This item has the following downloads:
1 THE RELATIONSHIPS AMONG THE BUILT ENVIRONMENT, HEALTH BEHAVIORS, QUALITY OF LIFE, AND WEIGHT STATUS IN OVERWEIGHT AND OBESE RURAL CHILDREN By MEGAN J. COHEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Megan J. Cohen
3 To my husband, Matthew Cohen. Thank you for believing in me.
4 ACKNOWLEDGMENTS I would like to express my sincerest appreciation to Dr. David Janicke for his mentorship I would also thank all past and current members of the University of Florida Pediatric Psychology Lab for their contributions to this project, as well as their character building friendship. Finally, I would unwavering support and love.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Statement of the Problem ................................ ................................ ....................... 11 Background ................................ ................................ ................................ ............. 11 The Shift Towards an Obesogenic Environment ................................ .............. 13 Ecological Models of Health ................................ ................................ ............. 14 The Built Enviro nment, Physical Activity, and Obesity ................................ ..... 16 Residential Population Density, Physical Activity and Obesity ......................... 16 Availability of Recreational Space, Physical Activity and Obesity ..................... 19 The Built Environment, Dietary Intake, and Obesity ................................ ......... 19 Availability of Convenience Stores and F ast Food, Dietary Intake, and Obesity ................................ ................................ ................................ .......... 20 Availability of Food Markets, Dietary Intake, and Obesity ................................ 21 Using Indices to Measure the Impact of the Built Environment on Health ........ 22 Rural Health Disparities ................................ ................................ .................... 23 Rural Built Environment and Barriers to Healthy Behaviors ............................. 23 Defining Rural ................................ ................................ ................................ ... 25 Summary of Limitations in the Built Environment Literature ............................. 25 Health Related Quality of Life ................................ ................................ ........... 26 Measurement of Environmental Variables ................................ ........................ 27 Geographic Information Systems ................................ ................................ ..... 28 Determining Spatial Scale Using GIS ................................ ............................... 29 Purpose of Study ................................ ................................ ................................ .... 30 Aims and Hypotheses ................................ ................................ ............................. 30 Individual Dietary Factors ................................ ................................ ................. 30 Individual Physical Activity Factors ................................ ................................ ... 31 Weight Status ................................ ................................ ................................ ... 31 Quality of Life ................................ ................................ ................................ ... 32 Rural Obesogenic Environment Index ................................ .............................. 32 2 METHODS ................................ ................................ ................................ .............. 40 Description of the Par ent Project ................................ ................................ ............ 40 Identifying Rural Areas ................................ ................................ ............................ 41 Participants ................................ ................................ ................................ ............. 41 Procedure ................................ ................................ ................................ ............... 42
6 Recruitment ................................ ................................ ................................ ...... 42 Measures ................................ ................................ ................................ .......... 43 Demographic, health behavior and anthropometric data ........................... 43 Child completed measures ................................ ................................ ......... 43 Parent completed measures ................................ ................................ ...... 44 Geospatial measures ................................ ................................ ................. 45 Statistical Analysis Plan ................................ ................................ .......................... 46 Preliminary and Geographical Analyses ................................ ........................... 46 Preliminary Analysis of Demographic Data and Health Behaviors ................... 46 Primary Analyses ................................ ................................ ............................. 47 Obesogenic Rurality Index ................................ ................................ ............... 48 3 RESULTS ................................ ................................ ................................ ............... 50 Aim 1. Relationship am ong unprepared food stores, prepared food stores, and dietary outcomes. ................................ ................................ ................................ 51 Aim 2. The relationship among population density, parks and recreation, and daily time spent in physical activity. ................................ ................................ ..... 53 Aim 3. Built Environment and Child BMI z ................................ .............................. 54 Aim 4 Built Environment and Quality of Life ................................ ........................... 54 Aim 5 Creating and using a rurality index as a predictor for child outcome variables. ................................ ................................ ................................ ............. 55 4 DISCUSSION ................................ ................................ ................................ ......... 68 Rural Built Environment and Dietary Intake ................................ ............................ 68 Rural Built Environment and Physical Activity ................................ ......................... 72 Rural Built Environment and Child Weight ................................ .............................. 74 Rural Built Environment and Quality of Life ................................ ............................ 76 Using an Obesogenic Rurality Index ................................ ................................ ....... 77 Limitations ................................ ................................ ................................ ............... 78 Clinical Implication s and Future Directions ................................ ............................. 80 APPENDIX: MEASURES ................................ ................................ .............................. 83 LIST OF REFERENCES ................................ ................................ ............................... 99 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 112
7 LIST OF TABLES Table page 1 1 The Built Environment, Physical Activity, and Weight Status in Children ........... 33 1 2 The Built Food Environment, Dietary Intake, and Weight Status in Children ...... 37 3 1 Demographic characteristics of the sample. ................................ ....................... 57 3 2 Built environment characteristics in one, five, and ten mile radius ...................... 58 3 3 Descriptive data for all outcome variables ................................ .......................... 58 3 4 Correlations Among Study Variables and Demographic Variables ..................... 59 3 5 Multiple Regression Coefficients Predicting Dietary Intake at 1 mile .................. 59 3 6 Multiple Regression Coefficients Predicting Dietary Intake at 5 miles ................ 60 3 7 Multiple Regression Coefficients Predicting Dietary Intake at 10 miles .............. 60 3 8 Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 1 Mile ................................ ................................ ................... 61 3 9 Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 5 Miles ................................ ................................ ................. 61 3 10 Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 10 Miles ................................ ................................ ................ 62 3 11 Multiple Regression Coefficient Predicting Child BMI z score at 1 Mile .............. 62 3 12 Multiple Regression Coefficients Predicting Child BMI z score at 5 Miles .......... 63 3 13 Multiple Regression Coefficients Predicting Child BMI z score at 10 Miles ........ 63 3 14 Multiple Regression Coefficient Predicting Child Total Quality of Life at 1 Mile .. 64 3 15 Multiple Regression Coeffi cient Predicting Child Total Quality of Life at 5 Miles ................................ ................................ ................................ ................... 64 3 16 Multiple Regression Coefficient Predicting Child Total Quality of Life at 10 Miles ................................ ................................ ................................ ................... 65 3 17 Multiple Regressio n Coefficient of Rurality Index Predicting Dietary Intake ....... 65 3 18 Multiple Regression Coefficient of Rurality Index Predic ting Average Daily Time Spent in Physical Activity ................................ ................................ ........... 66
8 3 19 Multiple Regression Rurality Index Coefficients Predicting Chi ld BMI z score ... 66 3 20 Multiple Regression Rurality Index Coefficients Predicting Child Total Quality of Life ................................ ................................ ................................ .................. 67
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE RELATIONSHIPS AMONG THE BUILT ENVIRONMENT, HEALTH BEHAVIORS, QUALITY OF LIFE, AND WEIGHT STATUS IN OVERW EIGHT AND OBESE RURAL CHILDREN By Megan J. Cohen August 2013 Chair: David M. Janicke Major: Psychology Childhood obesity rates have been on the rise in recent decades, and the toxic obesogenic environment has been indicated to play a significant role (Egger & Swinburn, 1997; Ogden, Carr oll, Kit, & Flegal, 2012) Currently, there is limited understanding of the impact of the built environment and child health behaviors (i.e. dietary intake and physical activity), weight status, and quality of life in rural areas. Yet, children from r ural areas are disproportionately overweight and obese compared to their urban and suburban counterparts (Tai Seale & Chandler, 2010) This study included 269 overweight or obese rural youth between ages 8 and 12. Using state of the art Geographic Information Systems (GIS) population density, number of parks and recreational facilities, number of prepared food stores (i.e. restaurants and convenience stores), and number of unprepared food stores (i.e. grocery stores, chain supermarkets, farmers markets) was measured within one, five, and ten mile buffers around each those who had more parks and recreational facilities within 10 miles of their home, had a lower BMI z score. No relationship was foun d among the built environment variables
10 and child dietary intake, physical activity, or quality of life. Population density, parks and recreational facilities, unprepared food stores, and prepared food stores were also compiled into an obesogenic rurality index in order to determine their joint effect on child health behaviors and weight status. However the obesogenic rurality index, showed no clinically significant association with child health behaviors or weight status. For rural families, there is some evidence that living in a rural census tract with lower population density, as well as increasing access to parks and recreational facilities may be beneficial. Yet, the effect sizes of these relationships are small, and more research regarding how childr en interact with their environment to improve health is warranted.
11 CHAPTER 1 INTRODUCTION Statement of the Problem Childhood obesity is an epidemic that has affected all social classes and races, however children living in rural areas are disproportionately overweight and obese (Tai Seale & Chandler, 2010) There is some support in the literature that characteristics of the rural environment may be the driving force behind the increased risk of obesity in rural children, but built environment factors on child behaviors and weight status have been understudied in this population (Durand, Andalib, Dunton, Wolch, & Pentz, 2011) Furthermore, there is very limited understanding of the variability within the rural environment and its impact on child weight because classification of rural is often in the literature. Background High adiposity rates in children in the United States continue to be a serious public health concern. The most recent reports estimate that 31.8% of children between the ages of 2 and 19 have a body mass index (BMI) above the 85 th percentile for age and gender, which is classified as overweight, and 16.9% of children have a body mass index above the 95 th percentile for age and gender, wh ich is classified as obese (Ogden et al., 2012) Not only are larger percentages of children overweight or obese in recent decades, children are overweight or obese to a larger degree than ever before. approximately 6.3%, but b increased by over 17% (Anderson & Butcher, 2006) Ogden and colleagues reported
12 that approximately 12.3% of children are above the 97 th percentile for age and g ender (2012) This is of great concern becau se obese children are at greater risk for short term physical consequences such as orthopedic, neurological, and endocrine conditions (Must & Strauss, 1999) Adverse psychosocial consequences, including increased body dissatisfaction (Ricciardelli & McCabe, 2001) low self esteem (Banis et al., 1988; Zeller, Saelens, Roehrig, & Kirk, 2004) social marginalization (Strauss & Pollack, 2003) increased peer victimization (Gray, Kahhan, & Jan icke, 2009) and low quality of life (Janicke et al., 2007) are also highly prevalent in overweight and obese children. Furthermore, overweight children are more likely to become overweight adults (Dietz, 1998) and adult overweight and obesity has been consistently linked to preventable early mortality (Calle, Rodriguez, & Walker Thurmond, 2003; Peeters et al., 2003) annual obesity related healthcare costs averaged $99.2 bi llion per year in the late 199 0 s (Wolf & Colditz, 1998). Taken together, it is imperative that research continues to identify factors that contribute to this epidemic. As the weight status of both children and adults has increased dramatically in recent decades (Flegal, Carroll, Ogden, & Johnson, 2002; Ogd en et al., 2012) a growing amount of research has focused on which factors underlie the dramatic change in body composition for a large proportion of the population. Weight regulation is the result of an energy balance between energy intake and energy e xpenditure, while obesity is the result of long term positive energy intake. The energy balance equation may be impacted by both endogenous and exogenous variables. Endogenous variables include endocrinological factors (e.g. hormonal regulation of basal m etabolic
13 rate), neurological factors (e.g. neurotransmitter regulation of satiety) and genetics, while exogenous variables include those things that would account for increased caloric intake and decreased physical activity energy expenditure. Epidemiolog ical studies of the changes in these endogenous factors over time do not show that human body chemistry has changed dramatically enough to account for the increased rates of obesity (Hill, Wyatt, Reed, & Peters, 2003) In fact, research indicates that less than 5% of obesity cases result from endogenous factors (Anderson & Butcher, 2006) These findings have indica ted the greater importance of exogenous factors, specifically changes towards a more obesogenic environment. The Shift Towards an Obesogenic Environment (Egger & Swinburn, 1997) has been used to describe a myriad of shifts across physical, economic and social environments, especially in Western society, that have lead to the current obesity epidemic. Throughout history humans have worked to develop technology that would ensure consistent high energy food sources, decrease need for physical labor, and increase leisure time (Hill, Wyatt, Reed, & Peters, 2003; Peters, Wyatt, Donahoo, & Hill, 2002). However, by almost eliminating the occurrence of starvation, shifting from a labor centered to information centered workf orce, moving towards increased motor vehicle usage, and increasing the availability of enjoyable sedentary leisure activities, society has sabotaged its own health (Peters, Wyatt, Donahoo, & Hill, 2002) The obesogenic shi ft in culture has also had a significant impact on infrastructure, including how people plan and build communities.
14 Ecological Models of Health As early as the 1930 s researchers and scholars have considered the environment as a key factor in human development, health, and health behaviors (Le win, Heider, & Heider, 1936) Early work in the field by Bronfenbrenner suggested that there are five environmental contexts (microsystem, mesosystem, exosystem, macosystem, and chronosystem) that affect human development and health (Bronfenbrenner, 1979) These five environmental systems incorporate built, social, political, chronological, and cultural environments that are both direct and indirect to the individual (1979). When Bronfenbrenner included levels of environments as explained influences on human health and behavior, a myriad of other, specific ecological theories were developed and applied to the fields of health psychology and public health in order ory for Health Promotion provided guiding principles for understanding the relationships among a variety of personal and environmental factors on health (1996) Wi thin Social Ecological Theory, the environment can affect health by serving as a medium for disease transmission (e.g. contaminated water), a stressor, a source of danger or safety (e.g. unsafe terrain), an enabler of health behavior (e.g. proximity to rec reational centers), or a provider of health resources (e.g. proximity to hospitals) (1996). However, the psychological state, or behavioral patterns) (1996). Sallis, Owen, a nd Fisher describe ecological models of health behavior as having four core concepts: 1. Health behavior has multiple levels of influence across intrapersonal, interpersonal, environmental, community, and policy. 2. Influences on health behaviors interact. 3. Ecological models
15 should be behavior specific when being developed to drive interventions. 4. Multi level interventions should be most effective in changing behavior. (2008) With regards to the obesogenic environments influence on the obesity epidemic, two specific ecological health mode ls have been proposed. The Ecological Model of Active Living describes five levels of influence on physical activity: Intrapersonal (e.g. demographics, family makeup, biological factors, psychological factors), Perceived Environment (e.g. attractiveness, comfort, accessibility), Behavioral Active Living Domains (e.g. recreation, active transportation, occupation, and household living), Access and Characteristics of Behavior Settings (e.g. neighborhood environment, home environment, workplace environment, a nd school environment), and Policy Environment (e.g. zoning laws, healthcare policies, average housing costs) (2008) As with other ecological models, these levels of influence on physical activity are all interrelated. On the other side of the energy equation, The Model of Community Nutrition Environments describes multiple levels of influence on individual nutrition (Glanz, Sallis, Saelens, & Frank, 2005) These levels of in fluence include individual levels, such as eating preferences and behaviors, and sociodemographic variables (e.g. psychosocial factors, and perceived nutrition environment), as well as environmental and policy variables. Regarding environmental variables, Glanz and colleagues outline the importance of the Organizational Nutrition Environment (e.g. eating at home, school, work, or other), the Community Nutrition Environment (e.g. types and locations of food outlets, accessibility of food outlets), the Consu mer Nutrition Environment (e.g. product promotions, nutrition labels), and the Information Environment (e.g. media and advertising). These two guiding theories point out that there are multiple influences on behavior to be studied,
16 as well as multiple lev els of potential intervention. Much of the literature has focused on a variety of potential influences within the built environment as put forth by these models. The Built Environment, Physical Activity, and Obesity Since the 1990 s much attention has been given to how the built environment has impacted adult and child energy expenditure. Starting in the second half of the 20 th century, a movement towards less physical activity and more sedentary activity has been recorded in both adults and children (Brownson, Boehmer, & Luke, 2005) A primary reason for this shift is likely due to the cultural and economic changes following W orld War II (Hill & Peters, 1998). The economic and population boom of the 1950 s meant that more and more families were able to own an automobile, a television set, and a house with more property (1998). Urban planning followed suit, and American neighbo rhoods began to spread out and revolve around automobile traffic versus pedestrian traffic. This changed the American landscape by decreasing residential population density and separating residential spaces from commercial, recreational and industrial spa ces (1998). This means that many Americans have access to fewer destinations within walking or cycling distance than they did fifty years ago, greatly limiting active transport (1998). Residential Population Density, Physical Activity and Obesity Residential population density refers to the concentration of dwellings within a defined area (e.g. a square mile). Residential population density has been hypothesized to impact physical activity and obesity across multiple ecological levels by determini ng destinations within walking distance (Saelens, Sallis, Black, & Chen, 2003)
17 as well as proximity to potential social outlets and social support for physical activity (Greiner, Li, Kawachi, Hunt, & Ahluwalia, 2004) In order to capture the effect the changing landscape has had on physical ac tivity, researchers have investigated residential population density, along with other factors such as street connectivity and land use, within indices, such as urban sprawl (Ewing, Brownson, & Berrigan, 2006) and walkability (Saelens, Sallis, Black, & Chen, 2003) in order to estimate their joint effect on physical activity and overall active transport. The literature shows that adults who live in areas with high urban sprawl (i.e. low residential population density, low land use mix, and poor street connectivity) are more likely to weigh more, walk less, and have higher rates of hypert ension (Ewing, Schmid, Killingsworth, Zlot, & Raudenbush, 2003). These results have been replicated in other studies, and there has been support for the impact of sprawl even when controlling for personal barriers to physical activity (Joshu, Boehmer, Brownson, & Ewing, 2008) Conceptually the opposite of sprawl, neighborhoods with high walkability (Owens et al., 2007; Saelens, Sallis, Black, & Chen, 2003) and lower body mass index (Heinrich et al., 2008; Owen et al., 2007; 2003) including in disadvantaged neighborhoods (2008) Neighborhood walkability has also been associated with leisure time activity, and residents in low walkability neighborhoods tend to watch more television than residents in high walkability neighborhoods (Sugiyama, Salmon, Dunstan, Bauman, & Owen, 2007) While the study of sprawl and walkability has increased understanding o f the built
18 that these variables have undergone limited investigation with children and rural populations. Using ecological models of active living, it stands to reas on that these populations would also be affected by the sprawl and walkability of their communities. There have been a few studies related to children. Much of the research on sprawl as it relates to adult behavior and weight status has been conducted by Ewing and colleagues. However when they conducted a similar study in children they discovered sprawl was even more strongly associated with overweight in children than adults (Ewing, Brownson, & Berrigan, 2006). More research has been conducted with chi ld and adolescent health behaviors and community walkability. Merchant and colleagues reported that when two Canadian communities were assessed for parent perceptions of s edentary than children living in the other community (Merchant, Dehghan, Behnke Cook, & Anand, 2007) Within a sample of California adolescents, community walkability was positively related to number of minutes spent in physical activity (Kligerman, Sallis, Ryan, & Frank, 2007) In a 2008 stud y of Canadian preschoolers the odds of being overweight was lower for girls if they lived in walkable neighborhoods versus those (Spence, Cutumisu, Edwards, & Evans, 2008) Table 1 1 includes a list of these studies concerning the built environment, physical activity, and weight status in children, including the impacts of sprawl and w alkability. While it can be argued population density over time, the influence of low population density on physical activity may still be an important factor.
19 Availabili ty of Recreational Space, Physical Activity and Obesity The availability of facilities and parks for recreational physical activity has also impacted physical activity and weight status. In adults, support has been found for the association between physica l activity, obesity and availability of space for physical activity such as parks and recreational facilities (Bjrk et al., 2008; Boehmer & Lovegreen, Haire Joshu, Brownson, 2006; Ellaway, Macintyre, & Bonnefoy, 2005; Giles Corti & Donovan, 2003; Heinrich et al., 2008; Mobley et al., 2006; Mujahid et al., 2008; Nielsen & Hansen, 2007; Poortinga, 2006; Saelens, Sallis, Black, & Chen, 2003) The availability of physical activity f acilities and parks has been more frequently studied in children than other domains within the built environment, and these studies can be found on Table 1 1 Within the literature, child physical activity levels and weight status have been associated with parent perceptions of facility access, child perceptions of facility access, and observational measures (i.e. GIS derived) of access. Variables schoolyards, available greenspa ce (e.g. parks), trails, and beaches. The majority of these studies have been conducted either in entirely in urban samples or in national samples where rural youth are underrepresented. The Built Environment, Dietary Intake, and Obesity The American die t has made big changes in recent decades. In a review of four National Health and Nutrition Examination Surveys, in 2004 the average American adult consumed over 500 more calories per day than the average American adult in 1970 (Briefel & Johnson, 2004) Perhaps th e most significant impetus for the changing diet is the increased reliance on and availability of prepared convenience foods. The schedules and structure for the 21 st century American family has included fewer parents
20 in a homemaker role, more single pare nts, more two working parent households, and longer work days for parents (Anderson & Butcher, 2006) This changing work culture has called for an increased demand in quick, prepared foods. For the first time in hi story, highly palatable, energy dense, low nutrient, low cost foods are widely available to everyone in the population through fast food, packaged products, vending machines and convenience stores ( Anderson & Butcher, 2006; McCrory, Fuss, & Saltzman, 2000; McDermott & Stephens, 2010) Not only are these foods energy dense, but the portion sizes of these foods are up to five times larger than they were 40 years ago (McCrory, Fuss, & Saltzman, 2000) Availability of Convenience Stores and Fast Food, Dietary Intake, and Obesity The connection between changes in the food environment and changes in the weight status of the population has been gaining increased attention since the last part of the 20 th century. This has lead to sudden growth within the literature, and several studies have supported the relationship between the built food environme nt and dietary behaviors and weight status. Associations between obesity and availability of poor quality food sources (i.e. fast food restaurants and convenience stores) has been found in both adults and children, across varying spatial scales (Fleischhacker, Evenson, D. A. Rodriguez, & Ammerman, 2010; Galvez, Hong, Choi, Liao, & Godbold, 2009; Jilcott et al., 2011; Laska, Hearst, Forsyth, Pasch, & L. Lytle, 2010; Mehta & Chang, 2008) On a large scale level, states with high fast food restaurant density per person (ac ross the ages) also have higher obesity rates than states with low fast food restaurant density per person (Maddock, 2004) This relatio nship has also been found when considering smaller geographic areas, including county level fast food restaurant availability (Mehta & Chang, 2008) and census tract fast food restaurant availability (Bodor, Rice, Farley,
21 Swalm, & Rose, 2010) The focus in the food environment literature has largely been on adults versus children. And the literature regarding convenience food access and children lags behind the literature co ncerning the physical activity environment and children in terms of the number of studies that have been conducted. Yet, it is likely that convenience food access has a significant impact and deserves increased attention. A large scale epidemiological st udy of child fast food consumption shows that children who regularly eat fast food consume more calories, more fat, more carbohydrates, more added sugars, more sugar sweetened beverages, and fewer fruits and vegetables than children who do not regularly ea t fast food (Bowman, Gortmaker, & Ebbeling, 2004) So, if these foods are readily available, the regular cons umption of fast food is likely to increase. The few studies that have been conducted in this area are outlined in Table 1 2. Overall, child weight status has been more frequently associated with the presence of convenience stores versus fast food restaur ants, and the majority of studies have been conducted in urban settings where convenience stores are likely to be within walking distance. Availability of Food Markets, Dietary Intake, and Obesity The obesity epidemic is, in part, driven by the high preval ence of food outlets for energy dense, large portioned, convenience food. Conversely, there is also evidence that places with low availability of high quality food (e.g. fruits and vegetables) via supermarkets and grocery stores are prone to higher risk o f obesity in the population. In studies of communities and food store locations, rural areas and low income minority neighborhoods consistently have a higher ratio of convenience stores to chain supermarkets, which has been hypothesized to be partly respo nsible for the health disparities in these areas (Larson, Story, & Nelson, 2009) However, the literature
22 shows that as access to healthy foods in these areas increase, consumption of healthy foods also increase (2009) In several large scale studies, increasing distance from a associated with higher BMI in adults, even when controlling for sociodemographic factors (Inagami, Cohen & Finch, 2006; Larson, Story, & Nelson, 2009; Morland & Roux, 2006) These patterns have also been seen in children and adolescents, which are also outlined in Table 1 2. So far the literature shows that as children have increasing access to chain s upermarkets they are at lower risk for overweight, even when controlling for household income level. Using Indices to Measure the Impact of the Built Environment on Health It is important to understand the relationships of individual environmental variable s and health outcomes in order to determine the relative importance of each all of these variables are likely to interact and have a joint effect on health outcomes. Indices of the built environment can be useful when studying these joint effects, especially given that they can control for multi collinearity of variables (Ewing, Handy, Brownson, Clemente, & Winston, 2006) Furthermore, indices provide researchers with the ability to parsimoniously report their results. This is especially true if the environmental variables under consideration have limited variability, which may impede (Lytle, 2009) Two specific indices have been previously highlighted: the county sprawl index and the walkability index. The county sprawl index developed by Ewing and colleagues takes into account residenti al population density, land use mix and street connectivity to capture the impact of urban sprawl on health behaviors and weight (2003). Similar to sprawl, walkability of communities (a composite of high dwelling density, mixed land use, and high
23 intersect ion density) has also been measured to capture the net influence of the built environment on physical activity and weight (Saelens, Sallis, Black, & Chen, 2003) and this index has shown to have go od construct validity when using both GIS environmental variables (Leslie, Coffee, Frank, Owen, Bauman, & Hugo, 2007) and perceived environmental variables (Saelens Sallis, Black, & Chen, 2003) Rural Health Disparities Almost all communities across the United States have seen a shift towards a more obesogenic environment. However, some communities are at particularly high risk. Rural America provokes images of locally grown fresh foods, room for children to play and be active, and work that is characterized by hard labor. Yet, rural children have a 25% higher risk of obesity than their metropolitan counterparts (Lutfiyya, Lipsky, Wisdom Behounek, & Ipanbutr Martinkus, 2007; Williamson et al., 2009) In re cent decades, rural children have been shown to eat diets higher in dietary fat and calories, engage in exercise less frequently, and watch more television (Tai Seale & Chandler, 2010) Furthermore, rural children have less access to routine health care (2010) Th is may be, in part, due to obesogenic factors like population density Rural Built Environment and Barriers to Healthy Behaviors children is significantly underrepresented in the literature. This is not surprising given that research with rural populations is limited in general, often due to lack of f unding and barriers to participant recruitment (Cudney, Craig, Nichols & Weinert, 2004; Lim, Follansbee Junger, Crawford & Janicke, 2011). Yet, people living in rural areas are considered part of a unique cultural group, so generalizing research findings f rom urban and suburban populations may not be appropriate. This is especially true considering
24 built environment variables, which are at the core when differentiating what it means to resented, built environment and health research focusing on the heterogeneity of rural environments is absent. This is often due to the overgeneralization of the classification of rural as a residual (i.e. not urban) (Brown & Cromartie, 2003) which will be la ter discussed. The literature in rural areas suggests that the rural environment poses significant barriers to achieving a normal weight status in adults. Using perceived measures of environmental variables, Boehmer and colleagues (2006) and Casey and co lleagues (2008) found that low accessibility of grocery stores, fast food only dining options, limited physical activity facilities, and limited destinations were all associated with obesity in rural adults. Similarly, when comparing rural children to non rural children perceived lack of access to recreational facilities, high availability of energy dense foods, and limited destinations or social outlets were all reported and related to weight status or physical activity levels (Moore et al., 2010; Simen Kapeu, Khule, &, Veugelers, 2010; Yousefian, Ziller, Swartz, & Hartley, 2009) Furthermore, a study by Liu and colleagues suggests that r ural children experience the influence of the built environment differently than their suburban and urban counterparts, and increased distance to supermarkets was related to increased obesity only in rural areas (2007) These seminal studies suggest that the rural built environment may have a significant impact on the health behaviors and weight status of rural residents, but more research is needed in this area. This is especially true given that the majority of these studies measured the perceived built environment versus the objective built environment. It is
25 also unknown whether rural children differ from one another based on the differences in their built environment. Defining Rural There are multiple definitions of rural, even according to the U.S. Census Bureau. The most commonly used definition for research purposes includes the O ffice area includes a core county urban area of 50,000 people or more, with adjacent metropolitan areas heavily integrated as measured by commuting to work and having a po pulation of at least 10,000 (Hart et al. 2005). Therefore, the OMB defines rural as any area that does not meet these criteria. Given this rather broad definition of rural, it is not surprising that rural areas in research are rather heterogeneous. Depe nding on their proximity to metro areas, rural areas have the potential to have a wide range of resources. For example, one small remote town may have a population so small it ay be act would receive the same resources set aside for rural communities (Hall, Kaufman, & Ricketts, 2006) The heterogeneity of rural areas as defined by the U.S. Census Bureau calls for investigation of the varying degrees of rurality as it rela tes to the toxic obesogenic environment. Summary of Limitations in the Built Environment Literature Interest in the built environment and childhood obesity is a relatively recent phenomenon with the majority of papers on this topic published after 2006. As outlined in the previous sections, a primary limitation in the literature has been that most studi es
26 have concentrated on urban and suburban settings, with very few studies looking at these relationships in rural settings (Durand, Andalib, Dunton, Wolch, & Pentz, 2011) However, there are several other limitations within the literature that have not been addressed. For example, it is notable that the majority of studies, to date, have investigated the relationship of the built environment and health behaviors or the built environment and weight status (Papas et al., 2007) Very few studies have investigated the relationships among the built environment, health behaviors, and BMI, which is important in discovering the mech anisms in which the built environment may ultimately impact weight status. Much of the literature thus far has focused primarily on adults, with significantly fewer studies determining the relationship of the environment and child health behaviors and weig ht status (Glanz & Sallis, 2006) Finally, there are other indicators of health as r elated to weight status that have not been investigated in relation to the built environment. One such health indicator is health related quality of life. Health Related Quality of Life Health related quality of life (QOL) is multidimensional construct that includes the physical, psychological, social, and occupational/educational functioning of an individual (Spieth & Harris, 1996) QOL differs from other measures of wellbeing in that it describes difficulties in functioning secondary to health condit ions. Many pediatric chronic medical conditions have been shown to significantly impact QOL, including pediatric obesity (Ingerski et al., 2010.) In a review of QOL in obese youth, an overall inverse relationship was found between QOL and BMI (Tsiros et al., 2009) and obese youth have been found to be particularly impaired in their physical and social functioning when compared to non obese youth (2009). Above and beyond weight status, it is
27 important to determine additional factors that impact QOL in overweight and obese youth. The association between the built environment and health not only includes physical health, but mental and social health as well. For example, social capital (e.g. community involvement) has been found to be higher in communiti es that are considered walkable (Leyden, 2003) Studies have also shown that poor environmental conditions put individuals at increased risk for poor psychological functionin g (Evans, 2003; Galea, Ahern, Rudenstine, Wallace, & Vlahov, 2005) There has been so me investigated the impact on total QOL, especially in children. Measurement of Environmental Variables One of the primary concerns recognized by researchers in the fi eld has been the inconsistency between study variables and the continued development of methods to measure environmental variables. In general, researchers have measured either bles (Boehmer, Hoehner, Deshpande, Brennan Ramirez, & Brownson, 2007) Measurement of perceived variables has entailed asking study participants to rate their access to a variety of structures in the built environment (e.g. grocery stores, parks, r ecreational facilities, walking trails). On the other hand, many researchers have begun to utilize geographic information systems (GIS; described in the following section), which has allowed for easy access to objective measures of aspects of the built en vironment in relation to study participants (Liu, Cunningham, & Downs, 2002; Wieczorek & Delmerico, 2009) Associations betwee n the built environment and health have been found using both perceived and observed measures, however the implications of the findings may differ depending on the measures used (Boehmer, Hoehner, Deshpande,
28 Brennan Ramirez, & Brownson, 2007) For example, findings using perceived measures may indicate that increased awareness of resources should be targeted versus changing the built environment. For the purpose of accuracy in capturing the environment, more and more researchers are turning to GIS. Geographic Information Systems According to the Environmental Systems Research Institute (ESRI), a managing, analyzing and displaying all forms of geographically refere nced (Wieczorek & Delmerico, 2009) All data in a G IS has a geographic component, linking each data point to a specific geographic point on the earth. The geographic information in a GIS is represented using Cartesian coordinates (e.g. latitude and longitude) (2009). A GIS goes beyond traditional maps by enabling researchers to encode geographic information in multiple map layers. Each map layer is a river, a park, a well, or an entire continent. These features can be r epresented in one of three geometrical forms: polygons, lines, or points (Ornsby, Napoleon, Burke, Groessl, & Bowden 2010). Each feature in a map layer can be linked to an extensive ple, a map earth. The attributes of these countries may include information such as average rainfall, population, highest elevation, gross national product, as well as any other characteristic one may want to know about each country. When map layers are compiled, each feature, along with all of its attributes, can be compared to other features that share the same geographic location. For example, a map layer contai ning
29 the attributes of the schools (e.g. public versus private). This layer may then be compiling these map layers, a researcher would be able to determine the ratio of public to private schools in each county in the state of Florida. Any information that can be linked to a geographic location on earth can be represented in a GIS. The proces s of adding geographically represented information to a GIS is known as geocoding (Wieczorek & Delmerico, 2009) In the United States, any street address can be geocoded by referencing the U.S. Census Topologically Integrated Geographic Encoding and Reference (TIGER) (2009). TIGER is a publically available geographic dataset containing all geographically linke d variables in the U.S. Census. This includes U.S. Economic Census information regarding potential obesogenic built environment variables such as location of fast food restaurants, grocery stores, parks, and healt h facilities. Determining Spatial Scale Using GIS Using GIS to measure environmental variables as they relate to health is a relatively new application of this technology. Spatial scales have varied widely within the literature, from state wide boundaries (Maddock, 2004) to approximations of the smaller neighborhood environment, which may be more relevant to health behaviors and weight sta tus (Franco, Diez Roux, Glass, Caballero, & Brancati, 2008; Larson et al., 2009; Richardson, Boone Heinonen, Popkin, & Gordon Larsen, 2011; Saelen, Sallis, Black, & Chen, 2003) Perhaps the most precise m easures of the neighborhood include radial buffers (Laska et al., 2010) Radial buffers allow for a boundary to include the
30 Having a uniform spatial scale across p articipants is an advantage to using buffers. However, there is little guidance in the literature as to what radial distances would best investigated within the fram ework of this study. Purpose of Study The purpose of this study was to determine the impact of the obesogenic built environment on child health behaviors, weight status, and quality of life for children in rural areas. This study was nested within a larger study of a rural lifestyle intervent ion for children who are overweight or obese. This study addressed gaps in the literature in several ways. First, this study was the first to investigate the impact of the rural environment within a heterogeneous rural sample. Second, it expanded the li terature on built environment and child health behaviors and weight status, which has been understudied compared to the built environment and adult health behaviors and weight status. Third, the impact of the environment on both health behaviors and weigh t status was investigated. Fourth, state of the art Geographic Information Systems were utilized in order to accurately measure environmental variables. This study investigated the individual contributions of each variable in the rural environment to chil d behaviors, QOL, and weight status. Finally, this study developed and assessed the validity of a new index, consisting of these environment variables that assess the rural obesogenic environment. Aims and Hypotheses Individual Dietary Factors Aim 1: To determine the impact of the built food environment on individual dietary factors.
31 Hypothesis 1a: Children with greater access to prepared food stores (i.e. convenience stores and restaurants) will have higher caloric intake. Hypothesis 1b: Children with g reater access to prepared food stores will have higher percent caloric intake from fat. Hypothesis 1c: Children with greater access to prepared foods stores will have lower fruit and vegetable intake. Hypothesis 1d: Children with greater access to unprep ared food stores (i.e. chain supermarkets, nonchain supermarkets, grocery stores) will have lower caloric intake. Hypothesis 1e: Children with greater access to unprepared food stores will have lower percent caloric intake from fat. Hypothesis 1f: Children with greater access to unprepared food stores will have higher fruit and vegetable intake. Indiv idual Physical Activity Factors Aim 2: To determine the impact of aspects of built environment on individual physical activity factors. Hypothesis 2a: Children living in areas of lower population density will have less time spent in physical activity. Hypothesis 2b: Children with access to fewer recreational facilities and parks (as a combined variable) will have less time spent in physical activity. Weight Stat us Aim 3: To determine the impact of the built environment on weight status. Hypothesis 3a: Children living in areas of lower population density will have a higher degree of overweight. Hypothesis 3b: Children with access to fewer recreational facilities and parks will have a higher degree of overweight. Hypothesis 3c: Children with access to more prepared food stores will have a higher degree of overweight. Hypothesis 3d: Children with access to more unprepared food stores will have a lower degree of overweight.
32 Quality of Life Aim 4: To determine the impact of the rural obesogenic environment on health related quality of life. Hypothesis 4a: Children living in areas of lower population density will h ave lower health related quality of life. Hypothesis 4b: Children with access to fewer recreational facilities and parks will have lower health related quality of life. Hypothesis 4c: Children with access to more prepared food venues will have lower healt h related quality of life. Hypothesis 4d: Children with access to more unprepared food stores will have higher health related quality of life. Rural Obesogenic Environment Index Aim 5: To assess the validity of the rurality index. Hypotheses 5a: Ruralit y index scores of each participate will be positively demonstrating convergent validity. Hypothesis 5b: Great rurality will be positively related to weight status. Hypothesis 5c: Gre at rurality will be positively related to calorie intake Hypothesis 5d: Greater rurality will be positively related to percent intake from fat. Hypothesis 5e: Greater rurality will be negatively related to fruit and vegetable intake. Hypothesis 5f: Greater rurality will be negatively related to time spent in physical activity. Hypothesis 5g: Greater rurality will be negatively related to health related quality of life.
33 Table 1 1 The Built Environment, Physical Activity, and Weight Status in Children Authors Study Population Measurement of Environment Findings Crawford and colleagues 2010 301 children ages 10 12 in Australia GIS measurement of land use, road connectivity, and traffic exposure within a 2 km radius Perceived measures of safety. Low traffic and perceived safety were associated with more moderate to vigorous physical activity. Moore and colleagues 2010 50 children from urban and rural communities in North Carolina Qualitative measurement of access to physical activity facilities Both rural and urban children reported low access to physical facilities as a barrier to physical activity. Oreskovic and colleagues 2010 21,008 children ages 2 18 living in Massachusetts GIS information on distance to and density of s treets, sidewalks, schools, subway stations, bike paths, green space, fast food restaurants, and population density. Distance to and density of fast food restaurants, distance to subway stations, amount of open space, and population density all associated with BMI in hypothesized directions. Simen Kapeau, Kuhle, & Veugelers 2010 3,421 fifth grade students in Alberta, Canada Parent perception of the neighborhood environment including access to stores, access to physical activity facilities, and safety Child ren from rural areas had lower access to physical activity facilities and had higher incidence of overweight. Yousefian and colleagues 2009 84 Rural children ages 10 17 in Maine Qualitative measurement of physical activity environment via focus groups Chi ldren reported lack of facilities, transportation, and destinations as barriers to physical activity. Bell, Wilson, & Liu 2008 3831 children ages 3 16 in Indiana Satellite derived measure of areas Youth living in greener neighborhoods had lower BMI z scores Spence, Cutumisu, Edwards & Evans 2008 501 preschoolers ages 3 5 in Edmonton, Canada Calculated walkability using GIS derived residential population density, land use mix, intersection density, and availability of facilities. Odds of being overweight or obese were lower for girls living in more walkable neighborhoods.
34 Table 1 1. Continued Authors Study Population Measurement of Environment Findings Evenson and colleagues 2007 1554 sixth grade girls in Arizona, California, South Carolina, Louisiana, Minnesota, and Maryland Perceived accessibility of physical activity facilities were assessed Perceived access to physical activity facilities was associated with lower BMI. Kligerman and colleagues 2007 98 adolescents ages 14 17 in California GIS was used to calculate a walkability index using land use mix, residential density, intersection density, and retail floor space. Number of parks, recreational facilities and beaches were also captured. Walkability was positively related to number of minutes in physical activity. Liu, Wilson, Qi, &Ying 2007 7,334 children ages 3 18 in Indiana Neighborhood vegetation and proximity to food stores were calculated with GIS Vegetation was related to decreased risk for overweight in urban areas, a nd increased distance to a supermarket was related to increased risk for overweight in rural areas. Merchant, Dehgan, Behnke Cook, & Anand 2007 160 elementary children in Canada Walkability of two neighborhoods was assessed via a parent questionnaire Chil dren in the neighborhood sedentary than neighborhood. Scott and colleagues 2007 1,556 sixth grade girls in Arizona, California, South Carolina, Louisiana, Minnesota, and Maryland Schools wer e classified as having locked or unlocked schoolyards. These schools were geocoded and distance of home residence to schools was measured with GIS. Proximity to fewer unlocked schoolyards was related to increased BMI. However, there was no relationship b etween schoolyards and physical activity levels. Ewing, Brownson, & Berrigan 2006 6,811 youth ages 12 17 in a national sample County sprawl was calculated via an index including residential density, land use mix, degree of centering, and street accessibility. Sprawl was related to higher levels of obesity in US youth.
35 Table 1 1. Continued Authors Study Population Measurement of Environment Findings Gordon Larsen, Nelson, Page, & Popkin 2006 20,754 adolescents in a national sample Presence of physical activity and recreational facilities in each census block was measured with GIS Lower SES block groups were less likely to have facilities. Controlling for SES, increasing numbers of facilities was associated with lower BMI and more moderate vig orous physical activity. Nelson, Gordon Larsen, Song, & Popkin 2006 20,754 adolescents in a national sample Neighborhood characteristics were measured with GIS and categorized into six distinct neighborhood types based on SES and density Children living i n more dense, older suburbs were more likely to be active than children living in less dense newer suburbs. Rural children were also shown to be less active. Ward and colleagues 2006 1,105 girls with a mean age of 14.6 from South Carolina Availability of recreational facilities were measured with a questionnaire There was no association between environmental variables and BMI in girls. Timperio, Salmon, Telford, & Crawford 2005 291 families of children ages 5 6 and 919 families of children ages 10 12. Parent perceptions of walking distance to destinations and traffic patterns. No parent perceptions of the environment were associated with 5 6 year old weight status. Parent perception of heavy traffic was associated with increased risk of obesity in 10 1 2 year old children. Burdette & Whitaker 2004 7,020 children between 36 and 59 months in the WIC database in Cincinnati, OH GIS determined the distance between child residence and nearest playground and nearest fast food restaurant. Neighborhood safety w as determined based on the number of police reported crimes was calculated for each Cincinnati neighborhood. No environmental variables were associated with child weight status.
36 Table 1 1. Continued Authors Study Population Measurement of Environment Findings Liu, Cunningham, & Downs 2002 2,554 children between 4 and 18 living in Indiana GIS calculated proximity between child residence and 4 types of public play space: YMCA programs, city parks, city trails, after school programs There was no relation ship between weight status and play space above and beyond SES Romero, Robinson, & Kraemer 2001 796 fourth grade students in Northern California Parents completed telephone interviews rating neighborhood hazards including traffic, trash, crime, lack of park access, and drugs. More hazards were related to more physical activity in children, suggesting the relationship between physical activity and environment is complex.
37 Table 1 2 The Built Food Environment, Dietary Intake, and Weight Status in Children Authors Study Population Measurement of Environment Findings Jilcott and colleagues 2011 744 children ages 2 18 in North Carolina GIS measurement of food environment: convenience stores, fast food restaurants, grocery markets, supermarkets, and other restaurants. BMI was negatively associated with positively associated with convenience stores and fas t food restaurants. Laska and colleagues 2010 349 adolescents ages 11 18 from Minneapolis/St. Paul GIS measurement of the food environment including distance to and density of fast food restaurants, any restaurants, convenience stores, and grocery stores within a buffer. BMI and body fat percentage was associated with presence of a convenience store within 1600 m. Oreskovic and colleagues 2010 21,008 children ages 2 18 living in Massachusetts GIS information on distance to and density of streets, sidewalks, schools, subway stations, bike paths, green space, fast food restaurants, and population density. Distance to and density of fast food restaurants, distance to subway stations, amount of open space, and population density all associated with BM I in hypothesized directions. Galvez and colleagues 2009 323 urban children ages 6 8 Number of convenience stores census block was measured via GIS Children living next to more convenience stores were more likely to be overweight
38 Table 1 2. Continued Authors Study Population Measurement of Environment Findings Timperio and colleagues 2008 340 children ages 5 6 and 461 children ages 10 12 in Australia markets, supermarkets, convenience stores, fast food outlets and t akeout restaurants were derived in GIS The more fast food and convenience stores close to the home, the lower likelihood children would consume at least 2 fruits of vegetables during the day. The likelihood children would consume more than 3 fruits or vegetables during the day was greater the farther children lived from fast food restaurants. Liu, Wilson, Qi, &Ying 2007 7,334 children ages 3 18 in Indiana Neighborhood vegetation and proximity to food stores were calculated with GIS Vegetation was relat ed to decreased risk for overweight in urban areas, and increased distance to a supermarket was related to increased risk for overweight in rural areas. Powell and colleagues 2007 73,079 children in 8 th and 10 th grade drawn from a national sample. GIS measurement of small grocery stores, chain supermarkets, nonchain supermarkets, and convenience stores were collected Increased availability of chain supermarkets was associated with lower adolescent BMI, while increased availability of convenience stores was associated with higher BMI Sturm & Datar 2005 6,916 children recruited in kindergarten and followed through third grade Number of grocery stores (associated with low fruit and vegetable prices) and convenience stores (associated with high fruit and vegetable prices) were measured for each Children living in areas with more convenience stores than grocery stores (i.e. higher fruit and vegetable prices) were at higher risk for overweight.
39 Table 1 2. Continued Authors Study Pop ulation Measurement of Environment Findings Burdette & Whitaker 2004 7,020 children between 36 and 59 months in the WIC database in Cincinnati, OH GIS determined the distance between child residence and nearest playground and nearest fast food restaurant. Neighborhood safety was determined based on the number of police reported crimes was calculated for each Cincinnati neighborhood. No environmental variables were associated with child weight status.
40 CHAPTER 2 METHODS This study is a cross sectional analysis of the relationships among rural obesogenic environmental variables, diet, physical activity, weight status, and quality of life in a sample of rural overweight and obese youth. This study is part of a larger study investigating t he effectiveness of a group based behavioral weight loss intervention for overweight and obese youth living in rural areas, The Extension Family Lifestyle Intervention Project for Kids (E FLIP for Kids). Description of the Parent Project E FLIP for Kids is a randomized controlled trial investigating the impact of two lifestyle interventions (parent only and family based) on changes in child weight status as compared to a health education control group. Participating parents and children in the family based condition and the health education condition met in separate but simultaneous groups, while only parents met in a formal group in the parent only condition. All families participate in 21 group meetings at a local Cooperative Extension Services office ov er the course of one year, with the first 8 meetings occurring once per week, the next 4 meetings occurring biweekly, and the remaining meetings occurring once per month. The lifestyle interventions seek to help families improve nutrition through the use o f the Stoplight System developed by Epstein (Wilfley et al., 2007) incr ease physical activity, and manage weight. Lifestyle intervention families engage in goal setting, self monitoring, and other behavioral strategies in order to improve their overall health. Families participating in the health education condition receive information across a wide array of health topics, but they do not engage in behavioral strategies to manage weight. All families participating in the program attended a
41 baseline assessment visit and a post treatment assessment visit one year later. Meas ures from the baseline assessment were used. Identifying Rural Areas For the purpose of the group based behavioral weight loss intervention, rural counties were identified in North Central Florida based on the Office of ions (Hart et al. 2005). As previously described, the OMB classification of rural can yield a rather heterogeneous research sample, which increased the variability of the rural obesogenic environment for the purposes of the current study. Participants Thi s study included 269 overweight and obese youth between the ages of 8 and 12 years and their parent or legal guardian, living in rural counties in North Central Florida. Families were included if the child had a body mass index at or above the 85 th percen tile for age and gender norms published by the CDC (Kuczmarski et al. 2000). Parents or legal guardians lived in the same home with the child in a rural county in North Central Florida, and were age 75 years or less. Children or parents/legal guardians w ere excluded if they had: any medical conditions that contraindicate participating in a program requiring modifications of caloric intake or physical activity; health conditions such as a resting blood pressure of 140/90 mm Hg; a medication regimen includi ng antipsychotic agents, monoamine oxidase inhibitors, systemic corticosteroids, antibiotics for HIV or tuberculosis, chemotherapeutic drugs, or use of prescription weight loss drugs within six months of starting the intervention. Participants could not b e participating in another weight loss program. Furthermore, the following conditions or behaviors that would likely affect the conduct of the trial were
42 excluded: parents unable to give informed consent, parents who are non English speaking, parents unab le to read at an 8 th grade reading level, children with major cognitive or developmental delays, children with patterns of aggressive or extreme oppositional behavior. Procedure Recruitment Participants for this study were recruited via a variety of direct solicitation methods, including both widespread and culturally tailored for minority populations. These included direct mailings to households and health care providers, brochures distributed through local schools, press releases in local communities, pr esentations given to local groups and churches, and brochure distribution by Extension agents and research team members at local events and community locations. All recruitment materials included a toll free phone number, and interested families were enco uraged to contact the research team in order to begin initial phone screening. The purpose of the initial phone screen was to provide prospective participants with information about the study and to determine initial eligibility prior to scheduling an in person assessment. Families who met initial eligibility and expressed interest were scheduled for an in person screening visit to determine final eligibility. Families were asked not to join another formal weight loss program or use weight loss medicatio n during the course of the study. At the initial screening visit families completed informed consent and assent protocol and measures of height and weight to determine eligibility. Families that met eligibility at the initial in person screening visit we re scheduled for a baseline assessment visit approximately one to two weeks prior to the start of the intervention
43 program. The following information was gathered from children and their parent at the baseline visit. Measures Demographic, health behavior and anthropometric d ata The following parent and child measures were collected at two initial visits prior to starting the program. A trained nurse or nurse technician collected anthropometric measures. A trained research assistant collected questionnaire data. Child complete d m easures Height and Weight. Height was measured to the nearest 0.1 centimeter using a Harpendon stadiometer, and weight was measured to the nearest 0.1 kilogram using a digital scale. Children wore only one layer of clothes, emptied their pockets, and removed shoes before having anthropometrics taken. Body mass index was calculated as kilograms per meters squared. Physical Activity. All children were asked to wear a Sensewear Armband accelerometer for seven consecutive days, 24 hours per day (except when exposed to water). The Sensewear Armband included a 2 axis accelerometer (measuring motion and steps taken), a heat flux sensor (me asuring the rate at which heat dissipates from the body), skin temperature sensor (measures body surface temperature), and galvanic skin response sensor (measures electric currents that increase when sweating) ). Taken together, the Sensewear Armband can differentiate between levels of physical activity, as well as sleep. Data gathered in metabolic equivalents (METS) from four weekdays and one weekend day were averaged to provide an objective mea sure of average total time spent in daily physical activity (greater than 3 metabolic equivalents) ( Adams
44 Caparosa Thompson Norman, 2009) Children and parents were provided with verbal and written instructions on proper usage of this accelerometer. Dietary Intake. preceding week was assessed using the Block Kids 2004. This 77 item questionnaire was developed from the NHANES 1999 2002 dietary recall data, and the nutrient database was developed from the USDA Nutriti on Database for Dietary Studies, version 1.0. The Block Kids 2004 has shown good reliability and validity as compared to 24 hour diet recalls (Cullen, Watson, & Zakeri, 2008) Total caloric intake, percent caloric intake from fat, and total fruit and vegetable intake were used for the purpose of this study. Children and parents worked together to complete this questionnaire. Health Relat ed Quality of Life. The PedsQL is a 23 item scale that measures health related quality of life across physical and psychosocial functioning (with subscales in emotional, social and school functioning). This measure has strong internal consistency, clinica l validity, and factor analytic support for its subscales (Varni, Seid, & Kurtin, 2001) The PedsQL has been utilized with populations of healthy children, as well as children with chronic and acute conditions, including pediatric obesity (Ingerski et al., 2010; Janicke et al., 2007; Zeller, 2006) The total quality of life was then calculated from the four subscales and used in the subsequent analyses. Parent completed m easures Demographic Information This questionnaire obtains family background information such as: parent/child age, parent/child gender, parent/child race, marital status, education, occupation, and family income. Home address. Parents provided their c urrent home and mailing address during the initial screening. The home address was used for the purpose of this study.
45 Geospatial m easures Geospatial data including Florida map information, parks and recreational space information, and population density were collected from the Florida Geographic Data Library (FGDL). The FGDL is an online warehouse of over 350 GIS data layers compiled approximately 35 local, state, federal and private agencies ( www.fgdl.org ). The FGDL is maintained by the University of F screened for quality by GIS professionals and students. All data housed by the FGDL can be downloaded and used without cost. All restaurant, convenience store, and grocery store information was obtained from InfoUSA a nd geocoded using ArcGIS 10.0 software. Food Environment. The number of grocery stores, convenience stores, fast food restaurants, and dining restaurants for each county in the study was collected from the InfoUSA and geocoded. Given that various aspects of the food environment have impacted child weight status differentially, there are two total food scores. A total convenience stores, fast food restaurants, and dining re staurants available within a one variable included the sum of all grocery stores and supermarkets within a one mile, five mile and ten mile radius of the home residence. Physical Activity Environment. Point locations for parks and recreational facilities were downloaded via the FGDL. The total number of both parks and recreational facilities were summed within the one mile, five mile and ten mile radius of the home resi dence.
46 Statistical Analysis Plan Prelim inary and Geographical Analyses Normality of all dependent variables were initially assessed by examining skewness and kurtosis of the distribution. Skewness and kurtosis values between negative two and positive two were considered adequate (Cameron, 2004). Any non normal variables underwent logarithmic transformation. All participant residences were geocoded by assigning Cartesian mathematical coordinates (e.g. latitude and longitude) to each street address. Geocod ed addresses were compiled to create a map layer in the GIS. Then, all individual level participant health parameter data were merged into a tabular database within ArcGIS with the geocoded addresses and matched to each address. One mile, five mile, and te n mile radial buffer zones were formed around each and map layers were created. Each geographical set of variables were joined to form a map layer. All map layers were com bined into one tabular database and matched to the Social Sciences Version 21.0. Preliminary Analysis of Demogr aphic Data and Health Behaviors Pearson correlations in vestigated the relationships among individual demographic variables (i.e. child age, child gender, child race, parent income) and health behaviors and weight status (i.e. total caloric intake, percent calories from fat, daily servings of fruit and vegetabl es, time spent in physical activity, total quality of life, and child baseline BMI z score). Significant associations were controlled for using hierarchical regression in subsequent analyses.
47 Primary Analyses Aim 1. The first aim was to determine the impa ct of the built food environment on individual dietary factors. Separate hierarchical linear regressions were conducted for outcome variables at one, five, and ten mile buffer zones. The first block contained any significantly associated demographic varia bles, and the second block contained the total number of prepared food stores and unprepared food stores in the buffer zones. Total calorie intake, percent caloric intake from fat, and total fruit and vegetable intake from the BLOCK Kids served as the ind ividual dependent variables for each hierarchical regression at each of the three buffer zones. Aim 2. The next aim was to determine the impact of the built environment on child physical activity. Three hierarchical regressions were conducted for one, fi ve, and ten mile buffer zones. Significant demographic variables were entered into the first block of each regression, and population density and total number of recreational facilities and parks were entered in the second block. Total average time spent in daily physical activity (greater than 3 metabolic equivalents) based on accelerometry served as the dependent variable. Aim 3. The third aim was to determine the impact of individual elements of the built environment on weight status. Separate hierarch ical regressions were conducted at the one, five, and ten mile buffer zones. Significant demographic covariates were entered into the first block, population density, total parks and recreational facilities, prepared food stores, and unprepared food store s were entered into the second blocks of each regression. Child BMI z score was the dependent variable. Aim 4. The fourth aim was to determine the impact of individual elements of the built environment on total quality of life. Separate hierarchical re gressions were
48 conducted at one, five, and ten mile buffer zones. Significant demographic covariates were entered into the first block, population density, total recreational facilities, prepared food stores, and unprepared food stores were entered in the second blocks. Child self reported total health related quality of life was the dependent variable. Obesogenic Rurality Index Aim 5. The fifth aim will be to determine the impact of the overall obesogenic environment on child health and weight status. In order to capture the joint effect of the individual built environment variables, an obesogenic rurality index was calculated based on the built environment at the 10 mile radius. The obesogenic rurality index included four variables that have been sho wn to be obesogenic and have the best representation in available databases for the state of Florida: 1. Population density of the census tract. 2. Total prepared food establishments of the buffer zones 3. Total unprepared food store establishments of the buffer zo nes 4. Total physical activity opportunities of the buffer zones. Each of the calculated values were normalized using a z score such that a value of 1 would indicate that the value is one standard deviation above the mean for each of the variable categories. The total obesogenic rurality index was then the inversed sum of the z scores from the four measures. Higher scores on the obesogenic rurality index indicate higher obesogenic rurality as characterized by low population density, fewer prepared food estab lishments, fewer unprepared food store establishments and fewer physical activity opportunities. In order to establish convergent validity of the obesogenic rurality index, this index was compared to another index of rurality, the Montana State University (MSU) Rurality Index (Weinert & Boik, 1995) The MSU Rurality Index was developed in order
49 to capture the variability of rurali ty of participants in research studies, and it has been used widely in the rural research literature (Weinert & Boik, 1995) T he MSU Rurality Index takes into account a variable describing the population density of the area surrounding the participant, and distance to the nearest medical center describes the s home residence to the nearest hospital was calculated. To calculate the MSU Rurality Index for each participant, the distribution of population densities and the distance to the nearest hospital underwent z score transformation. These two scores were su mmed for the MSU Rurality index score. To assess convergent validity, a Pearson correlation was determined between the MSU Rurality Index score and the obesogenic rurality index score. In order to determine the relationship of overall rurality to dietary factors, physical activity factors, child weight status, and quality of life separate hierarchical regressions were be conducted. In each hierarchical regression significant demographic covariates were entered into the first block, and the obesogenic rura lity index were entered into the second block. Total calorie intake, percent calories from fat, fruit and vegetable intake, total time spent in physical activity, BMI z scores, and total quality of life served as dependent variables in each regression.
5 0 C HAPTER 3 RESULTS Four hundred six parents completed the phone screening for the E FLIP for Kids study. Of these, 398 were scheduled for an in person screening visit, and 314 attended this visit. Two hundred seventy two child parent dyads attended a baselin e assessment. Of these, three children were excluded because either child or parent did not complete questionnaires properly and the data was unable to be used. Therefore, a total of 269 children (122 males, 147 females) between the ages of 8 and 12 ( M= 9.8 8 years, SD= 1.41) were included in the analyses for the current study. All children were overweight or obese with the average body mass index falling two standard deviations above the CDC normed population mean ( M= 2.19 SD=. 37 ). More than two thirds of the sample identified as Caucasian while the next highest racial group was African American (13%). Approximately 12% of the sample identified their ethnicity as Hispanic. There was variability in family income, which ranged from below $19,999 to above $100,00 0. Demographic data can be found in Table 3 1 There was significant variability regarding the rural built environment that stores, and parks and recreational facilities were summed within a one, five, and ten mile radius (buffer zone) of each home, and the mean and standard deviation for these variables at each buffer zone can be found on Table 3 2 Population density was unable to be measured within concentric buffer zones, so this was measured at the census tract level. Population density ranged from less than one person per acre to 12 people per acre with an average of 1.73 ( SD= 2.17) people per acre for the census tract of residence.
51 Total daily calories, daily servings of fruits and vegetables, and average number of minutes spent in daily physical activity were all outside of the parameters of normality, therefore logarithmic transformation was utilized to correct the distribution of these variables. Transformed distributi ons all had skewness and kurtosis values within acceptable limits (Cameron, 2004). Descriptive statistics for all outcome variables can be seen in Table 3 3 Pearson correlations for individual demographic data, built environment data and outcome data wer e conducted to investigate the rel ationships among the variables, and these results are displayed in Table 3 4 Demographic var iables that were significantly correlated with at least one outcome variable were controlled for in subsequent regression models examining that specific outcome variable These included child age and family income. Aim 1. Relationship among unprepared food stores, prepared food stores, and dietary outcomes. The first aim was to determine the association between the built environmen t on three dietary outcome variables obtained from the BLOCK Kids 2004 survey: average daily caloric intake, average daily percent calories from fat, and average daily servings of fruits and vegetable. Across the sample, reported daily caloric intake range d from 156.6 to 4566.6 calories per day, percent calories from fat ranged from 18.5% to 57.6%, and servings of fruits and vegetables ranged from 0 to 12.2. Of note, approximately 50% of the sample reported their average daily caloric intake as less than 12 00 calories, 25% of the children in the sample reported their average daily caloric intake as less than 800 calories per day, and 10% reported their daily caloric intake as less than 600 calories per day, which likely suggests that underreporting of calori es was common on the BLOCK questionnaire. According to www.choosemyplate.gov sedentary children
52 between 8 and 12 years old require between 1200 and 1800 calories per day. Therefore, it is unlikely that the childr en who reported less than 1200 calories per day were giving an accurate response. Within the literature, it is estimated that individuals typically under report daily calories by approximately 20% (Lichtman et al. 1992; Westerterp, Verboeket van de Venne, Meijer, Hoor ten, 1991). Thus, only the 162 children who indicated that they consumed over 960 were used for the dietary outcome analyses. Children included in the analyses did not differ significantly by demographics, weight status, or built environment f eatures compared to those excluded from analyses based on low calorie reports. To address this aim, multiple regressions were conducted to determine whether the number of prepared food stores and unprepared food stores were associated with each of the thr ee dietary outcome variable (i.e. average daily caloric intake, daily percent calories from fat, average daily servings of fruits and vegetables) within the one, five, and ten mile buffer zones. Results indicated that the number of unprepared food stores a nd prepared food stores were not associated with daily caloric intake at a one mile radius ( R 2 =.028 F= 2.163), five mile radius ( R 2 =.005 F =.379), or ten mile radius ( R 2 =.009 F= .708). Unprepared food stores and prepared food stores were also not associated with daily percent calories from fat in the one mile radius model (R 2 =.001, F =.072), the five mile radius model ( R 2 =.001 F =.090), and the ten mile radius model (R 2 =.000, F =.03 4). Likewise, unprepared food stores and prepared food stores were not associated with daily servings of fruits and vegetables in the one mile radius m odel ( R 2 =.0 16 F = 1.207), the five mile radius model ( R 2 =.001 F =.132 ), and the ten mile radius model ( R 2 = .0 06 F = .470). Results displaying multiple regression coefficients for
53 the models of one mile variables predicting all three dietary outcomes are displayed in Table 3 5 Multiple regression coefficients for models of five mile variables predicting dietary outcomes are displayed in Table 3 6 And, multiple regression coefficients for models of 10 mile variables predicting dietary outcomes are displayed in Table 3 7 Aim 2. The relationship among population density, parks and recreation, and daily time spent in physical activity. In order to calculate the average total daily time spent in physical activity, participants needed to wear the Sensewear armband for at least four weekdays and one weekend day ( Adams Caparosa Thompson Norman, 2009) Of the 269 pa rticipants included in this study, only 166 participants provided sufficient accelerometer data. Independent samples t tests were conducted for each demographic variable and BMI z score in order to determine if there were any differences between participan ts who wore the armband for the required length of time versus those who did not. No significant differences were found between groups. Of the participants who wore the armband, the average total daily time spent in physical activity (>3 METS) was calcula ted for each child, and it ranged from 5.75 minutes in physical activity per day to 431 minutes in physical activity per day with a mean of 112 ( SD= 70.2) minutes spent in physical activity. In correlational analysis, no demographic variables were related to time spent in physical activity. Therefore, population density was entered into one block along with the total number of parks and recreational facilities at buffer zones of one, five, and ten miles, and average daily time spent in physical activity ser ved as the dependent variable. Multiple regressions revealed no association among population density or parks and recreational facilities and average daily time spent in physical activity at one mile ( R 2 =.083 F =.795), five miles ( R 2 =.082 F =. 724 ), or ten miles
54 ( R 2 =.0 95 F= 1.778). Multiple regression coefficients for the one mile model are displayed in Table 3 8 the five mile model are displayed in Table 3 9 and the ten mile model are displayed in Table 3 10 Aim 3. Built Environment and Child BMI z scores ( r = .142 p <.0 5). Therefore family income was entered into the first block of all hierarchical regressions predicting child BMI z score. The second block included population density, u nprepared food stores, prepared food stores, and parks and recreational facilities. Separate models were run for one, five, and ten mile buffer zones, and each included the population density of the census tract. Results showed that the one mile radius model ( R 2 =.048 F =2.449) and the five mile radius model ( R 2 =.053 F =2.721) were not significant; multiple regression co efficients are shown in Tables 3 11 and 3 12 However, the ten mile radius model yielded significant results ( R 2 =.076 F =3.990), as rep orted in Table 3 13 Specifically, children who had a greater number of parks and recreational facilities within 10 miles of their home weighed less ( = .189, p=. 003), and children living in areas of higher population density weighed more (( =.135, p= .039) Neither unprepared food stores nor prepared food stores were associated with child BMI z in the ten mile model. The model accounted for 7.6% of the variance in child BMI z score. Aim 4 Built Environment and Quality of Life The total score from the PedsQL was used to determine the relationship between elements of the built environment and baseline quality of life. The total score of the PedsQL ranged from 30.4 to 100.0 with a mean score of 7 6.0 ( SD = 13.8 ). Multiple regression analyse s were conducted for all built environment variables (i.e. population
55 density, unprepared food stores, prepared food stores, and parks and recreational facilities) and their associations with total quality of life at one, five, and ten mile buffer zones ar ound R 2 = .0 10 F =. 521 ,), five ( R 2 = .0 17 F =. 899 ) and ten ( R 2 = .0 18 F = .950 ) mile models were all not significant. Multiple regression coefficients for the one, five, and ten mile models are displayed in Table 3 14 through 3 16 Aim 5 Creating and using a rurality index as a predictor for child outcome variables. The fifth aim was to create a rurality index by combining all built environment variables within the ten mile buffer zone and examine whether the combination of these va riables predicted child health behaviors and weight status. Population density, prepared food stores, unprepared food stores, and parks and recreational facilities all underwent z transformation, and these z index score. Index scores were multiplied by negative one in order to denote that higher scores indicated higher levels of rurality. Rurality index values ranged from 8.80 to 4.19 ( M= 0.10, SD= 2.77). An alternative rurality index, the Montana State Rurality index ( Weinert & Boik, 1995) was then calculated for each participant and correlated with the obesogenic rurality index. The MSU Rurality Index takes into account the population density, and distance to the nearest medical center describes the remoteness residence to the nearest hospital ranged from 0.5 miles to 40.7 miles with an average of 13.2 miles ( SD = 7.87). To calculate the MSU Rurality Index for each participant, the distribution o f population densities and the distance to the nearest hospital underwent z score transformation, and these two z scores were summed for the MSU Rurality index
56 score. MSU Rurality scores ranged from 6.71 to 2.30 ( M= .05, SD= 1.44). Correlational analysis r evealed significant relationship between the MSU index and the obesogenic rurality index ( r =.546, p= .000), establishing convergent validity None of the multiple regressions investigating the association between the rurality index and all outcome variables (i.e. daily caloric intake, daily percent calories from fat, daily servings of fruits and vegetables, daily average minutes spent in physical activity, child BMI z score, and total quality of life) were significant. The results of these analyses are displ ayed in Tables 3 17 through 3 20.
57 Table 3 1 Demographic characteristics of the sample. Mean SD Child Age 9.88 1.41 BMI z score 2.19 .37 % Gender Boys 45.4 Girls 54.6 Child Race Caucasian 68.3 African American 13.4 American Indian .7 Latino .4 Asian .4 Multiracial 11.9 Not Reported 4.9 Child Ethnicity Hispanic 11.9 Not Hispanic 88.1 Family Income Brackets Below $19,999 19.1 $20,000 $39,999 $40,000 $59,000 $60,000 $79,999 $80,000 $99,999 29.6 22.5 13.9 7.5 Over $100,000 7.5
58 Table 3 2 Built environment characteristics in one, five, and ten mile radius One Mile Five Miles Ten Miles Range Mean(SD) Range Mean(SD) Range Mean(SD) Unprepared food stores 0 6 0.3(0.9) 0 12 3.3(3.3) 0 27 7.6(5.7) Prepared Food Stores 0 37 3.0(6.7) 0 132 33.8(34.2) 0 320 81.8(75.9) Parks and Recreation 0 5 0.4(0.9) 0 77 7.9(10.2) 0 148 24.6(23.6) Range Mean(SD) Population Density of Census Tract <1 12 1.73(2.17) Table 3 3 Descriptive data for all outcome variables Range M(SD) Daily Caloric Intake 156.6 4566.6 1255.1(672.6) Daily Percent Calories from Fat 18.5% 57.6% 33.1(5.9) Daily Servings of Fruits and Vegetables 0 12.2 2.1(1.7) Average Daily Minutes Spent in Physical Activity 5.75 431.0 112.0(70.2) Child BMI z score 1.21 3.00 2.19(.37) Total Quality of Life 30.4 100.0 76.0(13.8)
59 Table 3 4. Correlations Among Study Variables and Demographic Variables Child BMI z Total Calories % Calories from Fat Fruit /Veg. Serv. Min. of Physical Activity Total Qo L Pop. Density Prepared F ood a Unprepared Food a Parks and Rec a Child Age .015 .080 .033 .017 .237** .043 .002 .121 .116 .017 Child Gender .092 .010 .059 .012 .089 .020 .077 .149* .097 .011 Child Race .067 .001 .078 .052 .144 .006 .016 .050 .048 .044 Total Yearly Family Income .142* .019 .043 .042 .071 .112 .107 .049 .081 .036 p <0.05 level ** p <.01 (2 tailed) a Built Environment density at the 1 mile buffer zone Table 3 5. Multiple Regression Coefficients Predicting Dietary Intake at 1 mile Total Daily Calories Percent Calories from Fat Fruit and Vegetable Intake Variable t p f R 2 t p f R 2 t p f R 2 Built Environment 2.163 .0 28 .0 72 .00 1 1.207 .0 1 6 1 Mile Unprepared Food Stores 268 1. 984 049 .001 .010 992 209 1.541 125 1 Mile Prepared Store s 266 1. 968 .0 51 .032 .232 817 185 1.359 176 is standardized beta weight *p is significant at <.0 5 R 2 includes all previously entered blocks
60 Table 3 6. Multiple Regression Coefficients Predicting Dietary Intake at 5 miles Total Daily Calories Percent Calories from Fat Fruit and Vegetable Intake Variable t p f R 2 t p f R 2 t p f R 2 Built Environment 379 .00 5 090 .00 1 411 .00 5 5 Mile Unprepared Food Stores .0 97 602 .5 48 014 088 930 122 759 449 5 Mile Prepared Store s .0 32 203 840 046 285 776 .0 65 405 686 is standardized beta weight *p is significant at <.0 5 R 2 includes all previously entered blocks Table 3 7. Multiple Regression Coefficients Predicting Dietary Intake at 10 miles Total Daily Calories Percent Calories from Fat Fruit and Veget able Intake Variable t p f R 2 t p f R 2 t p f R 2 Built Environment .708 .0 09 034 .00 0 .470 .0 06 10 Mile Unprepared Food Stores 219 1.182 239 023 122 903 .1 55 .835 405 10 Mile Prepared Store s 209 1. 125 .2 62 039 211 833 179 .966 335 is standardized beta weight *p is significant at <.0 5 R 2 includes all previously entered blocks
61 Table 3 8 Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 1 Mile Average Time Spent in Physical Activity (>3METS) Variable t p f R 2 Block 1 Demographics 11.617 .073 Child Age .271 3.408 .001 Block 2 Built Environment .795 .083 Population Density .085 1.057 .292 1 Mile Parks and Recreation .067 .837 .404 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 9 Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 5 Miles Average Time Spent in Physical Activity (>3METS) Variable t p f R 2 Block 1 Demographics 11.617 .073 Child Age .271 3.408 .001 Block 2 Built Environment .724 .082 Population Density .084 1.047 .297 5 Mile Parks and Recreation .060 .748 .456 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
62 Table 3 10. Multiple Regression Coefficients Predicting Average Daily Time Spent in Physical Activity at 10 Miles Average Time Spent in Physical Activity (>3METS) Variable t p f R 2 Block 1 Demographics 11.617 .073 Child Age .271 3.408 .001 Block 2 Built Environment 1.778 .095 Population Density .086 1.090 .277 10 Mile Parks and Recreation .130 1.629 .105 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 11. Multiple Regression Coefficient Predicting Child BMI z score at 1 Mile Child BMI z score Variable t p f R 2 Block 1 Demographics 5.214 .021 Total Family Yearly Income .143 2.283 .023 Block 2 Built Environment 2.449 .048 Population Density .112 1.660 .098 1 Mile Unprepared Food Stores .094 .833 .405 1 Mile Prepared Stores .204 1.791 .075 1 Mile Parks and Recreation .037 .549 .584 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
63 Table 3 12. Multiple Regression Coefficients Predicting Child BMI z score at 5 Miles Child BMI z score Variable t p f R 2 Block 1 Demographics 5.214 .021 Total Family Yearly Income .143 2.283 .023 Block 2 Built Environment 2.721 .053 Population Density .124 1.811 .071 5 Mile Unprepared Food Stores .163 1.269 .205 5 Mile Prepared Stores .235 1.822 .070 5 Mile Parks and Recreation .101 1.543 .124 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 13. Multiple Regression Coefficients Predicting Child BMI z score at 10 Miles Child BMI z score Variable t p f R 2 Block 1 Demographics 5.214 .021 Total Family Yearly Income .143 2.283 .023 Block 2 Built Environment 3.990* .076 Population Density .135 2.077 .039 10 Mile Unprepared Food Stores .124 .858 .392 10 Mile Prepared Stores .205 1.401 .162 10 Mile Parks and Recreation .189 2.979 .003 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
64 Table 3 14. Multiple Regression Coefficient Predicting Child Total Quality of Life at 1 Mile Child Total Quality of Life Variable t p f R 2 Built Environment .521 .010 Population Density .082 1.108 .269 1 Mile Unprepared Food Stores .034 .295 .769 1 Mile Prepared Stores .048 .410 .682 1 Mile Parks and Recreation .044 .597 .551 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 15. Multiple Regression Coefficient Predicting Child Total Quality of Life at 5 Miles Child Total Quality of Life Variable t p f R 2 Built Environment .899 .017 Population Density .076 1.030 .304 1 Mile Unprepared Food Stores .055 .388 .698 1 Mile Prepared Stores .049 .351 .726 1 Mile Parks and Recreation .093 1.313 .191 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
65 Table 3 16. Multiple Regression Coefficient Predicting Child Total Quality of Life at 10 Miles Child Total Quality of Life Variable t p f R 2 Built Environment .950 .018 Population Density .086 1.229 .221 1 Mile Unprepared Food Stores .153 .985 .326 1 Mile Prepared Stores .113 .729 .467 1 Mile Parks and Recreation .069 .995 .321 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 17. Multiple Regression Coefficient of Rurality Index Predicting Dietary Intake Total Daily Calories Percent Calories from Fat Fruit and Vegetable Intake Variable t p f R 2 t p f R 2 t p f R 2 Built Environment 1.88 .01 .634 .003 3.39 .013 Rurality Index .099 1.37 .172 .050 .796 .427 .116 1.84 .067 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
66 Table 3 18. Multiple Regression Coefficient of Rurality Index Predicting Average Daily Time Spent in Physical Activity Average Time Spent in Physical Activity (>3METS) Variable t p f R 2 Block 1 Demographics 11.617 .073 Child Age .271 3.408 .001 Block 2 Built Environment .614 .077 Rurality Index .062 .784 .435 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks Table 3 19. Multiple Regression Rurality Index Coefficients Predicting Child BMI z score Child BMI z score Variable t p f R 2 Block 1 Demographics 5.214 .021 Total Family Yearly Income .143 2.283 .023 Block 2 Built Environment 2.174 .029 Rurality Index .093 1.475 .142 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
67 Table 3 20. Multiple Regression Rurality Index Coefficients Predicting Child Total Quality of Life Child Total Quality of Life Variable t p f R 2 Built Environment .652 .003 Rurality Index .055 .807 .420 is standardized beta weight *p is significant at <.05 R 2 includes all previously entered blocks
68 CHAPTER 4 DISCUSSION The current study is the first to investigate the impact of the rural built environment on child health behaviors, weight status, and quality of life. Rural research in general tends to be more limited than urban research, despite known health disparities (Lutfiyya, Lipsky, Wisdom Behounek, & Ipanbutr Martinkus, 2007; Williamson et al., 2009) The rural built environment could be contributi ng to these health disparities since rural children generally have less access to healthy foods and opportunities for physical activity than their urban counterparts (Moore et al., 2010; Simen Kapeu, Khule, &, Veugelers, 2010; Yousefian, Ziller, Swartz, & Hartley, 2009) Furthermore, much of rural classification is exclusionary (i.e. not urban), despite the potential for significant variab ility in the built environment. Therefore, this study sought to investigate built environment around their homes. Rural Built Environment and Dietary Intake It was expecte d that proximity to prepared food stores (restaurants and convenience stores) would be associated with child dietary intake such that a greater be positively associated wit h daily calories and percent intake from fat and negatively associated with fruit and vegetable intake. It was also expected that proximity to unprepared food stores (grocery stores) would be associated with child dietary intake such that a greater number of unprepared food stores within expanding distances from from fat and positively associated with fruit and vegetable intake. However, these
69 hypotheses were not supported The relationship among proximity to prepared food stores and unprepared food stores and weight status in children has received more support in the literature than the relationship among food stores and dietary intake ( Galvez, Hong, Choi, Liao, & Godbold, 2009; Jilcott et al., 2011; Laska, Hearst, Forsyth, Pasch, & L. Lytle, 2010) Indeed, the literature showing a connection between food access and dietary intake among both adults and children has been sparse (Larson, Story, Nelson, 2009). Yet, with the re lationship among food access and weight status being more consistently supported, it stands to reason that that this relationship would be accounted for by dietary intake. Larson and colleagues conducted a literature review of access to restaurants, conve nience stores, and grocery stores and dietary intake in 2009. Of the 43 studies examined, only one study found a relationship among convenience store access and lower fruit and vegetable consumption in youth, and in this same study no relationship was foun d among grocery store access and fruit and vegetable consumption. (2009). Since this review, only a few studies have supported the association between the built environment and diet. For example, Laska and colleagues found that GIS derived density of conve nience stores within 1600 meters of a home residence was associated with sugar sweetened beverage consumption in teens, but no associations were found with other dietary variables (2010). Lamichhane and colleagues found that density of supermarkets within four miles of home was associated with better dietary intake in South Carolina children with diabetes, but they found no association between dietary intake and fast food density (2012). Conversely, in a study with 8226 children and 5236 adolescents living in California, no robust relationships were found among diet and the food environment (An & Sturm, 2012).
70 With these mixed findings, it begs the question as to why these associations are not more robust. One important consideration is the outcome measures used to capture dietary intake. Although the Block Kids Food Screener has been shown to be valid compared to 24 hour dietary recalls ( ), there is good reason to believe that it did not accurately capture intake for the children in the E FLIP for Kids study. Almost 50% of children reported calorie counts below the daily recomm endations from www. ChooseMyPlate .gov Inaccuracy in self report of dietary intake in children is not uncommon in the literature. This is especially evident in overweight and obese youth. For example, Vance and colleagues found that overweight and obese youth were more likely to underreport calories compared to normal weight youth (2008). Also, Wolkoff and colleagues found that youth who have episodes of loss of control over eating are more likely to inaccurately recall their food intake (2011). Systematic underreporting could be occurring due to factors such as child response bias, inaccurate estimation of portion sizes, or simple forgetting. Another important consideration is the accuracy of our environmental variables. Grocery store, convenience store, and restaurant data was purchased from a commercial listing company and geocoded in ArcGIS 10.0. Powell and colleagues found there is only fair agreem ent between commercial listings of grocery stores, convenience stores, and restaurants and ground truthing methods for the presence of these establishments (2011). Ground truthing refers to field observation to verify the existence of and accurately catego rize elements of the environment, most often with the aid of a global positioning system (GPS) (Carp, 2008). Commercial listings of grocery stores,
71 convenience stores, and restaurants are more likely to use telephone directory information, census informati on obtained via questionnaire, and other public record rather than ground truthing methods ( 2013 ). Therefore, it is likely that the locations of these food establishments may not have included the most up to date and accurate information. Also, no differentiation was made between fast food and full service restaurants, which may have a different relationship to dietary intake, weight, and health outcomes. For example, Duffey and colleagues found that fast food was differentially related to poorer metabolic outcomes in young adults compared to sit down restaurants (2009). However, other studies show no difference between how fast food versus full service dining options impact dietary intake or obesity in children (Ayala, Rogers, Arredondo, C ampbell, Baquero, Duerksen, Elder, 2008). There was also no instance, restaurants with a range of healthier options and restaurants that may sell only hamburgers and French fries were all considered as prepared food stores. This could be important, especially given that many restaurant menus are changing to emphasize healthier options, especially for children. In 2011, the National Restaurant Association announced its initiat ive to make healthy dining options for children ( Jalonick, 2011 ). Its Kids LiveWell program includes nearly 40,000 restaurants nationwide, and its goal is to proteins, whole grains, low fat dairy, and limited trans fats, sugars and sodium ( 2011 ). The relationship among proximity to restaurants, convenience stores, and grocery stores may also be very different for rural children and families compared to urban families. In urban areas, children may be more likely to access food from
72 convenience stores on their block. On the other hand, rural children likely only have acc ess to food from prepared food stores and restaurants when their parents choose to is evidenced by the fact that the United States Department of Agriculture co nsiders an urban food desert to be within a one mile radius, while a rural food desert is within a ten mile radius ( www.usda.gov ). These families likely require use of an automobile for the majority of their transportation, and they may travel well beyond their home to obtain food. For example, rural parents may work in more urban or suburban regions, even farther than 10 miles away from their home, and therefore utilize urban or suburban prepared food stores and unprepared food stores. Rural Built Environ ment and Physical Activity Proximity to parks and recreational facilities was expected to be associated with child physical activity, but this hypothesis was not supported. As with the food environment, the literature supports a stronger association betwee n the physical activity environment and BMI versus actual physical activity behavior. The studies that do support the association between parks and recreational facilities and physical activity, did so on a much larger scale, often with self reported physi cal activity or perceived measures of the environment (Gordon Larsen, Nelson, Page, Popkin, 2006; Wilson, Lawman, Segal, Chappell, 2011). GIS derived access to facilities was not associated with physical activity in youth in studies where physical activity was measured through accelerometry (Prins et al. 2011; Scott et al. 2007). It was also expected that population density of the census tract would impact physical activity by providing more destinations within walking distance (Saelens, Sallis, Black, & Chen, 2003) accessibility to social outlets and social support for physical
73 activity (Greiner, Li, Kawachi, Hunt, & Ahluwalia, 2004) This hypothesis was also not supported. In the literature, population density has most often been associated with child physical activity, as measured by accelero metry and through parent report, when it is part of an index, such as sprawl or walkability (Kligerman, Sallis, Ryan, Frank, Nader, 2007; Nelson, Gordon Larsen, Song, Popkin, 2006; Spence, Cutumisu, Edwards, Evans, 2008; Schwartz et al. 2011). Not surprisi ngly, these relationships were detected in communities with higher population density, and therefore, walkability and active transport, which is an unlikely occurrence in an entirely rural sample. Therefore, it is likely that, on a national level, or in a lready densely populated areas, it is more likely that one will detect a statistically significant relationship between population density and physical activity. Although the accelerometry technology used in the current study is supported for its accurac y in the literature (Arvidsson, Slinde, Larsson, Hulthen, 2007 ), there were some challenges with its use with the rural children in this sample. Only 166 out of 269 participants wore the accelerometer correctly and for the appropriate duration, which meant that these analyses were conducted with less power. Also, children were enrolled in this study at several time points across multiple years. Therefore, time of year and weather may have had an impact on whether children engaged in physical activity. For example, children whose baseline week occurred in the summer months in Florida may have engaged in less outdoor physical activity due to intense heat. Children with a summer baseline visit may also have been more likely to engage in physical activity via s wimming, in which case the accelerometer could not have been worn. Anecdotally, many of the children who participated in this study also participated in
74 camps and sports. If their baseline visit coincided with their participation in these activities, it is more likely that their use of parks and recreational facility would be captured by accelerometry. The inconsistency in the literature for the association among physical activity and the built environment may have to do with accuracy in measurement. As technology continues to advance, the precision with which researchers can tackle these associations also advances. For example, a recent study by Rodriguez and colleagues utilized both accelerometry and global positioning systems to measure minute by minute moderate to vigorous physical activity in teenagers, as well as where this physical activity took place (2012). Through the use of this technology, the researchers were able to determine that physical activity occurred in areas of higher population density, near schools, and near parks (2012). Rural Built Environment and Child Weight Children with a higher number of parks and recreational facilities within ten miles of home weighed less than children with a smaller number of parks and recreational facilities in this zone. This finding supports the hypothesis, and is congruent with much of the literature highlighting the importance of physical activity supports and weight status (Alexander, Brunner Huber, Piper, & Tanner, 2013). Of note, the effect size of this relationship is small, but consistent with the literature when objective meas ures of parks and recreational facilities are used (Gilliland et al. 2012; Gordon Larsen, Nelson, Page, Popkin, 2006). However, these relationships were evident in an entirely overweight and obese treatment seeking sample. Therefore, it is possible that ef fect sizes would be larger in a sample with more variability. This finding suggests that children with access to facilities may be more physically active or less sedentary, even
75 though the relationship among the built environment and physical activity has not always been consistent in the literature. It may also be that the children with access to more parks and recreational facilities also have access to more social activities that do not revolve around food, thus impacting dietary intake. The relationshi p between parks and recreational facilities and child weight status did not reach significance at one mile or five miles. This is likely due to the fact that the majority of children in the study did not have parks or recreational facilities available to t hem within one mile, and 30% did not have parks or recreational facilities available within five miles of their home. Indeed, distance to facilities in rural areas may have an entirely different meaning. As with unprepared food stores and prepared food sto res, in order to use these parks and facilities, parents must be willing to drive. Conversely, children living in areas of higher population density tended to weigh more. This finding was contrary to our hypothesis, as the literature shows children living in areas of higher population density tend to weigh less ( Ewing, Brownson, Berrigan, 2006; Oreskovic, Winickoff, Kuhlthau, Romm, Perrin, 2010 ). But, studies supporting population density as an indicator for lower weight status were primarily conducted in urban settings where higher density population likely means having more destinations available via active transport. It is likely that population density has a different relationship with weight status in rural areas, which, by definition, have low popula tion density. Proximity to unprepared food stores and prepared food stores had no relationship with child weight status. This was contrary to the hypothesis, which has previously been supported in the literature. Like the relationship among unprepared fo od stores and prepared food stores and dietary intake, the difference in this study
76 versus the literature may be related to issues of small sample size, environmental scale, and the accuracy of the commercial listings from which addresses were obtained. Si nce much of the literature focusing on proximity to restaurants, convenience stores, and grocery stores has been studied in urban settings, proximity measured was often within walking distance. For example, Gilliland and colleagues found that fast food ou tlets close to homes and schools was associated with higher BMI z scores in adolescents, however these outlets were within 500 meters (.31 miles) (2012). Powell and colleagues found that higher number of unprepared food stores in a census block was related to lower BMI, but this study included 73,079 teens. Also, having at least one grocery store per 10,000 people only decreased BMI by .35%. So, with a large sample size, the effect size remains relatively minor (2007). With studies such as these putting the current study into perspective, it is perhaps not surprising that no relationship between environmental variables and child weight was detected in the current study. Rural Built Environment and Quality of Life pact on child health related quality impact on psychosocial functioning has been documented, and features such as busy streets, crowding, and housing quality have all bee n associated with increased stress and depression (Evans, 2003; Galea, Ahern, Rudenstine, Wallace, Vlahov, 2005). Janicke and colleagues have also shown that poor psychosocial functioning in overweight youth is related to lower quality of life (2007). The refore, it was hypothesized that elements of the built environment that could lead to increased obesity and poorer psychosocial functioning may also lead to poorer quality of life. Of note, children in the current study reported an average total quality of life score that is more similar to
77 children from a chronically ill sample versus a health sample (Varni, Seid, Kurtin, 2001). However, their reported quality of life was not associated with the rural built environment. relationship between the built environment and quality of life was largely exploratory. If a relationship exists, it is likely not direct. For example, as previously hypothesized the built environment may be associated with health behaviors, which could a ffect physical quality of life. Since associations among the built environment, dietary intake, and physical activity were not detected in this study, it stands to reason that the indirect effect on quality of life would also not be present in the current study. Physical quality of life is not the only area that could be indirectly associated with the built environment. Alternatively, the built environment could have been associated with quality of life by providing gathering places (e.g. restaurants and pa rks) to increase social and emotional quality of life (Oldenberg, 1989). It may also be that the current study did not measure other elements of the built environment that could have impacted quality of life, such as available mental health and primary ca re settings (Bonnar & McCarthy, 2012). Using an Obesogenic Rurality Index The individual contributions of each element of the rural built environment studied (i.e. unprepared food stores, prepared food stores, parks and recreational facilities, and popula tion density) were either not significant or demonstrated a very small effect size. Therefore, one of the aims of the study was to capture the join effect of these rural built environment variables on child health behavior and weight status outcomes throug h the calculation of an obesogenic rurality index. As anticipated, this index captured the wide range of rural variability within our sample. It was also correlated with the Montana State
78 University rurality index. However, the obesogenic rurality index sh owed no association with any of the child health behaviors or weight status outcome variables. Since the rurality index combined both built environment elements thought to be associated with dietary intake and physical activity, it is perhaps not surprisin g that the joint effect of these elements was not associated with child behavior. The lack of a relationship among the index and child weight status was not expected, however. There could be several reasons for this. First, the association among weight sta tus and population density was in the opposite direction than what was hypothesized, so this relationship may have dampened the join effect of rurality on weight status overall. Second, as previously noted, the relationship among commercial listings of pre pared food venues and unprepared food stores may have only been a fair representation of what exists in the community and could therefore weaken the index. Third, there may have been variability in level of rurality, the sample may have been too homogenous (i.e. all overweight and obese children from treatment seeking families) to detect an impact of overall rurality. Limitations The current study had several strengths, including utilizing objective measurements of the environment, physical activity, and we ight status. It also utilized buffer zone analysis to capture equal of the previous literature focused on census tracts, which are incongruent in shape and size, as a spatial scale. Most importantly, this st udy applied these measurement methods to a rural sample, which is underrepresented in the literature. However, the majority of the hypotheses in this study were unsupported or yielded very small effect
79 sizes. The explanations for these findings likely ran ge from theoretical to methodological; these potential explanations are outlined below. Due to the paucity of rural environment research, the hypotheses for the current study were largely drawn from previous research in urban and suburban samples, as well as rural research using perceived measure of the environment rather than objective measures. The relationship that urban and suburban children have with their environment may not have generalized to the rural setting. For example, available transportatio n may be quite different, and urban and suburban children may be afforded more independence in walking or using public transportation to access convenience stores, restaurants, or parks. Therefore, the availability of resources within one mile may be more relevant to urban and suburban children versus rural children. As noted above, the USDA considers a ten mile radius to be more salient for rural families. Rural children are likely much more reliant on their parents to provide transportation via automobile to resources. And, if an automobile is a necessity in rural settings, access to resources within a large radius (i.e. ten miles or beyond) may be more important in terms of their impact on health. It also does not seem unreasonable that rural parents may commute to suburban and urban centers for work, and then use the nearby unprepared food stores and prepared food stores to feed their families. There were several methodological limitations that impact the interpretation of size likely made detecting smaller effect sizes a challenge. This study was also cross sectional, and directionality of associations cannot be determined. As previously mentioned, it is importa nt to consider that the study sample was comprised of
80 overweight and obese children from treatment seeking families, which likely affected the variability within our sample, and thus our ability to detect differences based on the built environment. Theref ore these results cannot be generalized to the population. There were also several limitations in behavioral outcome measures used. The Block Kids 2004 questionnaire indicated underreporting of daily caloric intake. Also, both the Block Kids questionnaire and the accelerometers only captured one isolated week of behavior, and associations of behavior and environment may not have been captured if effect sizes are already small. Although every attempt was made to use the most accurate environmental data, it is likely that the commercial list of grocery stores, convenience stores, and restaurants was not 100% accurate. Future research should consider using ground truthing methods in order to accurately capture the rural built environment. Though there was var iability in the rural environment, this study also did not have a comparison non rural sample to determine differences between urban and rural settings. Finally, other neighborhood characteristics, such as socioeconomic position, were not measured, so the impact of these variables is unknown. Clinical Implications and Future Directions The pediatric obesity public health crisis is complex, and means through which it is addressed must be multi faceted. As proposed by ecological models of health, the role of the built environment is only one piece of the puzzle, and research is still determining how big that puzzle piece is. The current study, like those before it, contribution s to the obesity epidemic. Furthermore, these contributions likely interact with a multitude of other variables (e.g. culture, cost, knowledge, personal preference) to impact behavior. In many ways, the small effect sizes noted in much of the research
81 coul d be viewed positively. Changing the built environment drastically is costly and time intensive, even if long term health benefits and cost outweigh the initial expense (Dallat, Soerjomataram, Hunger, Tully, Cairns, Kee, 2013; Wang et al. 2004). Instead, it may be that small environment changes, in combination with helping individuals and families make healthy lifestyle decisions will provide the best results. The current study focused on objective measures of the built environment rather than perceived me asures or use of the built environment. This is important because policy makers and urban planners are looking to research to determine whether presence of built environment supports impact behavior in both children and adults. However, looking at presence of the elements in the built environment is likely not enough. The next step in the research should include observed measures, perceived measures, and built environment use (i.e. restaurant patronage, gym membership, park use) in order to better understan d targets for intervention. Possible areas to address targeting diet and exercise. In a study of older adults, Reed and colleagues found that awareness and use of walking t rails in the local community were low, despite the presence of these trails (2004). Barriers to use, such as cost, parent work schedules, operating hours, or condition of facilities could also be addressed. Finally, it may be that availability and proximit y are, in fact, the largest contributors to use, in which case interventions involving changes to the built environment may be needed. Changing the built environment to support health behaviors in children may be ideal, but it certainly is not always feas ible. Before making significant changes, it is important to know: if we build it will they come? In a quasi experimental study, Fitzhugh
82 and colleagues examined whether the addition of an urban/greenway trail increased physical activity levels of neighborh oods from 2005 to 2007. They found that physical activity counts (i.e. presence of a community resident being active) increased in the experimental community following addition of the greenway, while these activity counts stayed the same for the control co mmunities (2010). Little is known about how environmental interventions impact health behaviors in children. Family based behavioral interventions have shown promise in addressing obesity in children (Wilfley et al. 2007), including those in underserved r ural settings (Janicke et al. 2008). However, the long term treatment effects of these interventions need improvement, and understanding how the built environment affects whether children participating in interventions are successful will better inform pro gram development. There is some preliminary evidence that outcomes of family based pediatric obesity interventions have been moderated by the built environment (Epstein et al. 2012). Epstein and colleagues found that more parks, fewer convenience stores, a nd fewer reductions in child BMI z on intervention outcomes for rural populations is unknown and should be addres sed to best serve these families.
83 APPENDIX MEASURES Background Information 1. (please circle) _____ Boy ______ Girl 2. _______ 3. Date of Birth: ____/____/_____ 4. _______ 5 Your Gender (please circle) ______ Male _____ Female 6 Your 7 Y Mother _____ Father _____ Step Mother _____ Step Father _____ Grandparent _____ Other Legal Guardian _____ 8 Including yourself, how many adults live in your home: ________________ 9 Including your child, how many children live in your home: ________________ 10 Which of the following best describes your education through high school? _____ I left school before earning a high school diploma or GED If this is your answer, how many years did you complete before leaving? ______Years _____ I earned a GED If this is your answer, how many years did you complete before leaving and earning your GED? ______Years _____ I completed high school and earned a high school diploma 11 Did you complete any schooling or trai ning beyond high school/GED? _____Yes _____ No If Yes: 11 a. Which of the following best describes the highest level of school ing or training you completed beyond high school/GED? ____ Vocational or Trade School diploma or certificate
84 ____ Some college classes but no college degree year college degree year college degree ____ Some college or professional school AFTER college graduation ____ Doctoral/Professional degree (such as Ph.D., M.D., J.D.) 12. What is your current job status? (Mark the one that best describes you. However, if more than on e describes you, mark both). _____ Not working _____ Employed full time _____ Retired _____ Employed part time _____ Homemaker, raising children, care for other _____ Disabled, unable to work _____ Other (Specify): _____________________ 13. Which of the statements below best describes your job? If you are not working now, which statement best describes your past job, that is, the job you held the longest? (If you are a homemaker, but works part time, you should mark both). _____ Homemak er, raising children, care of others _____ Managerial, professional speciality _____ Technical, sales, and administrative support _____ Service _____ Operators, fabricators, and laborers, _____ Other (Specify): _______________________________ 1 4 Which of the following best describes your ed ucation through high school? _____ My partner left school before earning a high school diploma or GED If this is your answer, how many years did your partner complete before leaving? ______ Years _____ My partner earned a GED If this is your answer, how many years did your partner complete before leaving and earning your GED? ______Years
85 _____ My partner completed high school and earned a high school diploma 1 5 Did your partne r complete any schooling or tra ining beyond high school/GED? _Yes __ No If Yes: 1 5 a. Which of the following best describes the highest level of schooling or training you r partner completed beyond high school/GED? ____ Vocational or Trad e School diploma or certificate ____ Some college classes but no college degree year college degree year college degree ____ Some college or professional school AFTER college graduation ____ Mast ____ Doctoral/Professional degree (such as Ph.D., M.D., J.D.) 1 6 What is ? (Mark the one that best describes your partner. However, if more than one describes your partner, mark both). _____ Not working _____ Employed full time _____ Retired _____ Employed part time _____ Homemaker, raising children, care for other _____ Disabled, unable to work _____ Other (Specify): _____________________ 17. Which of the statements below best the longest? (If your partner is a homemaker, but works part time, you should mark both). _____ Homemaker, raising children, care of others _____ Managerial, professional speciality _____ Technical, sales, and administrative support _____ Service _____ Operators, fabricators, and laborers, _____ Other (Specify): _______________________________
86 18. Wha t is your total family income falls into? _____ Below $19,999 _____ $60,000 $79,999 _____ $20,000 $39,999 _____ $80,000 $99,999 _____ $40,000 $59,999 _____ Above $100,000 19. Do you consider your child to be Hispanic or Latino? ____ Yes Hispanic or Latino (A person of Mexican, Puerto Rican, Cuban, South or Central American, or other Spanish culture or origin, regardless of race. The _____ No Not Hispanic or Latino _____ No Response 20. What race do you consider your child to be? You may choose more than one response. _____ American Indian or Alaskan Native (A person having origins in any of the original peoples of North, Central, or Sou th America and who ma intains tribal affiliations or community attachment.) _____ Asian (A person having origins in any of the original peop les of the Far East, Southeast Asia, or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam.) _____ Black or African American (A person having origins in any of the black racial groups of _____ Native Hawaiian or Other Pacific Islander ( A person having origins in any of the original peoples of Hawaii, Guam, Somoa, or other Pacific Islands.) _____ White (A person having origins in any of the original peoples o f Europe, the Middle East, or North Africa.) _____ Bi Racial _____ No Response 21. Do you consider yourself to be Hispanic or Latino? ____ Yes Hispanic or Latino (A person of Mexican, Puerto Rican, Cuban, South or Central American, or other Spani sh culture or origin, regardless of race. The _____ No Not Hispanic or Latino
87 _____ No Response 22. What race do you consider yourself to be? You may choose more than one respon se. 23. _____ American Indian or Alaskan Native (A person having origins in any of the origina l peoples of North, Central, or South America, and who ma intains tribal affiliations or c ommunity attachment.) _____ Asian (A person having origins in any of t he original peop les of the Far East, Southeast Asia, or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam.) _____ Black or African American (A person having origins in any of the black racial groups of _____ Native Hawaiian or Other Pacific Islander (A person having origins in any of the original p eoples of Hawaii, Guam, Somoa, or other Pacific Islands.) _____ White (A person having origins in any of the original peopl es of Europe, the Middle East, or North Africa.) _____ Bi Racial _____ No Response
99 LIST OF REFERENCES Adams, M.A., Caparosa, S., Thompson, S., & Norman, G.J. (2009). Translating physical activity recommendations for overweight adolescents to steps per day. American Journal of Preventive Medicine 37 137 140. do i: 10.1016/j.amepre.2009.03.016. Alexander, D.S., Brunner Huber, L.R., Piper, C.R., & Ta nner, A.E. (2013). The association between recreational parks, facilities and childhood obesity: a cross Journal of Epidemiology and Community Health, 67, 427 431. doi: 10.113/jech 2012 2013 01. An, R. & Sturm, R. (2012). School and residential neighborhood food environment and diet among California youth. American Journal of Preventive Medicine, 42 129 135. doi:10.1016/j.amepre.2011.10.012. And ers on, P. M., & Butcher, K. F (2006). Childhood Obesity: Trends and Potential Causes. The Future of Children 16 19 45. doi:10.1353/foc.2006.0001 Arvidsson, D., Slinde, F., Larsson, S., & Hulthen, L. (2007). Energy cost of physical activities in childre n: validation of SenseWear Armband. Medicine and Science in Sports and Exercise, doi:10.101249.mss.0b013e31814fb439. Ayala, G.X., Rogers, M., Arredondo, E.M., Campbell, N.R., Baquero, B., Duerksen, S.,C., & Elder, J.P. (2008). Away from home food intake and risk for obesity : examining the influence of context. Obesity, 16, 1002 1008. doi:10.1038/oby.2008.34. Banis, H.T., Varni, J.W., Wallander, J.L., Korsch, B.M., Jay, S.M., Adler, R., Garciatemple, E. & Ne grete, V. (1988). Psychological and social adjustment of obese children and their families. Child: Care, Health and Development 14 157 173. Bell, J.F., Wilson, J.S., & Liu, G.C. (2008). Neighborhood greenness and 2 year changes in body mass index of chi ldren and youth. American Jour nal of Preventive Medicine, 35 547 553. doi:10.1016/j.amepre.2008.07.006. Bjrk, J., Albin, M., Grahn, P., Jacobsson, H., Ard, J., Wadbro, J., stergren, P. O., & Skarback, E.(2008). Recreational values of the natural environ ment in relation to neighbourhood satisfaction, physical activity, obesity and wellbeing. Journal of Epidemiology & Community Health 62 Bodor, J. N., Rice, J. C., Farley, T. A., Swalm, C. M., & Rose, D. (2010). The association between obesity and urban f ood environments. Journal of Urban Health 87 771 781.
100 Boehmer, T K., Hoehner, C. M., Deshpande, A. D., Brennan Ramirez, L. K., & Brownson, R C. (2007). Perceived and observed neighborhood indicators of obesity among urban adults. International Journal o f Obesity, 31 968 977. Boehmer, T.K. & Lovegreen, S., Haire Joshu, D., & Brownson, R.C. (2006). What constitutes an obesogenic environment in rural communities? American Journal of Health Promotion, 20(6), 411 421. Bonnar, K.K., & McCarthy, M. (2012). Heal th related quality of life in a rural area with low racial/ethnic density. Journal of Community Health, 37, 96 104, doi: 10.1007/s10900 011 9422 2. Bowman, S.A, Gortmaker, S.L., Ebbeling, C.B Pereira, M.A., & Ludwig, D.S. (2004). Effects of fast food cons umption on energy intake and diet quality among children in a national household survey. Pediatrics, 113 112 118. Briefel, R. R., & Johnson, C. L. (2004). Secular Trends In Dietary Intake in the United States. Annual Review of Nutrition 24 (1), 401 431. d oi:10.1146/annurev.nutr.23.011702.073349 Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design Harvard University Press. Brown, D.L. & Cromartie, J.B. (2003). The nature of rurality in postindustrial society. In New forms of urbanization: Conceptualization and measuring human settlement in the 21 st century. Bellagio, Italy, March 11 15, 2002. Brownson, Ross C, Boehmer, Tegan K, & Luke, D. A. (2005). Declining rates of physical activity in the United States: What are the contributor? Annual Review of Public Health 26 (1), 421 443. doi:10.1146/annurev.publhealth.26.021304.144437 Burdette, H.L. & Whitaker, R.C. (2004). Neighborhood playgrounds, fast food restaurants, and crime: relationships to overweight in low income p reschool children. Preventive Medicine, 38, 57 63.doi: 10.1016/j.ypmed.2003.09.029. Calle, E., Rodriguez, C., & Walker Thurmond, K. (2003). Overweight, obesity, and mortality from cancer in a prospectively studied cohort of US adults. New England Journal o f Medicine, 348, 1625 1638. Cameron, A.C. (2004). Kurtosis. In M.S. Lewis Beck, A. Bryman & T.F. Liao (Eds.), The SAGE Encyclopedia of Social Science Research Methods doi:10.4135/9781412950589. Carp, J. (2008). "Ground truthing" representations of social space: Using Lefebvre's conceptual triad. Journal of Planning Education and Research, 28, 129 142. doi: 10.1177/0739456X08324685.
101 Casey, A.A., Elliot, M. Gianz, K., Haire Joshu, D., Lovegreen, S.L., Saelens, B.E., Sallis, J.F., & Brownson, R.C. (2008). Imp act of the food environment and physical activity environment on behaviors and weight status in rural U.S. communities, Preventive Medicine, 47, 600 604. doi: 10.1016/j.ypmed.2008.10.001. Crawford, D., Cleland, V., Timperio, A., Salmon, J., Andrianopoulos, N., Roberts, R., Giles Corti, B., Baur, L., & Ball, K. (2010). The longitudinal influence of home and neighbourhood environments on children's body mass index and physical activity over five years: the CLAN study. International Journal of Obesity, 34, 117 7 1187. doi:10.038/ijo.2010.57. Cudney, S., Craig, C., Nichols, E., & Weinert, C. (2004). Barriers to recruiting and adequate sample in rural nursing research. The Online Journal of Rural Nursing and Health Care, 4 78 88. Cullen, K. W., Watson, K., & Zake ri, I. (2008). Relative Reliability and Validity of the Block Kids Questionnaire among Youth Aged 10 to 17 Years. Journal of the American Dietetic Association 108 (5), 862 866. doi:10.1016/j.jada.2008.02.015 Dallat, M.A., Soerjomataram, I., Hunter, R.F., T ully, M.A., Cairns, K.J., & Kee, F. (2013). Urban greenways have the potential to increase physical activity levels cost effectively. European Journal of Public Health, doi: 10.1093/eurpub/ckt035. Dietz, W. (1998). Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics 101, 518 525. Duffey, K.J., Gordon Larsen, P., Steffen, L.M., Jacobs, D.R., & Popkin, B.M. (2009). Regular consumption from fast food establishmen ts relative to other restaurants is differentially assocated with metabolic outcomes in young adults. The Journal of Nutrition, 139, 2113 2118. doi:10.3945/jn.109.109520. Durand, C. P., Andalib, M., Dunton, G. F., Wolch, J., & Pentz, M. A. (2011). A system atic review of built environment factors related to physical activity and obesity risk: implications for smart growth urban planning. Obesity Reviews Smart growth urban planning and obesity risk, 12 e173 e182. doi:10.1111/j.1467 789X.2010.00826.x Egger, G., & Swinburn, B. (1997). An "ecological approach to the obesity pandemic. British Medical Journal, 315, 477 480. Ellaway, A., Macintyre, S., & Bonnefoy, X. (2005). Graffiti, greenery, and obesity in adults: secondary analysis of European cross sectional survey. British Medical Journal, 331 611 612.doi:10.1136/bmj.38575.664549.f7.
102 Epstein, L.H., Raja, S., Daniel, T.O., Paluch, R.A., Wilfley, D.E., Saelens, B.E., & Roemmich, J.N., (2012). The built environment moderates effects of family based childhood o besity treatment over 2 years. Annals of Behavioral Medicine, 44, 248 258. doi: 10.1007/s12160 012 9383 4. Evans, G. (2003). The built environment and mental he alth. Journal of Urban Health : Bulletin of the New York Academy of Medicine, 80(4), 536 555. Evenson, K.R., Scott, M.M., Cohen, D.A., & Voorhees, C.C. (2007). Girls' perception of neighborhood factors on physical activity, sedentary behavior, and BMI. Obesity,15 430 445. Ewing, R., Schmid, T., Killingsworth, R., Zlot, A., & Raudenbush, S. (2003) Relationship between urban sprawl, physical activity, obesity, and morbidity. American J ournal of Health Promotion, 18 47 57. Ewing, R., Brownson, R.C. & Berrigan, D. (2006). Relationship between urban sprawl and weight of United States Youth. American Journa l of Preventive Medicine, 31 464 474. doi: 10.1016/j.amepre.2006.08.020. Ewing, R., Handy, S., Brownson, R.C., Clemente, O., & Winston, E. (2006). Identifying and measuring urban design qualities related to walkability. Journal of Phy sical Activity and Health, 3 S223 S239. Feng, J., Glass, T.A., Curriero, F.C., Stewart, W.F., & Schwartz, B.S. (2010). The built environment and obesity: a systematic review of the epidemiologic evidence. Health & Place, 16, 175 190. doi: 10.1016/j.healthplac e.2009.09.008. Fitzhugh, E.C., Bassett, D.R., & Evans, M.F. (2010). Urban trails and physical activity: a natural experiment. American Journal of Preventive Medicine, 39, 259 262. doi:10.1016/j.amere.2010.05.010. Flegal, K., Carroll, M.,Ogden, C., & Johnso n C. (2002). Prevalence and trends in obesity among US adults, 1999 2000. The Journal of the American Medical Association 288, 1723 1727. Fleischhacker, S. E., Evenson, K. R., Rodriguez, D. A., & Ammerman, A. S. (2010). A systematic review of fast food ac cess studies. Obesity Reviews 12 e460 e471. doi:10.1111/j.1467 789X.2010.00715.x Franco, M., Diez Roux, A. V., Glass, T. A., Caballero, B., & Brancati, F. L. (2008). Neighborhood characteristics and availability of healthy foods in Baltimore. American Jo urnal of Preventive Medicine 35 561 567. doi:10.1016/j.amepre.2008.07.003
103 Frank, L.D., Sallis, J.F., Saelens, B.E., Leary, L., Cain, K., Conway, T.L., & Hess, P.M. (2010). The development of a walkability index: application to the Neighborhood Quality of Life Study. British Journal of Sports Medicine, 44, 924 933. doi: 10.1136/bjsm.2009.058701. Galea, S., Ahern, J., Rudenstine, S., Wallace, Z., & Vlahov, D. (2005). Urban built environment and depression: a multilevel analysis. Journal of Epidemiology & Co mmunity Health 59 822. Galvez, M.P, Hong, L., Choi, E., Liao, L., Godbold, J. & Brenner, B. (2009). Childhood obesity and neighborhood food store availability in an inner city community. Academic Pediatrics 9 339 343. Giles Corti, B., & Donovan, R. J. (2003). Relative influences of individual, social environmental, and physical environmental correlates of walking. American Journal of Public Health 93 1583. Gilliland, J.A., Rangel, C.Y., Healy, M.A., Tucker, P., Loebach, J.E., Hess, P.M., He, M., Irwin, J.D., & Wilk, P. (2012). Linking childhood obesity to the built environment; a multi level analysis of home and school neighbourhood factors associated with body mass index. Canadian Journal of Public Health, 103, S15 S21. Glanz, K., & Sallis, J. F (2006). The role of built environments in physical activity, eating, and obesity in childhood. The Future of Children 16 89 108. Brookings Institution Press. Glanz, K., Sallis, J., Saelens, B., & Frank, L.D. (2005). Healthy nutrition environments: conc epts and measures. American Journal of Health Promotion, 19, 330 333. Gordon Larsen, P., Nelson, M.C., Page, P., & Popkin, B.M. (2006). Inequality in the Built Environment Underlies Key Health Disparities in Physical Activity and Obesity. Pediatrics 117 417 424. doi:10.1542/peds.2005 0058 Gray, W. N., Kahhan, N. A., & Janicke, D. M. (2009). Peer victimization and pediatric obesity: A review of the literature. (L. A. Theodore, M. A. Bray, & T. J. Kehle, Eds.) Psychology in the Schools 46 720 727. doi:10.1 002/pits.20410 Greiner, K.A., Li, C., Kawachi, I., Hunt, D.C., & Ahluwalia, J. S. (2004). The relationships of social participation and community ratings to health and health behaviors in areas with high and low population density. Social Science & Medicin e 59 2303 2312. Hall, S. A., Kaufman, J. S., & Ricketts, T. C. (2006). Defining Urban and Rural Areas in U.S. Epidemiologic Studies. Journal of Urban Health 83 162 175. doi:10.1007/s11524 005 9016 3 Hart, L.G., Larson, E.H., & Lishner, D.M. (2005). Rur al definitions for health policy and research. American Journal of Public Health, 95 1149 1155.
104 Heinrich, K.M., Lee, R.E., Regan, G.R., Reese Smith, J.Y., Howard, H.H., Haddock, K. Carlos Poston, W. S., & Ahluwalia, J.S. (2008). How does the built environ ment relate to body mass index and obesity prevalence among public housing residents? American Journal of Health Promotion 22 187 194. Hill, J., Wyatt, H., Reed, G. & Peters, J. (2003). Obesity and the environment: where do we go from here? Science 299, 853 855. How we compile and verify our data. (2013). Retrieved March 30, 2013, from www.infousa.com/business lists/ Hunsberger, M., O'Malley, J., Block, T., & Norris, J.C. (2012). Relative validation of Block Kids Food Screener for dietary assessment in children and adolescents. Maternal and Child Nutrition. Epub ahead of print. doi:10.1111/j.1740 8709.2012.00446.x Inagami, S., Cohen, D., Finch, B., & Asch, S.M. (2006). You Are Where You Shop: Grocery Store Locations, Weight, and Neighborhoods. American Jou rnal of Preventive Medicine, 31 10 17. doi:10.1016/j.amepre.2006.03.019. Ingerski, L.M, Modi, A., Hood, K., Pai, A., Zeller, M., Piazza Waggoner, C., Driscoll K. A., Rothenberg, M.E., Franciosi, J., & Hommel, K.A. (2010). Health Related Quality of Life Across Pediatric Chronic Conditions. The Journal of Pediatrics 156 639 644. Mosby, Inc. doi:10.1016/j.jpeds.2009.11.008 Introducing the enhanced bodymedia sensewear system. (2013). Retrieved March 30, 2013 from http://sensewear.bodymedia.com/ Jalonick, M.C. (2011). Chain restaurants will make kids menus healthier. In The Huffington Post. Retrieved Ma rch 30, 2013 from http://www.huffingtonpost.com/2011/07/13/chain restaurants healthy kids meals_n_896866.html Janicke, D. M., Marciel, K. K., Ingerski L. M., Novoa, W., Lowry, K. W., Sallinen, B. J., & Silverstein, J. H. (2007). Impact of psychosocial factors on quality of life in overweight youth. Obesity (Silver Spring, Md) 15 1799 1807. doi:10.1038/oby.2007.214 Jilcott, S. B., Wade, S., McGuirt, J T., Wu, Q., Lazorick, S., & Moore, J. B. (2011). The association between the food environment and weight status among eastern North Carolina youth. Public Health Nutrition 1 8. doi:10.1017/S1368980011000668 Joshu, C.E., Boehmer, T.K., Brownson, R.C., & Ewing, R. (2008). Personal, neighbourhood and urban factors associated with obesity in the United States. Journal of Epidemiology & Community Health 62 202 208. doi: 10.1136/jech.2006.058321
105 Kligerman, M., Sallis, J., Ryan, S., Frank, L. & Nader, P.R. (2007). Association of neighborhood design and recreation environment variables with physical activity and body mass index in adolescents. American Jo urnal of Health Promotion, 21 274 277. Kuczamarski, R.J., Ogden, C.L., Guo, S.S., Grummer St rawn, L.M., Flegal, K.M., Mei, Z. Wei, R., Curtin, L.R., Roche, A.F., & Johnson, C.L. (2000). 2000 CDC growth charts for the United States: Methods and Development. Vital and Health Statistics Series, 11 1 190. Lamichhane, A., Mayer Davis, E.J., Puett, R ., Bottai, M., Porter, D.E., & Liese, A.D. (2012). Associations of built food environment with dietary intake among youth with diabetes. Journal of Nutrition Education and Behavior, 44, 217 224. doi:10.1016/j.jneb.2011.08.003. Larson, N. I., Story, M. T., & Nelson, M. C. (2009). Neighborhood environments: disparities in access to healthy foods in the U.S. American Journal of Preventive Medicine 36 74 81. doi:10.1016/j.amepre.2008.09.025 Laska, M. N., Hearst, M. O., Forsyth, A., Pasch, K. E., & Lytle, L. (2 010). Neighbourhood food environments: are they associated with adolescent dietary intake, food purchases and weight status Public Health Nutrition 13 1757 1763. doi:10.1017/S1368980010001564 Leslie, E., Coffee, N., Frank, L., Owen, N., Bauman, A., & Hug o, G. (2007). Walkability of local communities: Using geographic information systems to objectively assess relevant environmental attributes. Health & Place 13 111 122. doi:10.1016/j.healthplace.2005.11.001 Lewin, K., Heider, F. T., & Heider, G. M. (1936 ). Principles of Topological Psychology. McGraw Hill. Leyden, K. M. (2003). Social capital and the built environment: the importance of walkable neighborhoods. American Journal of Public Health 93 1546. Lichtman SW, Pi sarska K, Raynes Berman E, Pestone, M., Dowling, H., Offenbacher, E., Weisel, H. Heshka, S., Matthews, D.E., & Heymsfield, S.B (1992). Discrepancy between self reported and actual caloric intake and exercise in obese subjects. New England Journal of Medicine 327 :1893 8. doi: 10.1056/NEJM 199212313272701. Lim, C.S., Follansbee Junger, K., Crawford, M.S., & Janicke, D.J. (2011). Treatment outcome research in rural pediatric populations: the challenge of recruitment. Jour nal of Pediatric Psychology,36 696 707. doi: 10.1093/jpepsy/jsr018. L iu, G.C., Cunningham, C., Downs, S.M., Marrero, D.G., & Fineberg, N. (2002). A spatial analysis of obesogenic environments for children. In Proceedings of the American Medical Informatics Association Symposium. 459 463.
106 Liu, G.C, Wilson, J.S., Qi, R. & Ying, J. (2007). Green neighborhoods, food retail and childhood overweight: differences by population density. American Journal Health Promotion 21 317 325. Lutfiyya, M.N., Lipsky, M.S., Wisdom Behounek, J. & Inpanbutr Martinkus, M. (2007). Is rural resi dency a risk factor for overweight and obesity for US children? Obesity 15, 2348 2356. Lytle, L. A. (2009). Measuring the Food Environment: state of the science. American Journal of Preventive Medicine 36 S134 S144. doi:10.1016/j.amepre.2009.01.018 Maddock, J. (2004). The relationship between obesity and the prevalence of fast food restaurants: state level analysis American Journal of Health Promotion 19 137 143. McCrory, M.A., Fuss, P.J., Saltzman, E., & Roberts, S.B. (2000). Dietary determinants of energy intake and weight regulation in healthy adults. The Journal of Nutrition, 130 276s 279s. McDermott, A. & Stephens, M.B. (2010). Cost of eating: whole foods versus convenience foods in a low income model. Family Medicine 42 280 284. Mehta, N. K ., & Chang, V. W. (2008). Weight Status and Restaurant Availability. American Journal of Preventive Medicine 34 127 133. doi:10.1016/j.amepre.2007.09.031 Merchant, A., Dehghan, M., Behnke Cook, D., & Anand, S. (2007). Diet, physical activity, and adiposi ty in children in poor and rich neighborhoods: A cross sectional comparison. Nutrition Journal. doi: 10.1186/1475/2891/6/1 Mobley, L. R., Root, E. D., Finkelstein, E. A., Khavjou, O., Farris, R. P., & Will, J. C. (2006). Environment, obesity, and cardiovasc ular disease risk in low income women. American Journal of Preventive Medicine 30 327 327. Moore, J. B., Jilcott, S. B., Shores, K. A., Evenson, K. R., Brownson, R. C., & Novick, L. F. (2010). A qualitative examination of perceived barriers and facilita tors of physical activity for urban and rural youth. Health Education Research 25 355 367. doi:10.1093/her/cyq004 Morland, K., Roux, A. D., & Wing, S. (2006). Supermarkets, Other Food Stores, and Obesity: The Atherosclerosis Risk in Communities Study. Am erican Journ al of Preventive Medicine, 30 333 339. doi: 10.1016/j.amepre.2005.11.003. Mujahid, M. S., Roux, A. V. D., Shen, M., Gowda, D., Snchez, B., Shea, S., Jacobs, D. R., &, Jackson, S.A. (2008). Relation between neighborhood environments and obesit y in the Multi Ethnic Study of Atherosclerosis. American Journal of Epidemiology 167 1349.
107 Must, A., & Strauss, R. (1999). Risks and consequences of childhood and adolescent obesity. International Journal of Obesity Supplement 23 S2 S11. Nelson, M.C., Gordon Larsen, P., Song, Y., & Popkin, B.M. (2006). Built and social environments: associations with adolescents, overweight, and activity. American Journ al of Preventive Medicine, 31, 109 117. doi: 10.1016/j.amepre.2006.03.026. Nielsen, T. S ., & Hansen, K. B. (2007). Do green areas affect health? Results from a Danish survey on the use of green areas and health indicators. Health & Place 13 839 850. Ogden, C.L., Carroll, M.D., Kit, B.K., & Flegal, K.M. (2012). Prevalence of obesity and trends in body mass index among US children and adolescents, 1999 2010. Journal of the American Medical Association, 307 483 490. Oldenberg, R. (1989). The Great Good Place: Cafes, Coffee Shops, Bookstores, Bars, Hair Salons, and Other Hangouts at the Heart of a Community. New York, NY: Marlowe & Company. Owen, N., Cerin, E., Leslie, E., duToit, L., Coffee, N., Frank, L.D., Bauman, A. E., Hugo, G. Saelens, B.E., & Sallis, J.F. (2007). Neighborhood Walkabili ty and the Walking Behavior of Australian Adults. American Journal of Preventive Medicine 33 387 395. doi:10.1016/j.amepre.2007.07.025 Oreskovic, N.M., Winickoff, J.P., Kuhlthau, K.A., Romm, D., & Perrin, J.M. (2010). Obesity and the built environment am ong Massachusetts children. Clinical Pediatrics, 48 904 912. doi:10.1177/0009922809336073. Ornsby, Napoleon, Burke, Groessle, & Bowden (2010). Getting to Know ArcGIS for ArcGIS 10 Desktop. ESRI Press, Redlands, CA. Papas, M. A., Alberg, A. J., Ewing, R., Helzlsouer, K. J., Gary, T. L., & Klassen, A. C. (2007). The Built Environment and Obesity. Epidemiologic Reviews 29 129 143. doi:10.1093/epirev/mxm009 Peeters, A., Barendregt, J.,Willekens, F., Mackenback, J., Al Mamun, A., & Bonneaux, L.(2003). Obesity in adulthood and its consequences for life expectancy: a life table analysis. Annals of Internal Medicine, 138, 24 32. Peters, J., Wyatt, H., Donahoo, W., & Hill, J. (2002). From instinct to intellect: the challenge of maintaining healthy weight in the mo dern world. Obesity Reviews 3 69 74. Wiley Online Library. Poortinga, W. (2006). Perceptions of the environment, physical activity, and obesity. Social Science & Medicine 63 2835 2846. doi: 10.1016/j.socscimed.2006.07.018.
108 Powell, L.M., Auld, C., Chalo upka, F.J., O'Malley, P.M., & Johnston, L.D. (2007). Associations between access to food stores and adolescent body mass index. American Journ al of Preventive Medicine, 33 S301 S307. doi: 10.1016/j.amepre.2007.07.007. Powell, L.M., Han, E., Zenk, S.N., Kh an, T., Quinn, C.M., Gibbs, K.P., Pugach, O., Barker, D.C., Resnick, E.A., Myllyluoma, J., & Chaloupka, F.J. (2011). Field validation of secondary commercial data sources on the retail food outlet environment in the U.S., Health Place, 17, 1122 1131.doi:10.1016/j.healthplace.2011.05.010. Prins, R.G., Ball, K., Timperio, A., Salmon, J., Oenema, A., Brug, J., & Crawford, D. (2011). Associations between availability of facilities within three different neighbourhood buffer sizes and objectively assessed physical activity in adolescents. Health and Place 17, 1228 1234. doi:10.1016/j.healthplace.2011.07.012. Reed, J.A., Ainsworth, B.E., Wilson, D.K., Mixon, G., & Cook, A. (2004). Awareness and use of community walking trails. Preventive Medicine, 39, 903 908. Ricciardelli, L.A., & McCabe, M.P. (2001). Children's body image concerns and eating disturbance: A review of the literature. Clinical Psychology Review 21 325 344. Richardson, A. S., Boone Heinonen, J., Popkin, B. M., & Gordon Larsen, P. (2011). Neighborhood fast food restaurants and fast food consumption: A national study. British Medical Journal: Public Health 11 543. doi:10.1186/1471 2458 11 543 Romero, A.J., Robinson, T.N., Kraemer, H.C., Erickson, S.J., Haydel, K.F., Mendoza, F., & Killen, J.D. (2001). Are perceived neighborhood hazards a barrier to physical activity in children? Archives of Pediatric and Adolescent Medicine, 155, 1143 1148. Saelens, B.E., Sallis, J.F., Black, J.B., & Chen, D. (2003). Neighborhood based differences i n physical activity: an environment scale evaluation. American Journal of Public Health, 93 1552 1558. Sallis, J., Owen, N., & Fisher, E. (2008). Ecological models of health behavior. In: Health Behavior and Health Education: Theory, Research and Practic e. 4 th Edition. Eds: Glanz K., Rimer B.K., Viswanath K., San Francisco: Jossey Bass. Schwartz, B.S., Stewart, W.F., Godby, S., Pollack, J., DeWalle, J., Larson, S., Mercer, D.G., & Glass, T.A. (2011). Body mass index and the built and social environments in children and adolescents using electronic health records. American Journal of preventive Medicine, 41, e17 e28. doi: 10.1016/j.amepre.2011.06.038
109 Scott, M.M., Cohen, D.A., Evenson, K.R., Elder, J., Catellier, D., Ashwood, J.S., & Overton, A. (2007). Weekend schoolyard accessibility, physical activity, obesity: the Trial of Activity in Adolescent Girls (TAAG) study. Preventive Medicine, 44, 398 403. doi: 10.1016/j.ypmed.2006.12.010. Simen Kapeu, A., Khule, S., & Veugelers, P.J. (201 0). Geographic differences in childhood overweight, physical activity, nutrition and neighbourhood facilities: implications for prevention. Canadian Journal of Public Health, 101 128 132. Spence, J.C., Cutumisu, N., Edwards, J., & Evans, J. (2008). Influe nce of neighbourhood design and access to facilities on overweight among preschool children. International Journal of Pediatric Obesity,3 109 116. Spieth, L., & Harris, C.V. (1996). Assessment of health related quality of life in children and adolescents : an integrative review. Journal of Pediatric Psychology 21 175 193. Stokols, D. (1996). Translating social ecological theory into guidelines for community health promotion. American Journal of Health Promotion, 10, 282 298. Strauss, R. & Pollack, H.A. ( 2003). Social marginalization of overweight children. Archives of Pediatrics and Adolescent Medicine, 157, 746 752. Sturm, R. & Datar, A. (2005). Body mass index in elementary school children, metropolitan area food prices, and food outlet density. Public Health, 119 1059 1068. doi: 10.1016/j.puhe.2005.05.007. Sugiyama, T., Salmon, J., Dunstan, D. W., Bauman, A. E., & Owen, N. (2007). Neighborhood walkability and TV viewing time among Australian adults. American Journal of Preventive Medicine 33 444 449. Swinburn, B., & Egger, G. (1999). Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Preventive Medicine 29, 563 570. Tai Seale, T. & Chandler C. ( 2010). Nutrition and overweight concerns in rural areas: a literature review. Rural Healthy People 2010: A companion document to Healthy People 2010. Volume 2. College Station, TX: The Texas A&M University System Health Sciences Center, School of Rural Pub lic Health, Southwest Rural Health Research Center. Timperio, A., Salmon, J., Telford, A., & Crawford, D. (2005). Perceptions of local neighbourhood environments and their relationship to childhood overweight and obesity. International Journal of Obesity, 29 170 175. doi: 10.1038/sj.ijo.0802865.
110 Tsiros, M. D., Olds, T., Buckley, J. D., Grimshaw, P., Brennan, L., Walkley, J., Hills, A. P., Howe, P.R.C., & Coats, A.M. (2009). Health related quality of life in obese children and adolescents. International jou rnal of obesity (2005) 33 387 400. doi:10.1038/ijo.2009.42 Vance, V.A., Woodruff, S.J., McCargar, L.J., Husted, J., & Hanning, R.M. (2008). Self reported dietary energy intake of normal weight, overweight and obese adolescents. Public Health Nutrition, 1 2, 222 227. doi:10.1017/S1368980008003108. Varni, J. W., Seid, M., & Kurtin, P. (2001). Reliability and Validity of the Pediatric Quality of Life InventoryTM Version 4.0 Generic Core Scales in Healthy and Patient Populations. Medical Care 800 812. Vias, A (2004). Bigger stores, more stores, or no stores: paths of retail restructuring in rural America. Journal of Rural Studies 20 303 318. doi:10.1016/j.jrurstud.2003.10.003 Wang, G., Macera, C.A., Scudder Scouci, B., Schmid, T., Pratt, M., Buchner, D., He ath, G. (2004). Cost analysis of the built environment: the case of bik and pedestrial trials in Lincoln, Neb. American Journal of Public Health, 94, 549 553. Ward, D.S., Dowda, M., Trost, S.G., Felton, G.M., Dishman, R.K., & Pate, R.R. (2006). Physical ac tivity correlates in adolescent girls who differ by weight status. Obesity, 14 97 105. Weinert, C., & Boik, R. J. (1995). MSU rurality index: development and evaluation. Research in Nursing & Health 18 453 464. Wiley Online Library. Westerterp KR, Verb oeket van de Venne WPHG, Meijer GAL, & Hoor ten F. (1991). Self reported intake as a measure for energy intake. A validation against doubl y labelled water. Obese Europe, 91 :17 22. Wieczorek, W. F., & Delmerico, A. M. (2009). Geographic information systems. Wiley Interdisciplinary Reviews: Computational Statistics 1 167 186. doi:10.1002/wics.21 Wilfley, D. E., Tibbs, T. L., Van Buren, D., Reach, K. P., Walker, M. S., & Epstein, L. H. (2007). Lifestyl e interventions in the treatment of childhood overweight: A meta analytic review of randomized controlled trials Health Psychology 26 521 532. doi:10.1037/0278 6220.127.116.111 Williamson, D. A., Champagne, C. M., Han, H., Harsha, D., Martin, C. K., Newton, R. L., Ryan, D. H., Southern, M.S., Stewart, T.M.,& Webber, L.S., (2009). Increased obesity in children living in rural communities of Louisiana. International Journal of Pediatric Obesity 4 160 165. doi:10.1080/17477160802596148
111 Wilson, D.K., Lawman, H .G., Segal, M., & Chappell, S. (2011). Neighborhood and parental supports for physical activity in minority adolescents. American Journal of Preventive Medicine, 41, 399 406. doi:10.1016/j.amepre.2011.06.037. Wolf, A. M., & Colditz, G. A. (1998) Current est imates of the economic cost of obesity in the United States. Obesity Researc h, 6 97 106. Wolkoff, L.E., Tanofsky Kraff, M., Shomaker, L.B., Kozlosky, M., Columbo, K.M., Elliot, C.A., Ranzenhofer, L.M., Osborn, R.L., Yanovski, S.Z., & Yanovski, J.A. (2011 ). Self reported versus actual energy intake in youth with and without loss of control eating. Eating Behaviors, 12, 15 20. doi:10.1016/j.eatbeh.2010.09.001. Yousefian, A., Ziller, E., Swartz, J., & Hartley, D. (2009). Active living for rural youth: addre ssing physical inactivity in rural communities. Journal of Public Health Management and Practice, 15 223 231. Zeller, M.H, & Modi, A.C. (2006). Predictors of Health Related Quality of Life in Obese Youth. Obesit y 14 122 130. Zeller, M.H., Saelens, B.E., Roehrig, H., Kirk, S., & Daniels, S.R. (2004). Psychological adjustment of obese youth presenting for weight management treatment. Obesity Research, 12, 1576 1586.
112 BIOGRAPHICAL SKETCH Megan Cohen is a doctoral candi date in the Department of Clinica l and Health Psychology at the University of Florida. She obtained a Bachelor of Arts in psychology at Miami University in 2006 and completed an honors thesis on spatial cognition. Megan also held a position as a research a ssistant in the Department of Behavioral Medicine at Cincinnati Children's Hospital Medical Center, specifically working on interventions for children with chronic headaches, cystic fibrosis, and type 1 diabet es. She graduated with a Master of Science from the University of Florida in 2010, and her thesis was doctoral internship in clinical psychology at Nemours/Alfred I. duPont Hos pital for Children in Wilmington, Delaware. Megan will be staying on in a post doctoral fellowship position where she will be providing clinical care for chronically ill children and their families.