Public Health
Grand Rounds

University of Washington
School of Public Health & Community Medicine

 
 

Another Tale From the Frozen North

Tuesday, Otober 6, 1998
3:00 p.m., T-739 HSC

Prof. Clyde Hertzman, MD
University of British Columbia and the
Canadian Institute of Advanced Research

Professor Hertzman discussed the "Evolution of the Concept of 'Population Health' in Canada and its Impact on Public Policy."

RESPONSE PANEL:


Introduction:

+ Presentation on foundations of what makes some populations healthier than others

+ 10 established foundations on his list

Ten Foundations:

1. long life expectancies this century didn't come from medicine (McKeown) (individually based health care services)

+ effective interventions are after all effective
- but scaling them up doesn't show a role


2. plumbing diagram

- Bob Evans model, (probably in Why are some people....bk)
- contextual determinants of health, what are factors that operate outside of health care box?
- question of what produces health within our society tends to be left out


3. wealth up to certain point determines health (1900, 1930, 1960, 1990 curves)

- increasing income & life expenditure is monotonic at lower levels
- amongst rich countries, further increases in GDP/capita do not correlate with increasing life expectancy (or quality adjusted life measures)
- what is going on in Canada & US


4. Wilkinson and income inequality/life expectancy

- equitability of income distribution, monotonic relationship
- trends also were correlated
+ debates over validity of LIS data + grab-bag of info about how income distributed,
- Canada has officially collected statistics, Germany has random surveys - Kaplan's work an enormous tie-breaker, because used comparable data
- don't just look at poverty per se, but measures that go up into middle class
+ Kaplan Wolfson extended analysis into Canada, with provinces clustered at good end - 24-26% range for Canadian provinces
- more decentralization in Canadian provinces than US to accomodate Quebec


5. Market systems in most countries tending towards increasing inequality

+ Slide of changes in market & disposable income
- market inequality greater in US
+ Wolfson Ross slide Disposable Income - Finland, Canada & W. Germany (to 1990) have tendencies to compensate those lower down
- redistributive task of social programs becoming greater, political pressures to allow disposable income to look like market picture - tendency of market is to push in oppositve question of what makes a population healthy


6. Income inequality correlated with dense matrix of other factors

- Kawachi & social trust
+ slide of trust in strangers, La Port, Shleifer 1996 - 40 country study,
- efficiency of judiciary
- quality of bureaucracy
- level of civic participation
- quality of physical infrasructure
- infant mortality level
- adequacy of educational system - income inequality represents dense network of factors that deal with quality of day to day lives we lead in society, existing at level of se environment and civic society environment


7. How does psychosocial and socioeconomic environment influence health status?

+ princple agent is through these gradient in health status - Socio-economic gradient, monotonic gradient, from lowest to highest, whether divide society by income level, status & presteige of occupation, or by status and level of education attained
- in all of wealthy countries of the world, if use either of these 3 or some complex measure of social class that combines them, see monotonic gradient + not a question of poorest, or those with least education being different than everybody else
+ Canada, divide by quartiles of education, & healthy active life expectancy after age 30 - montonic increase from lowest to highest educated
- 5.3 years more life expectancy
- Blishen Roberts Scale of occupational presteige in Canada - affects middle class, not a welfare discussion, has much broader effect
- gradient cuts across almost all disease processes
+ gradient replicates self on new conditions as they present in society - turn of century with gradient in infectious diseases, then heart disease enters as disease of affluent, then settles into poor (like HIV) + ubiquitous and must be explained somehow, more biologically fundamental than risk factors for individual diseases - will have to do with vulernability to and resistance to disease a a much deeper level + questions of reverse causality and differential mobility - don't explain much of gradient (<15%) + countries with steeper income inequalities have steeper ses gradients for disease - US is at top end, followed closely behind with Britain, Sweden low down
- despite fact that gradient is found everywhere, their steepness varies quite a bit, varies with correlates of income inequality


8. hypothesis of causality of gradient

+ must somehow emerge from biological embedding - differential experiences and quality of environments from time of conception onward somehow embed themselves in human biology
- begins with development of cns and sculpting of the brain & differential exposure to appropriate expreriences during brain development
- then embed themselves through connections between cns and various body systems, in traditional pathological predispositions over time - begins with differential access to opportunities for cognitive and behavioral development, gradually over time sinks into more traditional biological outcomes
+ critical periods in development of neurobiolgical phenomenon that relate to readiness for school + differential milestones, increasingly correlated with scultping of brain: at time of birth connections of cells in cns is chaotic, but under weight of visual, tactile, auditory stimulus, some connections reinforced, and others die away, so brain "sculps" and more appropriate and full-bodied the experiences are that people are exposed to early in life the richer the network of connections will be that are maintained over time
- when interract that notion with critical and sensitive periods of brain development in first few years of life, by time kids begin school, already se gradients have established themselves in terms of cognitive and behavioral development
+ National longitudinal development study in Canada on receptive language for kids 4-5 years of age with mother's educational status, already a 3 fold gradient already emerged at that point
- early on in school, if look at ses gradients with income, see relationship with grade repetition, problems with social relationships, behavioral and emotional problems of one sort or another
- gradients that emerge later on in health status emerge very early on, as gradients in terms of cognitive and behavioral development


9. biological embedding

+ level of international comparisons + Slide of math scores (age12) and ses gradient (by father's occuption), - flat for Japan (high average score flat gradient)
- NZ & US have steep gradients, lowest average scores
- Canada, France in moderate gradients, mid-range scores
- so see reflection of gradients that emerge with health in mid-life with gradients that emerge in cognitive and behavioral development early on in life + Whitehall study, see biological embedding + slide of BP decline after work is greater in those higher up the ses ladder - during the day, BP goes up, for high & low grades of civil servants, but it is low grade civil servants whose declines are less after work
- Marmott and Thornell, Int. J. Health Services 18, 4, 1988 + differences in metabolism for glucose metabolism, waist-hip ratio, fibrinogen recruitement, triglyceride metabolism - correlate strongly with se gradiets in mortality there
- when look at 4 fold difference in CV mortality, most of gradient left unexplained by usual risk factors + what starts out as gradient in cognitive and behavioral factors turns into a gradient of biomedical markers and then into mortality and morbidity outcomes traditionally understood - substrate of that is question of biological embedding


10. population approach is trying to think of bulls eye


 

- natural socioeconomic enviroment, level of wealth per capita and distribution of resources and opportunity
+ civic society (neighors)

- the way in which networks of norms and trust play themselves out, neighborhood character and cohesion, insitutional responsiveness of one sort or another, way in which day to day life can play out - intimate social network (whether people are strongly connected, or isolated)
+ interacting with that is individual life course, and whether or not there is an individual iteration on a day to day basis between the conditions here and the developing individual to somehow produce or undermine health status - population health perspective is simultaneous analysis of all these factors in one way or another


Research Issues:
+ What are research issues that emerge?

+ flat gradients and higher means - hypothesis is that flat gradients are good, those countries with flatter gradients, do they or do they not have higher averages of health status? + cognitive development - the flatter the gradient of math achievement, the higher the mean level in society
- if reduce ses inequalities get benefits for whole population without creating problems for those more priviledged + dominant ideology in our society is that we have to have an opportunity society + people who are most talented and able to do things ought to be given a free rein to take whatever benefits they can - educatte kids privately
- life in gated communities
- purchase better health care, they should be able to ...
- keep all their income, they should, we shouldn't have a progressive income tax system - if we want to do something that improves the life of the people at the other end of the spectrum, we can only do so, by limiting the life chances of the people at the upper end
- evidence seems to show that is not true, societies that find ways to arrange things more equitable seem to be able to provide large benefits at lower ends of se spectrum, without necessarily producing problems, or taking away achievement at upper end (math achievement)
+ UK vs Sweden, men 30-64 side by side comparison - reduced mortality at low end in Sweden, cf UK, but also at upper end + have study on why north america has had more economic growth over last 200 years, than S. America - Ingermen, Sokolov in Calif. has to do with same principles as these
- conflicting evidence here + how factors play themselves out over the life cycle? + current time environment &question of life cycle (longitudinal variables) - large literature, David Barker in Britain, all se effects are latent effects, by and large, if understand what is going on in gestation and first few years of life, that is all you need to know, it all happens then, after you are set on lifelong trajectories + materialist view: all has to do with current time - people's psychosocial working conditions are strong predictors of mortality over the following 10 years of their lives + Steve Suomi follows monkeys, says can i + 2 groups of monkeys identified at birth, + oneproduces very low levels of serotonin, - can identify gene that does this
- if leave them alone, they lead short miserable lives - other groups is very anxious
+ cross-foster monkeys - if do this with best of monkeys of previous generation, can completely ameliorate these effects - gets model of total interaction between nature & nuture with great loading of early environmental effects
- Genetics and parenting and response to stress
+ Changes in Central and Eastern Europe and Newly Independent States 1989-93 + real wage declines, declines in enrolment in pre-primary school + enormous increases in mortality (Russia, Ukraine, Hungary)
+ model of cross-sectional changs with enormous consequences in current time - cross-sectional changes, short-term, not a developmental effect
- how could Barker explain this + issues of how latent effects and pathway and cumulative effects play themselves out is large question for research, and not easy to study + 1958 birth cohort study - impaneled 17000 kids born in first month of March 1958 + He and Chris Power 1995 publication on 1958 birth cohort - social class at age 33 looked at by social class at birth + for each of outcomes - fair'/poor self-rated health
- limiting long-standing illness
- chronic respiratory syptoms
- high malaise
- obesity + have gradient based on social class at b irth, or both - see relative size of slope index used there (latent effect) + if go to educational qualifications these people got, size of gradient got larger - pathway effect + but things inter-correlate, need multi-variate model, of the bulls-eye - self-rated health in relation to early variables
- variables along child's pathway
- current time circumstances, in relation to material circumstances
- people's perceptions of civil society
- find independent contribution of early events when late events taken into account
- see evidence of independent effects + question of substitutability of socioeconomic and psychosocial determinants of health + what's enough to get by on for people? + large differences between political changes in central and eastern europe and how they affected the heatlh status of countries, try to understand these differences - intensity of level of socio-economic deprivation
+ look within a society (Russia, Poland, Ukraine)
+ Poland: for 20 years before wall came down - economic growth came to a stop, started to decline
- institutions that held society together began to degrade
- mortality differences between married and divorced people (index of social support) seemed to get larger as time went on Watson 1996 + raises question of how various determinants fit together? - if have strong social support network, could somehow buffer you against stresses in broader se environemtn + raises questions of Subsitutability of Social-economic determinants of health? - is it worse to be unemployed in a high unemployment community or a low unemployment one? + tracking biological embedding + can infer from population health patterns that there must be biological embedding and it must play itself over the life cycle + how to empirically test biological embedding? - have tried to pull together inter-disciplinary groups to empirically test this hypothesis + opportunities + hypothalamic pituitary axis central to biological embedding: + differential experiences early in life can affect hypothalamic pituitary axis, so get adaptive and maladaptive patterns to stresses of early life
- adaptive pattern is low basal cortisol and rapid rise/fall with stress
+ maladaptive pattern is higher basal level and shower rise/fall with stress, obtunded respond
- if integrate exposure to cortisol over lifetime, see higher level of exposure in those with maladaptive pattern
- chronic overexposure to coritsol is bad for several organ system
- do challenge tests for 50 year old men in a high mortality country such as Lithuiania and an low moratlity one as Sweden, see expected changes
- worth following through with prospective studies in people of different se circumstances from as early in life as we can, and have technology for sputum cortisol testing
+ broader idea: try to measure static difference in which there is a complex adaptive system (human body)
- everybody is own control group
+ so can't do a cross-sectional count variable on something in the body, it doesn't work, need to do studies defining individuals as their own control and challenge some biological parameter you draw from them, or challenging them from day to day and seeing how some biological parameters change
- complex and expensive, need enormous investment, consensus up front, that what you are measuring is worth measuring, difficult to get at the beginning
- in practical terms, trying to get scientists to take on biological embedding of social circumstances throughout the life cycle: progress over time is open question


Canada's health service system:
- universal access system,
+ argument to counter notion that Health Services not important

1. health care important at the margin - (if cut 30% of funding, would hurt those at the margin, would have health status implications) 2. health care works on morbidity, not mortality - much of this analysis here deals with mortality 3. role health care system plays as mechanism of social redistribution - distrubiton of access to health care is one of most important distributional aspects 4. the ability to maintain universal access is test of a society as a measure of quality of civic community - can we maintain enough social solidarity to provide universal access to health care when fancy health care becomes more and more expensive, and people are more and more discontent about using tax systems as a means of redistribution,
- his thought is that in terms of matrix of factors that make up civil societies, this becomes very important 5. socioeconomic non-financial barriers exist, - still see ses gradients in use of effective services, and when shift towards outpatient care, get shift towards smart consumers of health care, ie health care resources tend to go to upper ses groups who access services more, and less to lower ses groups who used inpatient services more - Canada Health Forum: get pharmaceuticals under single payer so can get control of price structure there, also long term care + principle policy issues are outside of the health care system
+ National children's agenda + intersectoral government and ngo program to improve experiences - MOH, social services, education, children and families, and justice, at federal, provincial and territorial level and starting to get municipal and buy into this, so can start some sort of national endeavour to enhance early childhood development + if it is really true that those first few years are really important, why is it that our collective responsibility taking really only begins at age 6 when kids go into school - why do we leave the period from age 0 to 5 as basically a few-fire zone where everybody is on their own?
- have to find ways to extend principles of collective responsibilites to age 0
- sounds pie in the sky, up against stakeholder pressures, current economic crises make it more difficult

Public Health Grand Rounds are a school-wide seminar series coordinated through the University of Washington School of Public Health & Community Medicine's Northwest Center for Public Health Practice.  For more information contact Chuck Treser (206-543-4207).


Revised: 12/28/98