A multilevel investigation on the socio-demographic and urban neighborhood effects on out of school physical activity among inner city minority adolescents.
Subject: Teenagers (Investigations)
Youth (Investigations)
Physical fitness (Investigations)
Authors: Yan, Fang
Voorhees, Carolyn C.
Zhang, Guangyu
Beck, Kenneth H.
Huang, Shuo
Wei, Hua
Pub Date: 06/22/2011
Publication: Name: American Journal of Health Studies Publisher: American Journal of Health Studies Audience: Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2011 American Journal of Health Studies ISSN: 1090-0500
Issue: Date: Summer, 2011 Source Volume: 26 Source Issue: 3
Topic: Event Code: 980 Legal issues & crime Computer Subject: Company legal issue
Product: Product Code: E121930 Youth
Accession Number: 308741515

Many youth today are physically inactive. Considerable evidence documents that nearly 35% of youth in the US fail to meet the minimum physical activity guidelines, and another 14% are completely inactive (CDC, 1997; USDHHS, 2000). Engagement in physical activity declines significantly during the high school years (Berrigan & Troiano, 2002). Low levels of physical activity and the failure to meet physical activity recommendations have notable consequences among adolescents, including increased risk of obesity (Trost et al, 2001a), low bone density (Doak et al, 2006), and low physical fitness (Maziak, Ward & Stockton, 2008). Furthermore, adolescents who are not physically active are denied the positive social and emotional benefits of physical activity, including higher self-esteem, lower anxiety, and lower stress (Pretty et al, 2005). The observed increasing prevalence of obesity and overweight is particularly striking for minority adolescents (Ogden et al, 2002). Studies have documented that inner-city adolescents experience high levels of life stress, poverty, and exposure to violence (Garbarino et al, 1992; Weist et al, 2001). These stressors can negatively affect healthy adolescent development and are associated with higher rates of emotional and behavioral problems and psychopathology (Tolan & Henry, 1996). Thus, research efforts must focus on identifying environmental factors that contribute to inactivity, particularly inner-city minority populations who are more likely to suffer adverse health consequences.

Off-school time provides a challenge and an opportunity to promote moderate-intensity physical activity to adolescents. Most school-based intervention projects have generally produced disappointing results with respect to improving body composition (Baranowski et al, 1998; Larry et al, 2008). Among the explanations postulated for why these obesity reduction programs have not achieved their full prevention potential is that they fail to alter the environment in a way that is conducive to such change (Doak et al, 2006). We need to examine the neighborhood environmental factors that influence adolescents' physical activity behaviors outside of school. Given that students consume 67%-75% of their daily energy (Farris et al, 1992) and accrue 70%-80% of their daily physical activity away from school (Myers et al, 1996; Ross & Gilbert, 1985), efforts to understand the environmental factors that may influence adolescents' physical activity behaviors could have large cumulative impact or effects.

To address the problem of physical inactivity and its association with the environmental factors, programs and policies have begun to focus on ecological models. Behavior-specific and context-specific ecological models assume that behavior can be better predicted when there is greater correspondence between a behavior outcome measure and the specific environmental and personal variables hypothesized to be associated with that behavior (Giles-Corti et al, 2005; Sallis et al, 2002; Sallis & Owen, 1997). Among a number of reviews that examined links between the built environment and adolescents' physical activity, very few studies have simultaneously assessed individual sociodemographic factors, and both the physical and social dimensions of neighborhood environment and their associations with moderate-to-vigorous physical activity (MVPA) in an urban inner city setting, a setting usually confounded by socioeconomic status (SES). Furthermore, much less emphasis has been placed on research specific to urban adolescents. Among many inner city specific built environment features, which ones are more important remain unknown, an issue that therefore deserves further investigation. The purpose of this study is to examine the association of individual and neighborhood characteristics with physical activity, particularly out-of-school MVPA, among a sample of minority adolescents living in an urban environment. Given the Healthy People 2010 (DHHS, 2000) major goals to eliminate health disparities in terms of higher prevalence of physical inactivity, and recent recommendations to focus on environmental determinants of physical activity, our study is important to elucidate this relationship (Sallis & Glanz, 2006; Sallis et al, 1998).



Study participants were drawn from the Baltimore Active Living Teens Study (BALTS). BALTS is a cross-sectional study that investigated the effects of multi-level risk and protective factors on moderate-to-vigorous physical activity (MVPA) in a sample of predominately African American urban high school students (grades 9 through 12). Participants were solicited through in-class presentations about the study. A detailed description of recruitment can be seen elsewhere (Ries et al, 2008). The recruitment rate was 54%, which was based on 649 students who were recruited and 350 who agreed to participate. A comparison of demographic variables of those who declined and those who participated suggested no difference between the 2 groups. Students' socio-demographic characteristics were assessed through a Web-based survey. The participants replied to the Web survey from school computers and were able to re-take the survey if any technical problems prevented them from continuing (technical problems were extremely rare). The survey took approximately one hour to administer. Physical activity levels were objectively measured using standardized protocols described below. Due to the geographic nature of this analysis, participants were asked to provide their home address, where they lived most of the time in Baltimore city, so that their individual information could be matched to their neighborhood of residence. Exclusions included: (1) thirty three students (9%) who had missing information for their physical activity levels, (2) twenty five students (6.8%) who had missing demographic information, and (3) twelve students (3.2%) whose addresses were outside the Baltimore area. Therefore, the final working sample contained 297 (81%) students. There were no demographic differences between those included and those excluded in the analysis. The sample represents a broad geographic area of Baltimore City. Each participant's parent or guardian provided written, informed consent and all subjects assented to participation. Each participant received incentives valued at $15 for participation in each measurement visit. The Institutional Review Board approved the study. The data were collected from January to June 2006.


Dependent variable

Moderate-to-Vigorous Physical Activity measurement and Data reduction

Physical activity was measured by Actigraph accelerometers (Computer Science Applications) for seven days. The Actigraph accelerometer is widely accepted as a valid means of physical activity assessment (Ainsworth et al, 2003; Pate et al, 2002; Thompson et al, 2005; Trost et al, 2001; Trost et al, 2003; Trueth et al, 2004). Following a standardized protocol, each monitor was initialized prior to placing it on a belt to be worn on each participant's waist on their right hip. Participants were asked to wear it all the time, except at night while sleeping and while bathing or swimming, during three consecutive monitoring days. Activity counts were stored in 30-second time intervals. Students who failed to comply with minimal wear, had a monitor malfunction, or left fewer than 3 days of data (or non useable data) were asked to wear the monitor again until useable data were collected.

Actigraph counts were summarized by quantifying the time (minutes) spent at different intensity levels. The BALTS thresholds for the activity intensities were less than 50.99 counts for sedentary activity, 51 to 578.99 counts for light activity, and 579 or more counts for MVPA. The threshold of 579 or more counts for MVPA corresponds approximately to the lower bound for a 2.5-mph walk, representing an activity intensity level of three metabolic equivalents (METs). A 3-MET cut point to define MVPA was used because it has been used as the threshold for MVPA in previous studies of youth (Ainsworth et al, 2003; Pate et al, 2002; Thompson et al, 2005; Trost et al, 2003; Trueth et al, 2004). Accelerometer data reduction methods incorporated the following data processing issues suggested by literature (Masse et al, 2005): (1) individual records that either had valuse greater than 16,000 counts (max value for accelerometer) or constant and consecutive nonzero records for 10 minutes were excluded from the analysis, (2) valid wearing time was determined by subtracting the invalid minutes (i.e., interruption) from the total wearing minutes. The total weekday out-of-school minutes were the accumulated, valid wearing minutes used before (5 a.m.-8:30 a.m) and after school (3 p.m.-10 p.m) combined. Interruption was estimated using 20 minutes (Saksvig et al, 2007; Treuth et al, 2004; Treuth, Sherwood, & Butte, 2003) of continuous zero counts. (3) The minimal wear requirement for a valid out-of-school day was 6 hours (Jackson et al, 2003). Moderate-to-vigorous physical activity (MVPA) was used as an outcome for this study. The percentage of time spent in MVPA was defined as the combined valid MVPA minutes divided by the total valid physical activity time. Accelerometers' data were only used before (5 a.m.-8:30 a.m) and after school (3 p.m.-10 p.m) and on weekdays, because we hypothesized that the neighborhood environmental factors would only affect non-school activity.

Independent variables

Individual level sociodemographic variables

Demographic variables included age, gender, ethnicity, grade, parents' education, and participants' current health status.

Neighborhood level variables/indicators

Objective measures of the neighborhood built environment were collected using established Geographic Information System (GIS) databases, which contained vital signs/outcome indicators of the Baltimore City's Community Statistical Areas (http://www.ubalt.edu/bnia/indicators/reports. html) (Schachtel, 2001). Baltimore has 55 Community Statistical Areas (CSAs) (Source: Baltimore Neighborhood Indicator Alliance, http://www.ubak. edu/bnia/indicators/statistical_profiles.html). The current study was based upon those CSAs. Figure 1 demonstrates the distribution of students in those community statistical areas. Respondents' home addresses and addresses of schools were collected and geocoded with ArcGIS software; the real "city neighborhood" and community boundaries were census-based and were used to define the border of the neighborhood level analysis in which students were nested. All neighborhood variables were aggregated at the Baltimore Community/neighborhood levels. Baltimore city specific GIS based indicators (France, 2002; Schachtel, 2001) characterizing the participants' neighborhoods focused on five categories: 1) housing and community development, 2) children and family safety and violence and crime, 3) pleasantness, 4) workforce and Economic Development, and 5) transportation time to work. Four items that measured housing conditions and community development at the neighborhood level are: (a) mean of vacant and abandoned property rate year 01-04 (Vac_aban_mean), (b) mean of median house sale price of year 00-04 (pricemean), (c) racial diversity index of year 2000 (race_divers00), and (d) economic diversity index of year 2000 (econ_diverse00). The percent of residential properties that are vacant and abandoned at year's end is an indicator of the housing conditions. This indicator reflects the number of properties designated as vacant and abandoned out of all residential properties in a given year; it reflects specifically the blighted homes that are dilapidated, unlivable, and boarded. Such an indicator is a very visible way to know about housing conditions. The median sale price is an indicator of property value and demand for homes in an area. The racial diversity index is an indicator of the degree of racial diversity in an area. The economic diversity index is an indicator of the degree of economic diversity in an area. Five household income ranges are used in Economic Diversity Index measures: less than $25,000; $25,000-39,999; $40,000-$59,999; $60,000-$74,999; $75,000 and higher. Children and family safety and violence and crime condition of neighborhood were measured by three indicators: (a) mean of juvenile arrest rates of year 00-04 (arrest_mean), (b) mean of crime rates of year 00-04 (crime_mean), and (c) mean of domestic violence rates of year 00-04 (violence_mean). The mean of dirty streets and alleys rates of years 02-04 (mean_dirtystreet) and percent of area covered by tree canopy of year 2000 (tree01) were used to measure neighborhood pleasantness. Unemployment rate (2000) was used to measure workforce and Economic Development. A single variable that estimates the percent of travel time more than 30 min to work (travel_long) was used to measure transportation time to work. A complete list of the indicators including categories, definitions, sources, and each indicator's mean and standard deviation are shown in Table 1. The prevalence of indicators was investigated with respect to relevant neighborhood characteristics (e.g., violent crime counts, vacant and abandoned property rates). This spatial measure was calculated by summarizing the neighborhood environment variables/indicators within a 0.5-mile radius buffer around participant's homes.


Individual address geo-coding and neighborhood spatial analyses were conducted with ArcGIS 9.3 (ESRI, Redlands, Calif.). All statistical analyses were conducted using SAS version 9.2 (Cary, N.C.). This study examined the association of individual and neighborhood characteristics with out-of-school MVPA among minority adolescents living in an urban environment. Our data exemplify a naturally occurring hierarchical or nested structure in which individuals (level 1 units) lived within the neighborhood (level 2 units); linear mixed effects models are employed to study the joint effects of individual level and neighborhood level factors on MVPA. Bivariate and multivariate analyses were used. To control for the intra-class correlation, we included neighborhood as a random intercept in the hierarchical linear models (HLMs).




The sample consisted of adolescents between 14 and 18 years, with an average age of 15.7 years old. The majority of the participants were African American (69%) and females (59.9%) with a fairly equivalent distribution across grades 9th, 10th, and 12th, with the exception of fewer 11th graders. Current health status was measured on a 1 to 5 scale (1 = Excellent, 2 = Very good, 3 = Good, 4 = Fair, 5 = Poor) with a mean of 2.27. Participants were physically active for 6.8% of their leisure times (data not presented in Table). More than half (61.6%) of the participants' mothers obtained educations at some college or higher levels while 35.3% of participants' fathers' obtained the same levels of education. Descriptive statistics are reported in Table 2.


Results from the bivariate (data not shown) and multivariate mixed model (Table 3) analyses displayed the associations of individual sociodemographic characteristics and neighborhood characteristics with out-of-school MVPA. Both models showed that the increased age, being female were associated with decreases in MVPA. In the multivariate model, when adolescents become one year older, about 0.65% their leisure time goes from active to inactive (P < 0.001). Girls spend 1.27% less leisure time for moderate and vigorous activity than boys do (P = 0.01). And Blacks spend 1.25% less leisure time for moderate and vigorous activity than other groups (p<0.001). After controlling for personal-level characteristics, MVPA is also affected by the neighborhood-level variables. Percent of population with travel time more than 30 minutes (p < .0001), abandoned property rate (p < 0.05), were negatively related to adolescents' moderate and vigorous activities respectively.


This study simultaneously modeled the associations between objective-measured out-of-school physical activity and a series of individual and inner city specific neighborhood environmental factors in an urban, predominantly minority adolescent sample. Results from this study indicated that both individual and neighborhood characteristics are significant predictors of physical activity in the sample. Particularly, the influence of individual demographic and neighborhood social-economic determinants, and transportation time to work outweighed the role played by racial composition, violence, crime, and neighborhood pleasantness of physical activity. In addition to age and race, transportation time to work and the neighborhood's abandoned and vacant housing appeared more important in being directly associated with whether or not moderate-to-vigorous physical activity was achieved.

Our study contributes to the growing literature on social and environmental determinants of health disparities. Social and environmental determinants are inequitable and directly related to the historical and current unequal distribution of social, political, economic, and environmental resources (CDC, 2008). In our study, African Americans and females were at a higher risk for physical inactiveness when compared to their counterparts. The finding is consistent with the literature that African Americans were 50% less likely to engage in physical activity as Non-Hispanic Whites (CDC, 2010). Particularly, low-income African American girls face higher risk for physical inactivity, having limited or no access to physical activity opportunities. These groups are experiencing significant disparities in rates of chronic disease, therefore, promoting physical activity in minority populations is an effective strategy for reducing such health disparities (Klebanoff & Muramatsu, 2002).

Transportation time to work was an interesting determinant of PA discovered in our study. The increased percentage of residents living in the community who spend more than 30 minutes traveling to work could lead to the decrease in time adolescents spend on MVPA. Our finding contributes to the ongoing debate on the public's nuanced views of "time crunch" issues that may contribute to the obesity problems. In a national study, Hughes and colleagues (2010) surveyed program directors from 1583 Head Start program (Nation's largest federally funded education program for preschool children) across the country and found many barriers to addressing obesity in children, including lack of money and time. For example, 70% reported that parents did not have time to participate in physical activity with children at home. In addition, in a California study, Lopez-Zetina and colleagues (2006) found obesity to be associated with urban transportation indicators related to automobile use. Similar patterns were observed between obesity and physical inactivity and commute time. For adults, long hours of commuting time cut into time for physical activity, let alone participating in physical activity with children at home. For adolescents, greater work intensity and long commute time translate into more time in unstructured social activities and less time in sports (Feldman et al, 2003; Safron et al, 2001). Our findings directly support the hypothesis found in the previous study--increased car commuting time is likely to impact the amount of physical activity youth engage in daily physical activity (Cohen et al, 2006).

In our study, vacant and abandoned property rates decreased adolescents' time spent on MVPA. This finding elucidates an urban-suburban paradox. Given a built environment in many US inner cities and urban neighborhoods that include sidewalks and mixed land uses, and gridded street patterns which foster connectivity, we might expect that rates of physical activity would be more favorable in inner-city neighborhoods. But there appears, instead, a paradox whereby obesity and physical inactivity are more prevalent among inner-city residents than among suburbanites (Lopez & Hynes, 2006). Some, if not much, of this paradox may be explained by other factors, particularly the social economic challenges commonly faced by urban minorities. For example, the social effect of abandonment has been termed "the broken window syndrome": neighborhoods with broken windows and dilapidated housing encourage crime, pose safety hazards, isolate residents and reduce trust in the ability of the neighborhood to meet its challenges (Cohen et al, 2000; Wilson & Kelling, 1982). A consequence of this constellation of risk from a moldering environment would likely be reduced with walking, physical activity and recreation in public. Even when abandoned buildings are demolished, the resulting vacant land poses problems to the environment (Cohen, Farley, & Mason, 2003). If untended, these lots usually become overgrown with weeds and covered with litter--an invitation for illegal dumping, and fostering criminal activity. Our investigation raised an interesting question for future studies: "does vacancy contribute to less walking, playing, and other physical activity in the affected neighborhood?" This question, like its companion regarding abandoned housing, has yet to be examined and confirmed.


This study had limitations that merit noting. First, we cannot assess a causal relationship due to the cross-sectional nature of the data. Second, this study only utilizes the objective measures of the neighborhood environment. Ideally, we would prefer to have objective-measured neighborhood conditions and be able to link people's perceptions and objective neighborhood conditions to actual physical activity. In future studies, we will address these limitations.

Despite these limitations, this study has several major strengths. First, it was guided by the social ecological framework, which is important in understanding physical activity behavior in social and behavioral science research (Sallis & Owen, 1997; 1999). Second, the study included objective measures of physical activity outcomes (i.e., MVPA). Third, our findings will provide evidence for future community-built environment interventions because adolescents engage in a variety of activities after school and their ability to meet the recommended moderate to vigorous levels of exercise depends upon a combination of environment and individual factors (Clifton 2003).


We gratefully thank Berenic Rushovich, MSW, project coordinator, and the school principals, students, teachers, and the parents and guardians who made this study possible. Special thanks to Ivel Turkson and Gentry Kuehn for their editorial work on this manuscript. This study was supported by a research grant (PI: Voorhees) and a dissertation award (PI: Yan) from the Robert Wood Johnson Foundations' Active Living Research Program. It was approved by the University of Maryland Institutional Review Board.


Ainsworth, B. E., Wilcox, S., Thompson, W. W., Richter, D. L., & Henderson, K. A. (2003). Personal, social, and physical environmental correlates of physical activity in African-American women in South Carolina. American Journal of Preventive Medicine, 25(3, Supplement 1), 23-29.

Baranowski, T., Anderson, C. B., & Carmack, C. (1998). Mediating variable framework in physical activity interventions. How are we doing? How might we do better? American Journal of Preventive Medicine, 15(4), 266-297.

Berrigan, D., & Troiano, R. P. (2002). The association between urban form and physical activity in U.S. adults. American Journal of Preventive Medicine, 23(2, Supplement 1), 74-79. CDC. (1997). Youth Risk Surveillance Survey - United States. MWR Morb Mortal Wkly Rep, 47(SS-3), 1-89.

CDC. (2008). Health disparities among racial/ethnic populations.

CDC. (2010). Health, United States, 2009. Retrieved April 17, 2010, from http://www.cdc.gov/nchs/data/ hus/hus09.pdf

Clifton, K. J. (2003). Independent Mobility among Teenagers: An Exploration of Travel to After-school Activities. Transportation Research Record, 1854, 74-80.

Cohen, D., Ashwood, J. S., Scott, M. M., Overton, A., Evenson, K. R., Voorhees, C. C., et al. (2006). Proximity to school and Physical Activity Among Middle School Girls: The Trial of Activity for Adolescent Girls Study. Journal of Physical Activity & Health, 3(Suppl 1), S129-138.

Cohen, D., Farley, T., & Mason, K. (2003). Why is poverty unhealthy? Social and physical mediators. Social Science and Medicine, 57(9), 1631-1641.

Cohen, D., Spear, S., Scribner, R., Kissinger, P., Mason, K., & Wildgen, J. (2000). Broken Windows and the risk of gonorrhea. American Journal of Public Health, 90(230-236).

DHHS. (2000). Healthy People 2010.

Doak, C. M., Visscher, T. L., Renders, C. M., & Seidell, J. C. (2006). The prevention of overweight and obesity in children and adolescents: a review of interventions and programmes. Obesity Reviews, 7, 111-136.

Doak, C. M., Visscher, T. L. S., Renders, C. M., & Seidell, J. C. (2006). The prevention of overweight and obesity in children and adolescents: a review of interventions and programmes. Obesity Reviews, 7, 111-136.

Farris, R. P., Nicklas, T. A., Webber, L. S., & Berenson, G. S. (1992). Nutrient contribution of school lunch program: implications for Healthy People 2000. Journal of School Health, 62, 180-184.

Feldman, D. E., Barnett, T., Shrier, I., Rossignol, M., & Abenhaim, L. (2003). Is Physical Activity Differentially Associated With Different Types of Sedentary Pursuits? Archives Pediatric Adolescent Medicine, 157, 797-802.

Garbarino, J., Dubrow, N., Kostelny, K., & Pardo, C. (1992). Children in Danger: Coping With the Consequences of Community Violence. San Francisco, Calif: Jossey-Bass.

Giles-Corti, B., Timperio, A., Bull, F., & Pikora, T. (2005). Understanding physical activity environmental correlates: Increased specificity for ecological models. Exercise and Sport Science Review, 33(4), 175-181.

Hughes, C. C., Gooze, R. A., Finkelstein, D. M., & Whitaker, R. C. (2010). Barriers to Obesity Prevention in Head Start. Health Affairs, 29(3), 454-462.

Jackson, D. M., Reilly, J. J., Kelly, L. A., Montgomery, C., Grant, S., & Raton, J. Y. (2003). Objectively measured physical activity in a representative sample of 3- to 4-year-old children. Obesity. Research., 11, 420-425.

Jacob France. (2002). Vital Signs for Baltimore Neighborhoods Report. Baltimore, Md: Baltimore Neighborhood Indicators Alliance.

Klebanoff R., & Muramatsu, N. (2002). A Community-Based Physical Education and Activity Intervention for African American Preadolescent Girls: A Strategy to Reduce Racial Disparities in Health. Health Promotion Practice, 3(2), 276-285.

Larry, S. W., Diane, J. C., Leslie, A. L., David, M. M., Charlotte, A. P., Deborah, R. Y., et al. (2008). Promoting Physical Activity in Middle School Girls: Trial of Activity for Adolescent Girls. American Journal of Preventive Medicine, 34(3), 173-184.

Lopez-Zetina, J., Lee, H., & Friis, R. (2006). The link between obesity and the built environment. Evidence from an ecological analysis of obesity and vehicle miles of travel in California. [doi: DOI: 10.1016/j. healthplace.2005.09.001]. Health & Place, 12(4), 656-664.

Lopez, R. P., & Hynes, H. P. (2006). Obesity, physical activity, and the urban environment: Public Health research needs. Environmental Heath, 5(25).

Masse, L. C., Fuemmeler, B. F., Anderson, C. B., Matthews, C. E., Trost, S. G., Catellier, D. J., et al. (2005). Accelerometer data reduction: a comparison of four reduction algorithms on select outcome variables. Medicine & Science in Sports & Exercise, 37 (11 Suppl), S544-S554.

Maziak, W., Ward, K. D., & Stockton, M. B. (2008). Childhood obesity: are we missing the big picture? Obesity Reviews, 9, 35-42.

Myers, L., Strikmiller, P. K., Webber, L. S., & Berenson, G. S. (1996). Physical and sedentary activity in school children grades 5-8: the Bogalusa Heart Study. Medicine & Science in Sports & Exercise, 28, 852-859.

Ogden, C. L., Flegal, K. M., Carroll, M. D., & Johnson, C. L. (2002). Prevalence and trends in overweight among US children and adolescents, 1999-2000. JAMA: The Journal Of The American Medical Association, 288(14), 1728-1732.

Pate, R. R., Freedson, P. S., Sallis, J., Taylor, W. C., Sirard, J., Trost, S. G., et al. (2002). Complicance with physical activity guidelines: prevalence in a population of children and youth. Annalz of Epidemiology, 12, 303-308.

Pretty, J., Peacock, J., Sellens, M., & Griffin, M. (2005). The mental and physical health outcomes of green exercise. International Journal of Environmental Health Research, 15(5), 319-337.

Ries, A. V., Voorhees, C. C., Gittelsohn, J. G., Roche, K. M., & Astone, N. M. (2008). Adolescents' perceptions of environmental influences on physical activity. American Journal of Health Behavior, 32(1), 26-39.

Ross, J. G., & Gilbert, G. G. (1985). The National Children and Youth Fitness Study: a summary of findings. J Physical Educ Recreat Dance, 56(Suppl 1), 45-50.

Safron, D. J., Schulenberg, J. E., & Bachman, J. G. (2001). Part-Time Work and Hurried Adolescence: The Links among Work Intensity, Social Activities, Health Behaviors, and Substance Use. Journal of Health and Social Behavior, 42, 425-449.

Saksvig, B. I., Catellier, D. J., Pfeiffer, K., Schmitz, K. H., Conway, T., Going, S., et al. (2007). Travel by Walking Before and After School and Physical Activity Among Adolescent Girls. Archives Pediatric Adolescent Medicine, 161(2), 153-158.

Sallis, J., & Glanz, K. (2006). The role of built environments in physical activity, eating, and obesity in childhood. Future of Children, 16(1), 89-108.

Sallis, J., & Owen, N. (1999). Physical Activity and Behavioral Medicine. Thousand Oaks, CA: Sage.

Sallis, J. F., Bauman, A., & Pratt, M. (1998). Environmental and policy interventions to promote physical activity. American Journal of Preventive Medicine, 15(4), 379-397.

Sallis, J. F., Kraft, K., & Linton, L. (2002). How the environment shapes physical activity: A transdisciplinary research agenda. American Journal of Preventive Medicine, 22, 208.

Sallis, J. F., & Owen, N. (1997). Ecological models. San Francisco, CA: Jossey-Base.

Schachtel, M. R. B. (2001). CitiStat and the Baltimore Neighborhood Indicators Alliance: Using Information to Improve Communication and Community. National Civic Review, 90(3), 253-266.

Thompson, A. M., Campagna, P. D., Rehman, L. A., Murphy, R. J., Rasmussen, R. L., & Ness, G. W. (2005). Physical activity and body mass index in grade 3,7, and 11 Nova Scotia students. Medicine & Science in Sports & Exercise, 37, 1902-1908.

Tolan, P. H., & Henry, D. (1996). Patterns of psychopathology among urban poor children: Comorbidity and aggression effects. Journal of Consulting and Clinical Psychology, 64(5), 1094-1099.

Treuth, M., Sherwood, N. E., & Baranowski, T. (2004). Physical activity self-report and accelerometry measures from the Girls Health Enrichment Multi-site Studies. Preventive Medicine., 38(suppl), S43-S49.

Treuth, M., Sherwood, N. E., & Butte, N. F. (2003). Validity and reliability of activity measures in African-American girls for GEMS. Medicine & Science in Sports & Exercise, 35, 532-539.

Trost, S. G., Kerr, L. M., Ward, D. S., & Pate, R. R. (2001a). Physical activity and determinants of physical activity in obese and non-obese children. International Journal of obesity and related metabolic disorders, 25, 822-829.

Trost, S. G., Kerr, L. M., Ward, D. S., & Pate, R. R. (2001b). Physical activity and determinants of physical activity in obese and non-obese children. International Journal of Obesity, 25, 822-829.

Trost, S. G., Sallis, J. F., Pate, R. R., Freedson, P. S., Taylar, W. C., & Dowda, M. (2003). Evaluating a model of parental influence on youth physical activity. American Journal of Preventive Medicine, 25(4), 277-282.

Trueth, M. S., Schmitz, K., Catellier, D. J., McMurray, R. G., Murray, D. M., Almeida, M. J., et al. (2004). Defining accelerometer thresholds for activity intensities in adolescent girls. Medicine & Science in Sports & Exercise, 36(7), 1259-1266.

USDHHS. (2000). Heathly People 2010: conference edition. Washington, D.C.: U.S. Government Printing Office.

Weist, M. D., Acosta, O. M., & Youngstrom, E. A. (2001). Predictors of violence exposure among inner-city youth. Journal Of Clinical Child Psychology, 30(2), 187-198.

Wilson, J., & Kelling, G. (1982). Broken Windows. Atlantic, 249(3), 29-38.

Fang Yan, MD., PhD., is an Assistant Professor at Department of Community and Behavioral Health Promotion, Zilber School of Public Health, University of Wisconsin at Milwaukee. Carolyn C. Voorhees, PhD, is affiliated with the University of Maryland School of Public Health, Department of Behavioral and Community Health. Guangyu Zhang, PhD, is affiliated with the University of Maryland School of Public Health, Department of Epidemiology and Biostatistics. Kenneth H. Beck, PhD, FAAHB, is affiliated with the University of Maryland School of Public Health, Department of Behavioral and Community Health. Shuo Huang, MS, is affiliated with The National Center for Smart Growth, Department of Urban Studies and Planning, University of Maryland, College Park. Hua Wei, MS, is affiliated with the Department of Geography, University of Maryland, College Park. Corresponding Author Information: Fang Yan, M.D., Ph.D., Assistant Professor, Department of Community and Behavioral Health Promotion, Zilber School of Public Health, University of Wisconsin at Milwaukee, Alumni House 327, PO BOX 413, Milwaukee, WI 53201-0413, Email: yanf@uwm.edu, Phone: 414-229-3264
Table 1 Baltimore City's Community Statistical Areas Vital
Sign indicators examined at the current study

                                 Housing and Community
                                 Development Indicators

Indicator                            Definition

Percent of residential    Number of vacant and abandoned
  properties that are     homes out of all residential
  vacant and abandoned    properties in that area that year.
  at year's end           Properties are considered vacant/
                          abandoned by Baltimore City if
                          the property is not habitable.
Median Sale Price         Selling price of a home that falls
                          in the middle of the most expensive
                          and least expensive home sale price
                          in that area
Racial Diversity Index    The probability that two people
                          picked at random will be of a
                          different race/ethnicity. This
                          number does not reflect which race
                          is predominant in an area
Economic Diversity        The probability that two households
  Index                   chosen at random will earn a
                          household income in different median
                          income range groups

                                  Workforce and Economic
                                  Development Indicator

Unemployment rate         Number of people ages 16-64 who are
                          in the labor force (looking for work),
                          but are not employed

                             Children and Family Safety and
                               Violence/ Crime indicators

Domestic Violence         Number of 911 calls to police for
  rate                    domestic violence incidents out of
                          every 1,000 people in the area
                          using Census 2000 population data
Juvenile Arrest Rate      Number of arrests of youth per 1,000
                          youths ages 10-17 in an area using
                          Census 2000 population data
Crime Rate                Number of reported Part I criminal
                          offenses per 1,000 people in the area
                          using Census 2000 population data
                          Part I offenses include murder,
                          aggravated assault, rape, attempted rape,
                          burglary, larceny, and auto theft.

                          Environment Pleasantness Indicator

Tree canopy               The concentration of tree coverage
                          in an area

Rate of dirty streets     Number of reported incidents of dirty
  and alleys              streets and alleys per 1,000 people
                          in the area using Census 2000 data

                                   Transit Indicators

Travel time to work       Percent of travel time to work more than
                          30 min (e.g., travel time to work refers
                          to the total number of minutes that it
                          usually took the person to get from home
                          to work each day during the reference week)
                          for population that does not work at home

                                                 Mean          SD

Indicator                    Source

Percent of residential    Baltimore City        5.9 (%)        7.0
  properties that are     Department of
  vacant and abandoned    Housing and
  at year's end           Community

Median Sale Price         First American        85.3 (k)       48.5
                          Real Estate

Racial Diversity Index    U.S Census            27.8 (%)       18.7

Economic Diversity        U.S Census            69.4 (%)       8.3
  Index                   2000

Unemployment rate         U.S. Census 2000,     11.6           5.9
                          provided by the
                          Maryland Department
                          of Planning-State
                          Data Center

Domestic Violence         Baltimore City        49.2           19.4
  rate                    Police Department

Juvenile Arrest Rate      Baltimore City        115.5 (per     58.4
                          Police Department     1,000 youth
                                                ages 10-17)
Crime Rate                Baltimore City        95.1           89.7
                          Police Department

Tree canopy               Ikonos satellite      17.3           15.1
                          image from Fred
                          Irani of the
                          Maryland Department
                          of Natural resources
Rate of dirty streets     CitiStat              15             9.3
  and alleys

Travel time to work       U.S Census 2000       43.3 (%)       8.76

Table 2 Characteristics of the study population (n=297)

Variables                                     Mean (SD)

Age                                          15.7 (1.21)
Health Status                                2.3 (0.94)


Gender               Boys                          40.1
                     Girls                         59.9
Ethnicity            Black                         69.0
                     White                         16.2
                     others                        14.8
Grade                9th                           30.3
                     10th                          20.9
                     11th                          12.8
                     12th                          29.0
                     Missing                        7.1
Father's education   Some high school               9.8
                     High school graduate          27.6
                     Trade school graduate          6.4
                     Some college                  14.8
                     College graduate              13.1
                     Advanced degree                7.4
                     Not sure                      20.9
Mother's education   Some high school               6.4
                     High school graduate          22.9
                     Trade school graduate          2.4
                     Some college                  27.9
                     College graduate              22.2
                     Advanced degree               11.4
                     Not sure                       6.7

Table 3 Multivariate mixed effect model (physical activity =
intercept + predictors)

Variable                                   Parameter    SE    P-value

Age                                          -0.65     0.13   <.0001
Gender             Girls                     -1.27     0.34   0.0002
Ethnicity          Black                     -1.25     0.48    0.01
                   White                      0.33     0.60    0.59
                   other                      Ref
Health Status                                -0.31     0.17    0.07
travel_long        Percent of Travel         -0.10     0.03   <0.001
                   time more than 30
                   min to Work for
                   Population That
                   Doesn't Work at
vac_aban_mean      mean of vacant and        -0.12     0.06    <0.05
                   abandoned property
                   rate 01-04
pricemean1         mean of median house       0.016    0.01    0.38
                   sale price 00-04
race_divers00      Racial Diversity           0.00     0.01    0.81
                   Index of 00
arrest_mean        mean of Juvenile          -0.01     0.01    0.11
                   Arrest Rate 00-04
mean_dirtystreet   mean of dirty street      -0.03     0.04    0.48
                   rate 02-04
unemployrate00     unemployed rate of 00      0.04     0.07    0.61
tree01             Tree Canopy--Percent       0.02     0.02    0.44
                   of area covered by
                   trees of 01
Gale Copyright: Copyright 2011 Gale, Cengage Learning. All rights reserved.