br Conflicts of interest br Acknowledgements

Conflicts of interest
Acknowledgements The authors are grateful for the fieldwork data collection undertaken by student and assistants from the Faculties of Environmental Design (Karina Lamy), Medicine (Lindsay Burton, Ann Toohey, Monica Viapiana), Science (Stefano Liccioli and Sarah Gingerick), and Veterinary Medicine (Barbara Behiels) at the University of Calgary. This project was funded by a Canadian Institutes of Health Research (CIHR), Institute of Population and Public Health Population Health Intervention Research Strategic Competition Grant (GIR-112745). Melanie Rock was supported by an Alberta Innovates Population Health Investigator Award and CIHR Health Research New Investigator Award. Gavin McCormack is supported by a CIHR New Investigator Award. The collaborative support from the Calgary Humane Society, the Calgary Parks Foundation, the Federation of Calgary Communities, and the City of Calgary Animal and Bylaw Services and City of Calgary Parks is also acknowledged.
Introduction Individuals with prior military experience are a unique ruthenium red Supplier and may provide insights regarding the effects of social stressors on vulnerability and resilience (King, King, Foy, Keane, & Fairbank, 1999). Competing hypotheses exist as to whether this population is psychologically and physically more vulnerable or resilient to current stressors (Aldwin & Stokols, 1988; Elder & Clipp, 1989). Many individuals may join the military to escape dysfunctional and chaotic situations, suggesting that despite exposure to early adversity, this population may have an increased capacity for resilience (Blosnich, Dichter, Cerulli, Batten, & Bossarte, 2014; Katon et al., 2015). Nevertheless, compared with those without military service, those with such a history consistently report poorer physical and mental health (Hoerster et al., 2012; Lehavot, Hoerster, Nelson, Jakupcak, & Simpson, 2012). Thus, it is possible that early adverse experiences, coupled with military-related trauma may increase vulnerability to long-term social stressors (Aldwin, Levenson, & Spiro, 1994). Yet, it remains unclear whether social stressors may differentially contribute to observed health disparities among those with a history of military service. Food and housing insecurity are two types of social stressors that can have a profound impact on health (Braveman, Egerter, & Williams, 2011; Commision on Social Determinants of Health (CSDH), 2008; Office of Disease Prevention and Health Promotion, 2015). In 2013, approximately 14% of U.S. households were food insecure, defined as having restricted access to safe and healthy foods (Coleman-Jensen, Gergory, & Singh, 2014). Adverse living conditions, such as crowded living and high housing cost to income ratio, represent housing insecurity (Johnson & Meckstroth, 1998); In 2014, 40% of adults reported that mortgage or rent were significant sources of stress (American Psychological Association, 2015). A growing body of evidence demonstrates the link between food and housing insecurity and poor health behaviors and healthcare access. For example, 43% of US adults reported that Crossing-over have eaten too much or have eaten unhealthy foods because of stress (American Psychological Association, 2015). Twenty percent of US adults reported having or considered skipping an annual doctor visit due to financial concerns (American Psychological Association, 2015). Food and housing insecurity may also contribute to unhealthy weight status among youth and adults (Casey et al., 2006; Cutts et al., 2011; Pan, Sherry, Njai, & Blanck, 2012; Rose & Bodor, 2006), and are associated with increased emergency room use, hospitalization, and poor mental health (Heflin, Siefert, & Williams, 2005; Kushel, Gupta, Gee, & Haas, 2006; Whitaker, Phillips, & Orzol, 2006).
Methods Data were derived from respondents in nine states (Hawaii, Illinois, Massachusetts, Michigan, Nebraska, New Jersey, North Carolina, Oklahoma, and Wyoming) administering the Social Context module as part of the 2011 or 2012 Behavioral Risk Factor Surveillance System (BRFSS). BRFSS is an annual cross-sectional telephone survey coordinated between state health departments and the Centers for Disease Control and Prevention (Centers for Disease Control and Prevention, 2013c). BRFSS is administered via random-digit dialing of landline and cell phones of non-institutionalized adults aged≥18 years using a cluster stratified random sampling design. Details on methodology used in the collection of BRFSS data have been previously documented (Centers for Disease Control and Prevention, 2013c). State-level BRFSS datasets were appended after accounting for each state׳s independent sampling methodology. Response rates for the four states in 2011 ranged from 43.1% to 60.9% and for the seven states in 2012 ranged from 38.0% to 47.8% (Centers for Disease Control and Prevention, 2013a, 2013b). The analytic sample included 81,405 (71.4%) of 113,983 respondents to the Social Context module, after excluding those with missing (i.e., ‘not applicable,’ ‘refused,’ and ‘don’t know/not sure’ responses) responses to questions about military service status (n=184), food insecurity (n=22,009), housing insecurity (n=29,929), mental health (n=1937), and physical health (n=2445). To assess the potential non-respondent bias to the independent variables, we compared those with and without missing values on the independent variables.