Introduction Intervention in vulnerable groups might produce risk factors modification. This study is the baseline description of a health impact study of the implementation of FCs, a community citywide intervention.

Methods A natural step-wedge design compared areas where the intervention were implemented. FCs are community centers located in vulnerable areas of the city, built for fitness classes, staffed by municipality instructors. Individual indicators came from a multistage (census tracts-CT, households, and residents) survey, stratified by distance street buffers created around the planned point interventions, intersected with the CT, giving probabilistic ranging from 0.1 to 0.012 as the distance increase from the site intervention. Demographics, social determinants, health and lifestyles (including Physical activity-PA: >150 min/week), social support for PA (PASS) were collected. PA and co-variables were compared across one area that received earlier intervention with all others surveyed, using multiple Poisson regression models.

Results Of the 4048 adults, 54.1% were female; mean age 41.2 years; 16.1% reported PA. Comparing estimates across areas with and without FC, age, female gender, smoking and good self-perception of health were inversely related to PA while PASS, schooling and income were directly associated to PA in areas without FC. For those 17.6% living in the FC catchment area the only associated factor with PA was report of eating fruits 5 times/week; all other factors were no longer associated with PA.

Conclusion Tailored to address health disparities for chronic diseases, interventions in high risk urban environmental might narrow health inequality for some indicators of physical activity.