutrition, Volume 23(2) pp 163-168. Retrieved from: http://www.tandfonline.eom/doi/ftill/l 0.1080/07315724.2004.10719357#.Uz85zuKo4vt 23) Fitzgerald, A. and Veugelers, P. (2005). Prevalence of and risk factors for childhood overweight and obesity. CMAJ, Volume 173(6), pp. 607-613. Retrieved from: http://www.cmai.ca/content/173/6/607. short 24) Story, M., Kaphingst, K. and French, S. (2006). The Role of Schools in Obesity Prevention. Journal Issue: Childhood Obesity, Volume 16(1). Retrieved from: http://futureofchildren.org/publications/ioumals/article/index.xml?ioumalid=36&artieleid =98&sectionid=608 25) Delva, J., O’malley, P. and Johnston, L. (2007). Availability of more-healthy and less-healthy food choices in American Schools: A national study of grade, racial/ethnic, and socioeconomic differences. American Journal o f Preventative Medicine, Volume 33(4), pp. S226-S239. Retrieved from: http://web.b.ebscohost.com/ehost/detail?sid=0448cc8c-d282-48c6-beae7ee6d4b 10e3b%40sessionmgr 115 &vid= 1&hid= 113&bdata=JnNpdGU9ZWhvc30tbG12 ZQ%3 d%3d#db=ps vh&AN=2007-14813-012 26) Story, M., Kaphingst, K., Glanz, K. and Robinson-O’brien, R. (2007). Creating Healthy Food and Eating Environments: Policy and Environmental Approaches. Annual Review o f Public Health, Vol. 29, pp 253-272. Retrieved from: http://www.annualreviews.org/doi/full/10.1146/annurev.publhealth.29.020907.09Q926 27) French, S., Story, M., Neumark-Sztainer, D., Fulkerson, J. and Hannan, P. (2001) Fast food restaurant use among adolescents: associations with nutrient intake, food choices, and behavioral and psychosocial variables. International Journal o f Obesity and Related Metabolic Disorders, Volume 25(12), pp. 1823. Retrieved from: http://web.b.ebscohost.com/ehost/detail?sid=c3786268-6186-4427-bee8e6d 1f8103 91 e%40sessionmgr 115&vid=l &hid= 113 &bdata=JnNpdGU9ZWhvc3 OtbG12Z 0%3d%3 d#db=aph&AN=8 853176 30 28) Morrison, R., Mancino, L., and Variyam, J. (2011). Will Calorie Labeling in Restaurants Make a Difference? United States Department of Agriculture, Food Choices and Health. Retrieved from: http://ers.usda.gov/amber-waves/2Qll-march/will-calorielabeling.aspx#.Uz9GreKo4vt 29) Fuss, P. et al. (2012). Overeating in America: Association between Restaurant Food Consumption and Body Fatness in Healthy Adult Men and Women Ages 19 to 80. Obesity Research, Volume 7(6), pp. 564-571. Retrieved from: http://onlmelibrarv.wiley.eom/doi/l 0.1002/i. 1550-8528.1999,tb00715.x/full 30) Smith, T. and Variyam, J. (2010). Americans Are More Realistic About the Quality of Their Diets, United States Department o f Agriculture, Food Choices and Health. Retrieved from: http://ers.usda.gov/amber-waves/2010-march/americans-are-morereaIistic-about-the-qualitv-of-their-diets.aspx#.Uz9Jv-Ko4vu 31) Drewnoski, A. (2004), Obesity and the Food Environment: Dietary Energy Density and Diet Costs. American Journal o f Preventative Medicine, Volume 27(3) pp. 154-162. Retrieved from: http://www.sc.iencedirect.com/science/article/pii/S0749379704001503 32) (2014). Official USDA Food Plans: Cost of Food at Home at Four Levels, U.S. Average, January 2014. United States Department o f Agriculture. Retrieved from: http://www.cnpp.usda.gov/Publications/FoodPlans/2Q14/CostofFoodJan2014.pdf 33) Jetter, K., and Cassady, D. (2006). The Availability and Cost of Healthier Food Alternatives. American Journal o f Preventative Medicine, Volume 30(1), pp. 38-44. Retrieved from: http://www.ncbi.iibm.nih.gOv/pubmed/l 6414422 34) Grant, D., and Maxwell, S. (1999). Food Coping Strategies: A Century on from Rowntree. Nutrition Health, Volume 13(2), pp.45-60. Retrieved from: http://www.ncbi.plm.nih.gOv/pubmed/l 0453450? dopt=Abstract 35) Darmon, N., Ferguson, E. and Briend, A. (2002). A Cost Constraint Alone Has Adverse Effects on Food Selection and Nutrient Density: An analysis of human diets by linear programming. The American Society for Nutritional Sciences, Volume 132(12), pp. 3764-3771. Retrieved from: http://in.nutrition.org/content/132/12/3764.short 36) Huang, K,, and Lin, B. (2000). Estimation of Food Demand and Nutrient Elasticities from Household Survey Data. Economic Research Service, United States Deparment of Agriculture. Retrieved from: http://ageconsearch. umn. edu/bitstream/3 3579/l/tb001887.pdf 37) Powell, L. (2009). Food prices and obesity: Evidence and policy implications for taxes and subsidies. Milbank Quarterly, Volume 87(1), pp. 229-257. Retrieved from: http://web.a.ebscohost.com/ehost/detail?sid=9226a2ae-1553-4ecb-aca056fl)lba44e91%40sessionmgr4003&vid=l&hid=42-12&bdata=JnNpdGU9ZWhvc3QtbGl 2ZO%3 d%3d#db=psvh&AN=2009-03561 -010 31 38) Todd, J., and Lin, B. What Role Do Food and Beverage Prices have on Diet and Health Outcomes. United States Department o f Agriculture, Food Choices and Health. Retrieved from: http://ers.usda.gov/amber-waves/2012-september/what-role-do-foodand-beverage-prices.aspx#.UOBl -Ko4vt 39) Sharpe, L. (2013). U.S. Obesity Rate Climbing in 2013. Gallup, Well-Being Index. Retrieved from: http://www.gallup.eom/poll/l 65671 /obesitv-rate-climbing-2013.aspx 40) Todd, J. (2014). Changes in Eating Patterns and Diet Quality among Working-Age Adults, 2005-2010. United States Department o f Agriculture, Economic Research Report, No. (ERR-161). Retrieved from: http://ers.usda.gov/publications/err-economic-researchreport/errl 61 .aspx#.U0Bqb-Ko4vt 41) Todd, J. and Morrison, R. (2014). Less Eating Out, Improved Diets, and more Family Meals in the wake of the Great Recession. United States Department o f Agriculture, Food Choices and Health. Retrieved from: http://ers.usda.gov/amber-waves/2014march/less-eating-out-improved-diets,-and-more-fam..ily-meals-i.n-the~wake-of-the~greatrecession.aspx#.U0BrI-Ko4vt 42) Brunner, E., Marmot, M., and Martikainen, P. (2003). Socioeconomic differences in dietary patterns among middle-aged men and women. Social Science and Medicine, Volume 56 (7)pp. 1397-1410. Retrieved from: http://www.sciencedirect.com/science/article/pii/S0277953602001375 43) Turrel, G. et al. (2003). Measuring socio-economic position in dietary research: is choice of socio-economic indicator important? Public Health Nutrition, Volume 6 (2), pp, 191-200. Retrieved from: http://ioumals.cambridge.org/action/displavAbstract?fromPage=online&aid=567312 44) Grabner, M. (2012). BMI Trends, Socioeconomic Status, and Choice of Dataset. Obesity Facts, Volume 5, pp. 112-126. Retrieved from: http ://www .karger.com/Article/Pdf/337018 45) Stewart, If., Blisard, N., and Jollife, D. (2006). Let’s Eat Out americans weight taste, convenience, and nutrition. United States Department o f Agriculture, Economic Information Bulletin Number 19.