Short Report Residential segregation and exercise among a national sample of Hispanic adults Jenelle Mellerson a , Hope Landrine a,b,n , Yongping Hao c , Irma Corral a , Luhua Zhao a , Dexter L. Cooper a a Behavioral Research Center, American Cancer Society, USA b Multicultural Health Behavior Research, Behavioral Research Center, American Cancer Society, 250 Williams Street, Atlanta, GA 30303, USA c Centers for Public Health Research and Evaluation, Battelle, USA article info Article history: Received 6 August 2009 Received in revised form 21 November 2009 Accepted 21 December 2009 Keywords: Hispanics Exercise Residential segregation Neighborhoods abstract We examined the role of residential segregation in exercise among American Hispanics for the first time. Data on the 8785 Hispanic adults in the 2000 Behavioral Risk Factor Surveillance System (BRFSS) were linked to 2000 census data on the segregation of metropolitan statistical areas (MSAs). Multi-level modeling revealed that after adjusting for individual-level variables, the odds of exercise among Hispanics residing in high-segregated MSAs were 18% lower than those residing in low-segregated MSAs. This suggests that segregation contributes to lack of exercise among Hispanics; this effect might be mediated by the relative lack of recreational resources in segregated-Hispanic neighborhoods. & 2010 Elsevier Ltd. All rights reserved. 1. Introduction Lack of physical activity is a problem for all Americans, and for Hispanic Americans in particular (Centers for Disease Control, 2005a; Centers for Disease Control, 2005b). Levels of physical activity among Hispanics are significantly lower than those of Whites and of Blacks as well (Ahmed, Smith, Flores et al., 2005; Centers for Disease Control, 2005b). Gender (Ham, Yore, Kruger et al., 2007; Marquez and McAuley, 2006), acculturation (Ham et al., 2007; Marquez and McAuley, 2006), specific Hispanic ethnic group (Ham et al., 2007; Neighbors et al., 2008), and socio- economic status (SES; Neighbors et al., 2008) contribute to Hispanic disparities in physical activity but do not fully explain them (Ham et al., 2007). Residential segregation is an additional possible contributor that has not been examined. Segregated Hispanic neighborhoods are significantly more likely to lack recreational facilities than White neighborhoods of comparable SES (Moore et al., 2008; Powell et al., 2006). For example, Moore et al. (2008) examined the availability of recreational facilities in segregated White, Black, and Hispanic census tracts (CTS) in North Carolina, New York, and Maryland. Moore found that Black CTS (across CT SES) were 3 times more likely, and Hispanic CTS 8.6 times more likely than White CTS to lack recreational facilities; 82% of Hispanic (vs. 70% of Black and 38% of White) CTS have no recreational facilities. Given the strong role of access to such facilities in adult physical activity (Diez et al., 2007), low levels of exercise among Hispanics might in part reflect low access to recreational resources in segregated Hispanic neighborhoods (Moore et al., 2008). To begin to examine this possibility, we compared exercise-levels among high vs. low segregated American Hispanics for the first time, by using data from the 2000 Behavioral Risk Factor Surveillance System (BRFSS; CDC, 2000). 2. Method 2.1. Participants and procedure The sample consisted of N = 8785, self-identified Hispanic American adults (3608 men, 5177 women, ages 18–99, Mean age=38.9 years, SD=14.8) in the 2000 BRFSS, a nationwide, random digit-dial telephone health survey conducted by the Centers for Disease Control. 2.2. Measures and data-analytic strategy BRFSS data on exercise (any exercise in the past month? Yes/ No) were linked to 2000 census data for all metropolitan statistical areas (MSAs) in the BRFSS. MSA segregation was assessed by the Dissimilarity Index (D), a measure of the distribution of Hispanics vs. Whites across the neighborhoods of an MSA (Johnston et al., 2007). D ranges from 0 (totally integrated) to 100 (a totally segregated MSA in which Hispanics ARTICLE IN PRESS Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/healthplace Health & Place 1353-8292/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.healthplace.2009.12.013 n Corresponding author at: Multicultural Health Behavior Research, Behavioral Research Center, American Cancer Society, 250 Williams Street, Atlanta, GA 30303, USA. Tel.: + 1 404 329 4425; fax: + 1 404 929 6832. E-mail address: Hope.Landrine@cancer.org (H. Landrine). Health & Place 16 (2010) 613–615