Population Health in Canada: A Working Paper

Appendix B: Population Health Frameworks

Identifying correlates of health status is only a first step. The various "causes" should then be placed in a framework that makes a coherent picture of the whole. In the next section, we describe major frameworks and models available in the literature for thinking about relationships among most of the correlates of health identified in previous sections and two schemes for researching these relationships.

The CIAR Framework

Evans and Stoddart (1990; 1994) provide a clear model for describing disease, health and well-being and their determinants. In this model, disease influences health and function which in turn influences well-being. Of these three, the health-care system is related most naturally to disease. Factored into the picture are (economic) prosperity (influenced, in part, by the size of the health-care system and the levels of healthy functioning in the populace), individual behaviour and biology, genetic endowment and the physical and social environments (the latter two influenced by prosperity). This model can be used to assess both positive and negative effects of the health-care system. It can incorporate lifestyles of individuals, social-psychological concepts such as stress, self-esteem and perception of control, and social concepts such as social relationships and hierarchical status.

Frank (1995:233-237) has identified five major foci of the CIAR perspective. First, this framework recognizes the limits to medical care. The health care system is not the primary determinant of population health status and, thus, money spent on health care is not necessarily the best investment in a society's health and productivity. Second, the framework recognizes macro socioeconomic and cultural factors that influence a population's health. For example, an unequal distribution of income could be related to health, which in turn could be driven, in some nations, by dysfunctional economies and/or totalitarian, rigid social systems. Third, the framework identifies micro-level factors in individuals' immediate social environments that influence health, such as social position, socioeconomic status, stress and/or coping skills. Fourth, they recognize the importance of early childhood development, arguing that how children are cared for at an early age influences basic capacities to function and, thus, health throughout the life-cycle. Finally, they recognize new biologic pathways for socially mediated health effects. If one's position in the social hierarchy is related to many different illnesses, for example, then perhaps there are generic biologic causal mechanisms or pathways that translate social environmental inputs into adverse body responses. Relationships among these five foci are being investigated. For example, could adverse social and economic environments influence early childhood experiences?

The Wilkinson Framework

More recently Wilkinson (1996) has proposed a framework for understanding health and its determinants that incorporates many of these foci. In summarizing this framework we have drawn exclusively from Hayes and Dunn (1998).

Wilkinson focuses particular attention to the corrosive effect of social inequality upon the health of populations, through multiple pathways.

"Impact upon health is not entirely or exclusively a matter of material inequalities - it does not reduce to income or economic circumstance, but material distribution has an overwhelmingly important impact ... Economic opportunities play a crucial role in shaping the meanings attached to position in the social hierarchy directly (through the exercise of choices made possible by money) and indirectly (through conferring social status). The exchange value of money also allows the individual direct access to material goods required for sustenance (food, shelter, clothing, etc.) that are distributed across society. Ultimately, these pathways operate through the level of culture, for it is here that experiences are shaped .. [H]ow resources are shared, one's entitlements and expectations, rituals and beliefs .. all operate at the level of social construction, and through the distributive mechanisms of power and knowledge, materiality and meaning (Hayes and Dunn, 1998: 29).

"[I]dentity, power and control, ability to participate meaningfully in the routines of daily life, a sense of hope about the future, perceived self-worth relative to others, friendships, communication skills, and various other aspects of individual behaviour and perception are thought to be the ultimate influences upon health, as mediated by or conditioned through individual biology (Hayes and Dunn, 1998: 29)."

Thus, for Wilkinson (1996), social structure, and the inequities therein, is a particularly pertinent concern for population health research. In this spirit, relying upon the summary provided by Hayes and Dunn (1998: 26-28), Tarlov (1996) speculates on the relationship between individual experience - psychosocial and behavioural - and biological responses. Fongay (1996) explores how some psychosocial attributes - identity, coping skills and social competence, for example - are fostered in the relationships between children and parents. Blane et al. (1996) look at how the educational system contributes to social inequality. Bartley et al. (1996) discuss how the social-structural effects of employment, its availability and the conditions of work and its security have demonstrable effects upon health. Brunner (1996) explores how job stress and its distribution in the workplace affects heart disease. Wilkinson (1996) argues that social cohesion, including trust and participation in networks, may have an important relationship with health.

This ambitious framework includes both material and non-material inequities which link economic, social and political dimensions and their relationships with social hierarchies. It also links material circumstances with psychosocial perceptions and meanings and in turns connects these through physiological and biological pathways to health.

The attempt at comprehensiveness by Wilkinson and his colleagues means that the relationships between and among identified elements are numerous, and additional research investigating relationships among pieces of the whole is required. Some of the work on the social determinants of health will fill in parts of the picture, but there are gaps, and, of course, the generalizability of the framework requires testing. It is possible that differences among nations with respect to welfare state policies and access to health-care services, for example, will affect the relationships among factors within nations. As we noted above, for example, social capital and/or income inequality may not be strongly related to health in Canada as has been found elsewhere.

The comprehensiveness of the Wilkinson framework leads one to suspect that almost everything is interconnected. This does, however, support the King and Williams (1995) approach to investigating the relationship between race and health, for example. They conclude that race influences health since race is a proxy for biology, cultural, socioeconomic and political factors, and for racism, which then influence health practices, stress, environmental stress, psychosocial resources and medical care, all of which affect biological processes and hence, health. The picture is becoming complicated, which leads us to frameworks for investigation into relationships among concepts.

The Hertzman, Frank and Evans Framework for Investigating Heterogeneities in Health

Hertzman et al. (1994) note that population health research aims to explain heterogeneity in health status among population groups by finding related characteristics of the groups. They have created a framework with three dimensions. First, because of the relevance of childhood influences upon later health states, and because of differential effects of some circumstances upon different age groups, they argue that population health research should include life cycle stage. They support a four-part partition that focuses attention specifically upon the first year of life, the misadventure years (1-44 years), the chronic disease years (45-74) and, finally, the senescence years (75 plus).

Second, they advocate clear definitions of desired populations of interest, which could include, for example, socioeconomic status of population members, ethnicity/migration, geographic locality and gender. Finally, they delineate some known or suspected sources of heterogeneity. Of particular interest are their observations about causality and method. They warn of reverse causality, noting that one should always be aware that health could influence one's position in the social hierarchy, as well as the other way around. Also, some populations or sub-populations may have differential susceptibility to determinants. Sensitivity to issues of time is important, and researchers could differentiate between elapsed time (some causes have delayed effects), biological time (some early experiences manifest at a later stage in the life cycle), cumulative time (sometimes toxins accumulate to a critical mass at which point they influence health, for example) and historical time (sometimes latency effects are related to causes specific to a given period but not in operation today).

The Hamilton and Bhatti "Population Health Promotion Model"

Recently, efforts have been made to set up population health in opposition to health promotion (Frank, 1995). Evans and Stoddart (1994) have faulted the "lifestyle" movement for its emphasis on individual health behaviours and resultant "victim blaming;" a theme debated within health promotion circles as well. They also note that Canadians continue to experience significant health problems and health inequities (both as individuals and as a society) despite an enormous growth in health care spending and concomitant expansion of medical programs over the past several decades.

The factors identified in a population health approach are not substantially different from the prerequisites for health identified in the Ottawa Charter (1986), however. The Ottawa Charter recognized the fundamental conditions and resources for health to be peace, shelter, education, food, income, a stable ecosystem, sustainable resources, social justice and equity. Health requires a secure foundation in these prerequisites. Self-responsibility for health must be distinguished from self-reliance. Individuals, families and communities cannot be blamed for failing to rely on resources (economic, social, psychological) they do not possess or have reasonable access to.

Downplaying lifestyle as the pivotal point of the causal chain between programs or policies and health impact, the Epp Framework later added the words "for all" to the word health to lay the emphasis of his direction on attaining "equity in health". The causal links to health were then arrayed in three tiers, one of "health challenges" (reducing inequities, increasing prevention, enhancing coping), one of "health promotion mechanisms" (self-care, mutual aid, healthy environments) and one of implementation strategies (fostering public participation, strengthening community health services, and coordinating healthy public policy).

The Ottawa Charter on Health Promotion engendered the most widely adopted definition of health promotion as a statement of its goal of enabling people to gain control over and to improve their health. Its emphasis on empowering people to have greater control over their health shifted the spotlight away from health care services and toward other determinants of health in the environment and in living conditions and lifestyles. The Charter also helped position health in this implicit causal chain not as an end unto itself but as "a resource for living" by which it referred to other qualities of life for which health was to be seen as a determinant itself.

The health promotion perspectives described above have been combined with the population health perspective adopted by the Federal, Provincial, and Territorial Health Ministers to outline strategies for action on the full range of health determinants at all levels of populations from individual to societal. Hamilton and Bhatti formulated "An Integrated Model of Population Health and Health Promotion" in which they combined the foregoing formulations to suggest a framework that could guide actions to improve health. They present a three-dimensional policy and practice cube that suggests an intersection for each of the determinants of health named above with each of the levels of population from individual to society, and each of those intersections with the five strategies of the Ottawa Charter for Health Promotion. This model makes more explicit the ecological perspective that has been a foundation of public health and health promotion from their earliest articulations, but has been operationalized only partially in most health promotion projects and policies undertaken by federal and provincial agencies in Canada. It also links back to the health fields of the Lalonde Report.

Of further interest is their discussion on the need for evidence-based decision-making and the nature of evidence. They delineate three kinds of evidence:

"research studies that have systematically linked health issues, the underlying factors, the interventions and their impact, both intended and unintended; experiential knowledge that has been gained through practice and synthesized in ways that can guide practice and policy-making; and evaluation studies (formative and summative of policies, programs and projects" (Hamilton and Bhatti, 1996: 8).

This adds an important dimension -- the need for evaluation of interventions - along with suggestion that experiential evidence has a place within the population health movement. Finally, Hamilton and Bhatti note the role of values, beliefs and assumptions as part of the foundation underlying all policy and program decisions.

Table Of Contents