Background
Latinas are less likely to engage in the recommended levels of physical activity (PA) for health compared to women of other racial/ethnic backgrounds (CDC, 2012). The reasons for these disparities are unknown but several studies suggest that factors across various levels of the socio-ecological model influence PA (López et al., 2008; Martinez et al., 2009). To date, few studies have examined the influence of factors at both the individual- and environmental- levels on either domain-specific activities (e.g., leisure time PA) or accelerometer-assessed moderate-to-vigorous physical activity (MVPA) among Latinos. Therefore, the purpose of this study was to test a principle of the socio-ecological model by assessing the moderating influence of personal-level characteristics on neighborhood-level correlates of self-report leisure-time MVPA (LTMVPA) and accelerometer-assessed MVPA among Latina women in San Diego, CA.

Objectives
The objective of this study was to assess whether Latinas with more advantaged personal-level characteristics (higher income, education, and acculturation levels and lower body weight status) and favorable perceptions of neighborhood features (built and social environment) were more likely to meet the Physical Activity Guidelines for Americans (PA Guidelines) using LTMVPA and accelerometer-assessed MVPA than those with less-advantaged characteristics.

Methods
This study used cross-sectional baseline data obtained from 437 Latina women participating in a multilevel intervention for PA promotion in San Diego, CA, Fe en Acción [Faith in Action].  Participants were recruited from 16 churches that met the inclusion criteria and were randomized to either the PA intervention or attention-control group. After providing informed consent, participants had their anthropometric measures taken, completed a face-to-face interview, and wore an accelerometer for at least 5 complete days (i.e., 10 or more valid hours/day). Subscales from the Neighborhood Walkability Scale were used to examine five neighborhood features (e.g., safety from crime, safety from traffic, and aesthetics). Mixed effects models accounting for church clustering effects, and adjusted for confounders, examined the associations of neighborhood features with whether or not participants met the PA Guidelines using LTMVPA and accelerometer-assessed MVPA. To test for moderation by income, education, acculturation, and body weight status on the associations of neighborhood features with LTMVPA and accelerometer-assessed MVPA, interaction terms were created between each personal-level characteristic and neighborhood feature.

Results
Latinas with higher levels of perceived safety from crime were more likely to meet the PA Guidelines using LTMVPA than those with lower levels of perceived safety (ORadj = 1.27, 95% CI: 1.03-1.58). None of the neighborhood features were significantly related to meeting the PA Guidelines using accelerometer-assessed MVPA. Tests for moderation showed that the associations between neighborhood features and meeting the PA Guidelines varied by personal-level characteristics. Specifically, Latinas were significantly more likely to meet the PA Guidelines using LTMVPA if they had at least a high school education and higher levels of perceived safety from crime compared to those with lower education (ORadj=1.69, 95% CI: 1.16-2.45); had more assimilated acculturation levels and higher perceived safety from crime compared to less acculturated women (ORadj=1.93, 95% CI: 1.19-3.15); had a normal body weight and favorable neighborhood aesthetics compared to overweight/obese women (ORadj=2.73, 95% CI: 1.37-5.42); and had a monthly household income of $2000 or higher and better sidewalk maintenance compared to lower income women (ORadj = 2.05, 95% CI: 1.05-4.01). In addition, Latinas with more assimilated acculturation levels and higher levels of perceived safety from crime were twice as likely to meet the PA Guidelines using accelerometer-assessed MVPA as less acculturated Latinas (ORadj=2.11, 95% CI: 1.16-3.82).

Conclusions
Higher perceived safety from crime was the only significant correlate of meeting PA Guidelines using LTMVPA among Latinas. Of 20 interactions tested for LTMVPA, four personal-level characteristics were found to moderate neighborhood features: income, education, acculturation, and body weight status. None of the neighborhood features were significantly associated with meeting PA Guidelines using accelerometer-assessed MVPA; however, of the 20 interactions tested for accelerometer-assessed MVPA, acculturation was found to moderate perceived safety from crime. Our findings provide support for the socio-ecological model with significant interactions found across levels of influence of PA among Latinas.

Implications
Improvements in the neighborhood environment may only increase PA among subgroups of Latinas, such as women with higher income and education, more assimilated acculturation levels, or normal body weight. Subgroup differences among Latinas should be considered when developing interventions and policies targeting improvements of the neighborhood environment to promote PA. In addition, those with lower education and less acculturated to the US may need interventions beyond environmental changes.

References

1. Centers for Disease Control and Prevention (CDC). Behavioral Risk Factor Surveillance System Survey Data. 2012.
2. López IA, Bryant CA, McDermott RJ. Influences on physical activity participation among Latinas: an ecological perspective. American Journal of Health Behavior. 2008;32(6):627-639.
3. Martinez SM, Arredondo EM, Perez G, Baquero B. Individual, social, and environmental barriers to and facilitators of physical activity among Latinas living in San Diego County: Focus group results. Family & Community Health. 2009;32(1):22-33.

Support / Funding Source
This study was supported by a diversity supplement from the National Cancer Institute of the National Institutes of Health (3R01CA138894-04S1).