- Open Access
Crime rates and sedentary behavior among 4thgrade Texas school children
International Journal of Behavioral Nutrition and Physical Activityvolume 5, Article number: 28 (2008)
Although per capita crime has generally fallen over the period which coincides with the obesity epidemic, it has not fallen uniformly across communities. It also has not fallen enough to allay fears on the part of parents. Over the past 30 years, technological changes have made the indoor alternatives to playing outside, where children are more vulnerable to criminal activity, more enjoyable (cable TV, video games, and the internet) and comfortable (the spread of air conditioning to low income neighborhoods). We determined whether indoor sedentary behavior patterns are associated with community crime statistics. 4thgraders in the U.S. are typically 9 or 10 years old.
We used data from the 2004–2005 Texas School Physical Activity and Nutrition (SPAN) survey linked with U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics data for the years 2000 through 2005 and Texas State data on sexual offenders. The probability-based sample included a total of 7,907 children in grade four. Multistage probability sampling weights were used. The dependent variables included were hours of TV watching, video game playing, computer use and total indoor sedentary behavior after school. Incremental Relative Rates were computed for community crime rates including robberies, all violent crimes, murders, assaults, property crimes, rapes, burglaries, larcenies and motor vehicle thefts as well as for sexual offenders living in the neighborhood. The neighborhood refers to the areas where the students at each school live. In the case of sexual offenders, sexual offenders per capita are estimated using the per capita rate in the zip code of the school attended; all other crime statistics are estimated by the crimes per capita in the police department jurisdiction covering the school attended. After controlling for sex, age, and African-American and Hispanic, cross-sectional associations were determined using multivariate Poisson regression.
4thgrade boys were more likely to play video games in communities with increased per 100 population rates of larceny and burglary as well as in communities with increased per capita sexual offenders; 4thgrade girls were more likely to watch television in communities with increased per capita sexual offenders. While 4thgrade girls were more likely to watch TV in communities with increased per capita sex offenders, they were less likely to use computers. Per capita sexual offenders were negatively related to computer use amongst 4thgrade girls.
By combining community crime and cross-sectional individual level data on indoor sedentary behavior, we found that there is an association between community crimes/sex offender rates and certain types of indoor sedentary behavior. The development of technologies in recent decades which makes supervising children easier indoors, where children are much less vulnerable to crime, may be contributing to the epidemic of childhood obesity.
The prevalence of obesity  among children has doubled in the last twenty years , disproportionately affecting minorities [3–6]. Part of the problem is the lack of physical activity among youth, which is associated with higher rates of obesity [7–13] Time spent at indoor play comes at the expense of time spent at outdoor play, which may be more physical. Resultingly, sedentary behavior has been shown to be associated with higher rates of child and adolescent obesity [14–18]. During the childhood obesity epidemic, new technologies have made supervising children indoors, where children are less vulnerable to crime in comparison to playing outdoors, both easier and more comfortable. Twenty-five years ago, there were few households with cable television, video games, and no households with internet connections. Many households, especially in low income areas, did not have air conditioning. Technologies which improve the alternatives to playing outside have greatly improved and spread into low income neighborhoods.
Over the period corresponding to the childhood obesity epidemic, crime rates have fallen in the U.S . However, crime has not fallen in all communities and neighborhoods uniformly and many still have very high rates. At the same time, the ease of supervising children indoors, where children are less vulnerable to crime, has improved markedly. Thirty years ago, the options for television viewing after school were limited to local broadcast channels. From 1970 to 2007, the percentage of households with cable grew from 6.7 percent to 64.1 percent . Video game development was in its infancy during the 1970s and personal computing was limited to only the most technologically savvy kids whose parents were wealthy enough to afford the technology. The internet was used by only 15 percent of the population in 1995, but that grew to 73 percent in 2006 . Further, many homes were without air conditioning 30 years ago . Playing indoors has become more and more attractive every year for children and parents/guardians alike.
Physical activity, which is the alternative to sedentary behavior, has been shown to be negatively associated with neighborhood crime and road safety . Weir, Etelson, and Brand have shown that parental perceptions of crime in inner cities is negatively associated with lower physical activity among youth . Road safety and perceptions about the trustworthiness of neighbors have been shown to be positively associated with physical activity among youth [25, 26] Access to safe play environments, such as at schools, has been shown to be positively associated with increased physical activity . It's important to note, all of these studies focus on parental perceptions, rather than objective statistics about crime rates. In the one exception, violent crime densities were negatively associated with girl's physical activity . In this paper, we look at a range of crime statistics, including the prevalence of convicted sexual offenders in the area, and relate them to after school indoor electronic entertainment.
Other recent studies have emphasized the effect of contextual variables on health outcomes [29, 30]. However, none have examined crime and indoor sedentary activity. Our research, while related to the contextual literature, differs in that we examine a specific health behavior which could be directly attributable to a neighborhood characteristic rather than related to a general health outcome measure. Although childhood obesity measures could ultimately be related to neighborhood crime through their effect on physical activity, obesity is determined by a combination of behaviors including diet. Therefore, instead of focusing on obesity as an outcome, we focus on an important obesity-related behavior which is potentially directly related to neighborhood crime: indoor sedentary behavior.
The purpose of this study is to examine cross-sectional associations between indoor activities, namely computer use, video game playing, and television viewing, with community crime statistics for 4thgrade students in Texas. 4thgraders in the U.S. are typically 9 or 10 years old. Inexpensive technologies which promote sedentary indoor activities are abundant. We therefore hypothesize that community crime rates are positively associated with sedentary indoor activities among 4th-graders in Texas. The community crime statistics are robberies, all violent crimes, murders, assaults, property crimes, rapes, burglaries, larcenies and motor vehicle thefts. The number of released sexual offenders living in the area is also hypothesized to be positively associated with sedentary indoor activities.
The data used were the School Physical Activity and Nutrition III (SPAN). SPAN is a child obesity surveillance system developed and implemented by the University of Texas School of Public Health with support from the Texas Department of State Health Services during the 2004–2005 school year. As part of the SPAN project, student's eating behaviors, nutrition knowledge, and physical activity behaviors were surveyed. SPAN employed a stratified, multistage probability sample of public school children in Texas that included representative samples of White/other, Black or African-American, and Mexican-American-Latino or Hispanic youth. A full description of the SPAN study design and its participants has been documented for the first wave , and a brief summary is provided here. In 2004–2005 academic year, Texas was divided into 11 administrative Health Service Regions (HSR) (For a full description, see ). Schools which had fewer than 75 students or were charter, magnet or special schools were excluded from the sampling frame. Each HSR was divided into three strata: urban center, other urban/suburban and rural . Schools were randomly selected from HSR or from school districts depending on the strata and then at least two classes. We focused on the 4thgrade because children in the 8thand 11thgrades are much more independent. The sample size was 7,907, representing the 4thgrade population of 248,838. Approval for the SPAN study was obtained from the Committee for the Protection of Human Subjects at the University of Texas Health Science Center at Houston, as well as the Texas Department of State Health Services Institutional Review Board and participating school districts. Parents completed either an active or passive consent (depending on school district procedures for parental consent), and children completed a child assent form.
SPAN surveys were administered using standard protocols for recalled physical activity. Sedentary behavior was assessed by asking about the amount of the specific behavior on the previous day. The question was, 'Yesterday, how many hours did you watch TV or video movies away from school?'. Sedentary behavior was also assessed by asking about the amount of the specific behavior on an average day. One question was 'How many hours per day do you usually spend on the computer away from school?' (Time on the computer included time spent surfing the internet and instant messaging [on mobile or cell phones].). The other was 'How many hours per day do you usually spend playing video games like Nintendo, Sega, Play Station, Xbox, GameBoy or arcade games away from school?'. Response categories for each item included 1) none, 2) 1 hour, 3) 2 hours, 4) 3 hours, 5) 4 hours, 6) 5 hours, 7) 6 hours or more. Total sedentary behavior was defined as the sum of these 3 sedentary behaviors, which means there is a minimum of zero and a maximum of 18 hours. Other variables from SPAN were African-American (1 = yes, 0 = no), Hispanic (1 = yes, 0 = no), speaks Spanish at home (1 = yes, 0 = no), age, speaks language other than Spanish or English at home (1 = yes, 0 = no). The percentage disadvantaged at the school comes from the Texas Education Association (TEA) .
The crime variables came from two sources. First, we used the U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics data for the years 2000 through 2005 . The community crime statistics were robberies, all violent crimes, murders, assaults, property crimes, rapes, burglaries, larcenies, motor vehicle thefts, and sexual offenders living in the police department where the school is located. We included the average crime rates per hundred population for the years 2000 to 2005 in each police department. Police departments correspond to cities over 100,000 in population. Note that our six-year average meant that recent crimes "count" the same as crimes five years ago. Parents in a normally safe community suffering a recent up tick in crime may not bring their children inside, especially if the offending person(s) have been apprehended. Additionally, if criminal activity in an area declines after a long period of high crime activity, parents would likely remain reluctant to allow their children to play outside in the near term. Thus, criminal activity from five to six years ago should also count because it will influence parent's current decisions about their children's activity. Second, we included the number of sexual offenders, as of September 2007, living in the zip code address of the school . These data are listed by the Texas Department of Public Safety under the Texas Sex Offender Registration Program. The sexual offenders were per 100 population in the zip code of the school. We did not have the zip code of the student.
Overall, there were 160 schools with an average of 49.4 students per school. There were 111 unique zip codes for the schools. Schools were linked to 39 police districts reporting crime statistics.
All estimates and statistical tests were performed taking into account SPAN's sample design features. STATA (version 8.0, StataCorp LP, College Station, Texas) was used to analyze the data. Weighted means, population standard errors, and proportions for demographic characteristics were computed. Adjusted incremental relative risk (IRRs) for indoor sedentary behavior associated with crime levels and their corresponding 95% confidence intervals (95% CI) were calculated using sampling weighted Poisson regression models for each sedentary behavior measure (count of hours during the previous day). Adjustment variables were sex, African-American, age, percentage disadvantaged in school, Hispanic, Spanish spoken at home and language other than Spanish or English spoken at home. An a level of 0.05 was established a priori as the probability of incurring a type I error.
The mean age of 4thgrade students was 9.7 and was similar for boys and girls (Table 1 [See additional file 1]). Fifty-one percent of the 4thgraders were male, and 45 percent were Hispanic and approximately 25 percent spoke Spanish at home. Boys spent more hours in sedentary behavior per day than girls, spending 5.2 hours per day watching TV (0.206 standard error), playing video games, or using the computer compared to 3.6 hours per day (0.098 standard error) for girls. Boys watched slightly more TV than girls, but TV made up over half of the sedentary behavior for girls. Boys played video games almost three times more than girls, spending approximately two hours a day. Both boys and girls spent approximately an hour per day on the computer. Note that all sedentary behavior was outside of school.
Table 1 [See additional file 1] lists all of the per 100 population crime rates, averaged from 2000 to 2005 in the city of the school. Sexual offenders in the zip code of the school were included in per capita form.
Incremental relative risks for total daily hours of video game playing, adjusted by sex, African-American, Hispanic, speaks Spanish at home, age, speaks language other than Spanish or English at home and percentage disadvantaged at the school, were reported in Table 2 [See additional file 1]. For the total sample, sex offenders per capita was associated with increased video game playing (IRR 2.558, 95% CL 1.177, 5.560). For boys, sex offenders per capita was associated with increased video game playing (IRR 2.347, 95% CL 1.270, 4.337) as were burglaries per 100 population (IRR 1.246, 95% CL 1.054, 1.472). For girls, crime was not associated with increased video game playing. The magnitude of the association between the number of sexual offenders per capita and video game playing is quite large. When evaluated at the mean level of sex offenders per capita, a one percentage point increase in per capita sex offenders would be associated with an increase of approximately 1.2 hours of video game playing per day. For boys, a one percentage point increase in per capita sex offenders would be associated with an increase of approximately 1.7 hours of video game playing per day. The magnitude of the association between burglaries for boys were much lower. When evaluated at the mean level of burglaries per 100 population, a 10 percentage point increase would be associated with an increase of approximately 0.044 hours, or 2 minutes and 40 seconds, of video game playing per day.
Incremental relative risks for total daily hours of computer use, adjusted by sex, African-American, Hispanic, speaks Spanish at home, age, speaks language other than Spanish and English at home and percentage disadvantaged at the school, were reported in Table 2 [See additional file 1]. For the total sample and boys, none of the crime variables was associated with increased computer use. For girls, sex offenders per capita was protective against computer use (IRR 0.394, 95% CL 0.161, 0.964). When evaluated at the mean level of sex offenders per capita, a one percentage point increase in per capita sex offenders would be associated with a decrease of approximately 48 minutes of computer use by girls per day.
Incremental relative risks for total daily hours of TV viewing, adjusted by sex, African-American, His-panic, age, speaks language other than Spanish and English at home and percentage disadvantaged at the school, were reported in Table 2 [See additional file 1]. For girls, burglaries and hours of television watching were positively associated (IRR 1.186, 95% CL 1.038, 1.356) as were larcenies (IRR 1.051, 95% CL 1.001, 1.104). When evaluated at the mean level of burglaries per 100 population, a 10 percentage point increase would be associated with an increase of approximately 0.031 hours, or approximately 2 minutes of television viewing per day. When evaluated at the mean level of larcenies per 100 population, a 10 percentage point increase would be associated with an increase of approximately 0.032 hours, or approximately 54 seconds of television viewing per day.
Our results showed that 4thgrade boys played more hours of video games in communities with increased levels of larceny and burglary crimes as well as for increased per capita sexual offenders; 4thgrade boys played more hours of video games for increased levels of per capita sexual offenders; 4thgrade girls were more likely to watch television in areas with more burglaries and larcenies. Computer use was negatively associated with per capita sexual offenders among girls. However, it was not significant for boys, even though boys used the computer more hours per day than girls. Community crime rates were not associated with total indoor sedentary behavior.
Our results revealed that community crime rates affected whether girls and boys engaged in indoor sedentary behavior, although the types of activities differed by gender. We had expected the crime rates to affect indoor sedentary behavior for boys more than girls because the latter have traditionally played, and continue to play, indoors more than boys . So while it is not surprising to see girls using leisure technologies, the effect of crime in neighborhoods is more surprising due to the smaller time spent playing outdoors before the development of leisure technologies. It is likely that time spent with leisure technologies is being substituted for the small outdoor play time alloted to girls relative to boys, especially in high-crime neighborhoods.
Sexual offenders per capita were protective against computer use for 4thgrade girls. Parents in these communities may be reluctant to allow their girls to be potentially exposed to online sexual advances.
Sexual crimes, particularly those against children, are highly emotional, which may partly explain the greater influence on indoor sedentary behavior in our sample. However, there is another potential explanation. Perceptions of crime may not coincide with actual crime rates in an area [37–39]. By Texas law, jurisdictions can promote the presence of sexual offenders through mailings, advertisements, and the inter-net . Therefore, while the public may overstate the danger posed by sexual offenders, their knowledge of the presence of offenders may be more accurate than their knowledge of the prevalence of other crime measures.
New technologies which complement sedentary behavior may be having subtle effects on outdoor play, even for children without access to technologies promoting sedentary behavior. For instance, there may be "threshold effects" associated with children's outdoor play which have been negatively affected by the availability of new technologies. If new video game technologies have drawn children indoors, especially in neighborhoods with high crime rates, thresholds may not be met. Then, even children without video games and cable TV have less opportunity to play outdoors because many of their peers are indoors. Further, with fewer children playing outside, perhaps parents will not allow their children to play freely outside-parents may believe there is safety in numbers in regards to criminal activity. For all of these reasons, today's children may face decreased numbers of playmates available outdoors.
If parental fear of crime in their communities, in conjunction with the development of inexpensive leisure activity technologies for children, is tipping the balance towards indoor sedentary play, then there are important policy implications. Crime will never be completely eliminated; the development of video games, computers, and ever more television programming will continue apace. However, children still likely prefer to socialize, play team sports and to roam free outdoors. Therefore, after school programs promoting physical activity which feature adult supervision should be expanded because they will ease parental fears about crime. Team sports, which in years past would form spontaneously, must be organized by parents. Indeed, the secular trends are towards more time spent in supervised sports and activities among children, perhaps because of safety concerns due to crime . As noted earlier, access to safe play environments has been shown to be associated with increased physical activity 
Because the SPAN survey takes place during the school year, our results likely underestimate the true association between crime and indoor sedentary activity because more time is spent indoors in the summer. In Texas, where this study takes place, the percentage of houses with air conditioning has reached 90 percent . Heat-related mortality declined significantly from the 1960s to the 1990s due to air conditioning, especially in the Southeast and Southwest . Therefore, indoor play has become preferable relative to outdoor play due to temperatures in the summer.
Our results have limitations. Of course, a large cohort study would have been preferred to compare changes in crime and their effects on physical activity. Another problem is that computer use for homework cannot be excluded from our total sedentary and our total hours of computer use measure. It is possible that computers are used for homework in low crime areas. Further, the number of sexual offenders per capita is in the zip code area of the school, but it may be different in the residential area where the children live. Our crime statistics are likely imperfect estimates of neighborhood crime. Our crime rates for robberies, all violent crimes, murders, assaults, property crimes, rapes, burglaries, larcenies and motor vehicle thefts, are estimated by city police department jurisdiction. In several instances, two or more schools are located within the same police department jurisdiction. However in almost all cases, schools had unique per capita sexual offender estimates.
Our controls for socioeconomic status are imperfect because of the diffculty in asking children questions related to income. Further, income itself has conflicting effects on indoor sedentary behavior. Although the technologies are more affordable, it is possible that many wealthy suburban communities have gated residential lots which would lead children to substitute outdoor play for indoor sedentary activity. It has been long noted in the area of urban economics that wealthier families live in the suburban areas with large lots [42, 43]. However, we do not have information on residential lot size nor whether the child's yard is gated, both of which likely are conducive to outdoor play.
Children most likely enjoy outdoor activities. Our results show that parental fears of crime may be as important as child preferences for sedentary leisure. Therefore, policies which alleviate parental fears may increase childhood physical activity.
Mokdad AH, Marks JS, Stroup DF, Gerberding JL: Actual causes of death in the United States, 2000. JAMA : The Journal of the American Medical Association. 2004, 291 (10): 1238-1245. 10.1001/jama.291.10.1238.
National Center for Health Statistics: Prevalence of overweight among children and adolescents: United States, 1999. National Center for Health Statistics. 1999, [http://www.cdc.gov/nchs/products/pubs/pubd/hestats/overwght99.htm]
Troiano RP, Flegal KM: Overweight children and adolescents: Description, epidemiology, and demographics. Pediatrics. 1998, 101 (3 Pt 2): 497-504.
Campaigne BN, Morrison JA, Schumann BC, Falkner F, Lakatos E, Sprecher D, Schreiber GB: Indexes of obesity and comparisons with previous national survey data in 9- and 10-year-old black and white girls: The National Heart, Lung, and Blood Institute Growth and Health Study. The Journal of Pediatrics. 1994, 124 (5 Pt 1): 675-680. 10.1016/S0022-3476(05)81354-X.
Dwyer JT, Stone EJ, Yang M, Feldman H, Webber LS, Must A, Perry CL, Nader PR, Parcel GS: Predictors of overweight and overfatness in a multiethnic pediatric population. Child and Adolescent Trial for Cardiovascular Health Collaborative Research Group. The American Journal of Clinical Nutrition. 1998, 67 (4): 602-610.
Hoelscher DM, Day RS, Lee ES, Frankowski RF, Kelder SH, Ward JL, Scheurer ME: Measuring the prevalence of overweight in Texas schoolchildren. American Journal of Public Health. 2004, 94 (6): 1002-1008.
Datar A, Sturm R: Physical education in elementary school and body mass index: evidence from the early childhood longitudinal study. American Journal of Public Health. 2004, 94 (9): 1501-1506.
Kaestner R, Xu X: Effects of Title IX and Sports Participation on Girls' Physical Activity and Weight. NBER Working Paper 12113. 2006
Berkey CS, Rockett HRH, Field AE, Gillman MW, Frazier AL, Camargo CA, Colditz GA: Activity, Dietary Intake, and Weight Changes in a Longitudinal Study of Preadolescent and Adolescent Boys and Girls. Pediatrics. 2000, 105 (4): e56-10.1542/peds.105.4.e56.
Hernandez B, Gortmaker S, Colditz G, Peterson K, Laird N, Parra-Cabrera S: Association of obesity with physical activity, television programs and other forms of video viewing among children in Mexico City. International Journal of Obesity. 1999, 23: 845-854. 10.1038/sj.ijo.0800962.
Utter J, Neumark-Szainer D, Jeffery R, Story M: Couch potatoes or french fries: Are sedentary behaviors associated with body mass index, physical activity, and dietary behaviors among adolescents?. Journal of the American Dietetic Association. 2003, 103 (10): 1298-1305. 10.1016/S0002-8223(03)01079-4.
Page A, Cooper A, Stamatakis E, Foster L, Crowne E, Sabin M, Shield J: Physical activity patterns in nonobese and obese children assessed using minute-by-minute accelerometry. International Journal of Obesity. 2005, 29: 1070-1076. 10.1038/sj.ijo.0802993.
Kim J, Must A, Fitzmaurice GM, Gillman MW, Chomitz V, Kramer E, McGowan R, Peterso KE: Relationship of Physical Fitness to Prevalence and Incidence of Overweight among Schoolchildren. Obesity Research. 2005, 13 (7): 1246-1254. 10.1038/oby.2005.148.
Kaur H, Choi WS, Mayo MS, Harris KJ: Duration of television watching is associated with increased body mass index. The Journal of Pediatrics. 2003, 143 (4): 506-511. 10.1067/S0022-3476(03)00418-9.
Delva J, Johnston LD, O'Malley PM: The epidemiology of overweight and related lifestyle behaviors: racial/ethnic and socioeconomic status differences among American youth. American Journal of Preventive Medicine. 2007, 33 (4 Suppl): S178-86. 10.1016/j.amepre.2007.07.008.
Henderson VR: Longitudinal associations between television viewing and body mass index among white and black girls. The Journal of Adolescent Health. 2007, 41 (6): 544-550. 10.1016/j.jadohealth.2007.04.018.
Must A, Bandini LG, Tybor DJ, Phillips SM, Naumova EN, Dietz WH: Activity, inactivity, and screen time in relation to weight and fatness over adolescence in girls. Obesity. 2007, 15 (7): 1774-1781.
Epstein LH, Roemmich JN, Robinson JL, Paluch RA, Winiewicz DD, Fuerch JH, Robinson TN: A randomized trial of the effects of reducing television viewing and computer use on body mass index in young children. Archives of Pediatrics & Adolescent Medicine. 2008, 162 (3): 239-245. 10.1001/archpediatrics.2007.45.
Levitt SD, Dubner SJ: Freakonomics : A rogue economist explores the hidden side of everything. 2005, New York: William Morrow
Holloway D: Tracking TV: New numbers show shifts in how we're watching (and how much). American-Statesman. 2007
Pew Charitable Trust: Pew Internet and American Life Survey 2007. Tech. rep., Pew Charitable Trust. 2007
Davis RE, Knappenberger PC, Michaels PJ, Novicoff WM: Changing heat-related mortality in the United States. Environmental Health Perspectives. 2003, 111 (14): 1712-1718.
Carver A, Timperio A, Crawford D: Playing it safe: The influence of neighbourhood safety on children's physical activity-A review. Health & Place. 2007
Weir L, Etelson D, Brand DA: Parents' perceptions of neighborhood safety and children's physical activity. Preventive Medicine. 2006, 43 (3): 212-217. 10.1016/j.ypmed.2006.03.024.
Carver A, Salmon J, Campbell K, Baur L, Garnett S, Crawford D: How do perceptions of local neighborhood relate to adolescents' walking and cycling?. American Journal of Health Promotion. 2005, 20 (2): 139-147.
Timperio A, Crawford D, Telford A, Salmon J: Perceptions about the local neighborhood and walking and cycling among children. Preventive Medicine. 2004, 38: 39-47. 10.1016/j.ypmed.2003.09.026.
Molnar BE, Gortmaker SL, Bull FC, Buka SL: Unsafe to play? Neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents. American Journal of Health Promotion. 2004, 18 (5): 378-386.
Gomez JE, Johnson BA, Selva M, Sallis JF: Violent crime and outdoor physical activity among inner-city youth. Preventive Medicine. 2004, 39 (5): 876-881. 10.1016/j.ypmed.2004.03.019.
Roux AVD: The study of group-level factors in epidemiology: Rethinking variables, study designs, and analytical approaches. Epidemiologic Reviews. 2004, 26: 104-111. 10.1093/epirev/mxh006.
Roux AVD: Estimating neighborhood health effects: The challenges of causal inference in a complex world. Social Science & Medicine (1982). 2004, 58 (10): 1953-1960.
Hoelscher D, Day R, Kelder S, Ward J: Reproducibility and validity of the secondary level School-Based Nutrition Monitoring student questionnaire. Journal of the American Dietary Association. 2003, 103 (2): 186-194. 10.1053/jada.2003.50031.
Hoelcher D, Pérez A, Kelder S, Day R, Sanders J, Frankowski R, C CS, Lee E: Leveling off the prevalence of overweight in elementary school children in Texas: Results from the 2000–2002 and 2004–2005 School Physical Activity and Nutrition (SPAN) Surveillance System. 2008
Texas Education Agency: Snapshot 2000 summary tables: State totals. 2000, Accessed on 4-25-2002, [http://www.tea.state.tx.us/perfreport/snapshot/2000/state.html]
US Department of Justice: [http://bjsdata.ojp.usdoj.gov/dataonline/Search/Crime/Crime.cfm]
Texas Department of Public Safety: [https://records.txdps.state.tx.us/DPS_WEB/Sor/index.aspx]
Soori H, Bhopal RS: Parental permission for children's independent outdoor activities. Implications for injury prevention. European Journal of Public Health. 2002, 12 (2): 104-109. 10.1093/eurpub/12.2.104.
Sampson RJ, Raudenbush SW: Seeing Disorder: Neighborhood Stigma and the Social Construction of "Broken Windows. Social Psychology Quarterly. 2004, 67 (4): 319-342.
Quillian L, Pager D: Black Neighbors, Higher Crime? The Role of Racial Stereotypes in Evaluations of Neighborhood Crime. The American Journal of Sociology. 2001, 107 (3): 717-767. 10.1086/338938.
Fear itself. Economist. 2003, 368 (8333).
Texas Department of Public Safety: [http://www.txdps.state.tx.us/administration/crime_records/pages/sexoffender.htm]
Sturm R: Childhood obesity – what we can learn from existing data on societal trends, part 2. Preventing Chronic Disease. 2005, 2 (2): A20.
Muth RJ: Cities and Housing. The Spatial Pattern of Urban Residential Land Use. 1969, Chicago: The University of Chicago Press
Alonso W: Location and Land Use. 1964, Cambridge, Mass: Harvard University Press
Funding for this project came from the Hispanic Health Research Center, supported by: NIH CMHD P20 MD000170-03 located at the Brownsville Regional Campus of the University of Texas School of Public Health. Special thanks to the Texas Department of State Health Services and the participants of the SPAN seminar series for helpful comments.
The authors declare that they have no competing interests.