A large body of research shows that social determinants of health have significant impact on the health of Canadians and Americans. Yet, very few studies have directly compared the extent to which social factors are associated with health in the two countries, in large part due to the historical lack of comparable cross-national data. This study examines differences in the effect of a wide-range of social determinants on self-rated health across the two populations using data explicitly designed to facilitate comparative health researchdJoint Canada/United States Survey of Health. The results show that:
- sociodemographic and socioeconomic factors have substantial effects on health in each country, though the size of the effects tends to differdgender, nativity, and race are stronger predictors of health among Americans while the effects of age and marital status on health are much larger in Canada; the income gradient in health is steeper in Canada whereas the education gradient is steeper in the U.S.;
- Socioeconomic status (SES) mediates or links sociodemographic variables with health in both countriesdthe observed associations between gender, race, age, and marital status and health are considerably weakened after adjusting for SES;
- psychosocial, behavioural risk and health care access factors are very strong determinants of health in each country, however being severely/morbidly obese, a smoker, or having low life satisfaction has a stronger negative effect on the health of Americans, while being physically inactive or having unmet health care needs has a stronger effect among Canadians;
- risk and health care access factors together play a relatively minor role in linking social structural factors to health. Overall, the findings demonstrate the importance of social determinants of health in both countries, and that some determinants matter more in one country relative to the other.
As wealthy developed nations, Canada and the United States share many similarities in history, culture, and living standards and styles. The two countries also differ in other important ways, namely in the funding, organization, and delivery of health care and other social welfare programs, distribution of income, and social inequities, which likely have implication for health determinants within and between the two countries (Evans & Roos, 1999; Navarro et al., 2006; Siddiqi & Hertzman, 2007; Siddiqi & Nguyen, 2010). It is within this context that this study examines the effects of social determinants of health across Canada and the United States.
Stronks, van de Mheen, Looman, & Mackenbach, 1996), is not overly complex and lends itself to empirical analysis. Several studies in the U.S. have demonstrated support for the social determinants of health model. Using data from two national health and well-being surveys, Ross and Wu (1995) show that after controlling for sociodemographic factors (age, sex, race, and marital status) education was positively associated with self-rated health and level of physical functioning through work and economic conditions, psychosocial resources, and lifestyle.
Higher education led to more full-time employment and higher income, which in turn enhanced psychosocial resources and healthy lifestyle behaviours, and consequently better health. These findings are consistent with similar studies in the U.S. (e.g., House et al., 1990, 1994; Lantz et al., 1998; Ross & Bird, 1994). Research on social determinants of health in Canada has also revealed that social and economic factors interact with and shape psychosocial and behavioural factors, and, ultimately, health status (e.g., Birch, Eyles, & Jerrett, 2000; Denton, Prus, & Walters, 2004; Denton & Walters, 1999; Kosteniuk & Dickinson, 2003).
Cross-country comparisons of social determinants of health between Canada and the United States are also of interest given the social, economic, and cultural values shared by the countries yet the vastly different social welfare policies and programs in health care, income and employment security, and so on in each country. Intercountry research provides a unique opportunity to examine how such differences shape social determinants of health in the two nations. A few studies, by and large facilitated by the Joint Canada/ United States Survey of Healthdthe first survey of its kinddhave directly compared correlates of health across the two populations (Eng & Feeny, 2007; Huguet, Kaplan, & Feeny, 2008; Lasser, Himmelstein, & Woolhandler, 2006; McGrail, van Doorslaer, Ross, & Sanmartin, 2009; Sanmartin et al., 2006; Siddiqi & Nguyen, 2010; Siddiqi, Zuberi, & Nguyen, 2009).
Still, no research to date has explicitly compared a social determinants of health model across the countries. the current study examines the size and pattern of effects of social determinants on health in and across Canada and the U.S. using data from the Joint Canada/United States Survey of Health. It is expected that health is influenced by similar social factors in each country; however, the impact of the factors is likely different given the considerable differences in the social context and welfare policies of the two countries. U.S.-Canada comparative research on health outcomes has often encountered problems because of the lack of comparable data and/or sample design.
The Joint Canada/United States Survey of Health (JCUSH) was conducted by Statistics Canada and the National Center for Health Statistics between November 2002 and June 2003 to overcome these problems. Given the use of a single survey and a standard methodology across countries, the JCUSH provides a unique opportunity to directly compare social determinants of health across Canada and the U.S. Based on a stratified multi-stage probability sampling design, the JCUSH collected information through telephone interviews on health and illness, use of health services, correlates of health, and demographic and economic characteristics of individuals aged 18 years or older living in private residences in the ten Canadian provinces and the 50 U.S. states and the District of Columbia. Identical questions were asked to all sample respondents, except for those on race and health insurance given key differences between the countries on these variables (Statistics Canada & United States National Center for Health Statistics, 2004).
The JCUSH sample contains 3505 Canadians and 5183 Americans. Response rates were 66 percent in Canada and 50 percent in the United States. Despite the relatively low response, two recent studies showed that key characteristics, such as the age distribution, of the U.S. and Canadian samples in the JCUSH were quite similar to those in leading national health surveys in each country (Feeny, Kaplan, Huguet, & McFarland, 2010; Kaplan, Huguet, Feeny, & McFarland, 2010).
Health was defined using the concept of self-rated health (SRH). SRH indicates a respondent’s health status based on his or her own judgement. SRH is a very useful indicator of the overall health and well-being of individuals and populations. SRH is operationalized in the JCUSH as follows: “In general, would you say your health is: excellent, very good, good, fair, or poor?” It was grouped into three categories for purposes of analysis, “poor” (poor or fair), “good” (good), and “excellent” (very good or excellent), as has been done in many other studies of social determinants of self-rated health (e.g., Cott, Gignac, & Badley, 1999; Lantz et al., 2001; Manderbacka, Lahelma & Martikainen, 1998).
As a measure of health, the reliability and validity of self-rated health have been well-established (Idler & Benyamini, 1997). SRH provides a valid assessment of overall health (Idler, Russell, & Davis, 1992), and is a strong predictor of mortality (Mossey & Shapiro, 1982; Smith, Shelley, & Dennerstein, 1994; van Doorslaer & Gerdtham, 2003), disability (Mansson & Rastam, 2001), functional limitation (Idler & Benyamini, 1997), health-related behaviour (Cott et al., 1999; Manderbacka et al., 1998), and health care utilization (Pinquart, 2001). Banks, Marmot, Oldfield, and Smith (2006) find that self-reported measures of health are almost identical to those with biological measures such as physical and laboratory examinations.
