Delmar Boulevard, Geo-mapping, and the Social Determinants of Health

The social determinants of health are those factors that affect people’s health status that are the result of the social situation in which they find themselves. Thus, in the well-known graphic from Healthy People 2010 (dropped, for some reason, from Healthy People 2020), which I have reproduced several times, they complement the other determinants such as the biological (genetics), but are represented in most of the other areas. Physical environment and socioeconomic environment, certainly, but even “behaviors” are affected by the circumstances into which one is born and lives. So is biology, actually, as we learn more about genetic coding predisposing some people to addictive behaviors. Certainly it is not all volitional or evidence of weak character.

The social determinants of health can be partially enumerated, and include adequate housing (including sufficient heat in the winter), adequate food, education, and also a reasonable amount of nurturing

and support from your family. In short, they are “the rest of life”, outside and often ignored by the healthcare system. Camara Phyllis Jones, in her wonderful “cliff analogy” (which I have also reproduced before) creates a metaphor in which medical care services are provided for those who need them (or “fall into them”) along a cliff face, while the social determinants of health are represented by how far a person, or a group of people, lives from that cliff face. As such, it illustrates the degree of protection that we all have from falling off that cliff, more for some and less for others.[1]

One of the clearest ways to show the impact of these determinants is by a technique called “geo-mapping” in which certain characteristics (income, educational level, gang violence, drug use, number of grocery stores or liquor stores, public transportation routes, whatever you can think of) are laid over maps of a city, town, or region. We have seen these portrayed on TV or in the newspapers as national and state maps for political events (such as what areas voted for who), but they can also be very useful for understanding the different challenges faces by people living in different areas. The work of Steven Woolf and his colleagues at Virginia Commonwealth University has greatly contributed to this work; in addition to their incredibly useful County Health Calculator, has produced graphs that can be found on the Robert Wood Johnson Commission for a Healthier America site that show how life expectancy can vary dramatically in different neighborhoods, as in the map displayed of the Washington, DC area, mapped along Metro lines for greater effect, or the one of my area, Kansas City, Missouri (which doesn’t have a Metro!)

A recent contribution to this field has been made by Melody S. Goodman and Keon L. Gilbert, of Washington University in St. Louis, who mapped the dramatic differences across Delmar Boulevard in that city, in “Divided cities lead to differences in health”. Their graphic shows the disparities in education, income, and housing value, and, unsurprising, racial composition, on either side of Delmar. This work was covered in a BBC documentary. Dr. Goodman, speaking to a symposium from her alma mater, the Harvard School of Public Health, is quoted as saying “Your zip code is a better predictor of your health than your genetic code.”

This is a pretty sad commentary, given not only the incredible amount of money that has been spent on unraveling the genetic code but the amount of faith and expectation that we have been convinced to have in how this new genetic knowledge will facilitate our health. By knowing what we are at risk for, genetically, the argument goes, science can work on “cures” that target the specific genes. This is a topic for a different discussion, but in brief one problem is that the most common diseases we suffer from are not the result of a single gene abnormality. It is probable that, at least in the short-to-medium term, knowledge of our genetics will be more likely to lead to higher life insurance rates than cures of our diseases. The more profound issue, however, is that there is evidence from the social determinants of health, from the work of Woolf and Goodman and many others, that we do not address the causes of ill health even when we know what they are.

Why is this so? Why is there such great resistance to understanding, believing, that investment in housing, education, jobs, and opportunities will have a much greater impact on people’s health than more and more money spent on high-tech medical care (and, of course, profit for not only the providers, but the drug and device companies and middleman insurance companies)? It is in part because we hope (and, when we are more privileged, expect) that we will be the beneficiaries. And it is also because we choose to believe that those who do not have the benefits we have (of money, education, family) somehow “deserve” it because of character flaws.

The issue of “fault” is articulately addressed by Nicholas Kristof in a New York Times Op-Ed on August 10, 2014, “Is a hard life inherited?” Kristof argues that it is, not genetically but because the circumstances to which one is born and in which one grows up, the presence of caring parents who read to you rather than beat you, who take care of you instead of abusing drugs, as well as adequate food and housing make a tremendous difference in how you turn out.

Indeed, another major study by Johns Hopkins sociologist Karl Alexander, to be published in his “life’s work”, “The Long Shadow: Family Background, Disadvantaged Urban Youth, and Transition to Adulthood”, and covered on NPR, confirms this. Alexander and his colleagues tracked nearly 800 children for more than 20 years, and found that those from less privileged backgrounds with lower incomes and less supportive families did worse. Only 33 of the children moved from the low income to the high income bracket. Problems with drugs and alcohol were more prevalent among white males than other groups, but they did better financially anyway. Some people, rarely, overcome the deck being stacked against them, but most of those who do well after being born with relative privilege would likely not be among them had they been in the same situation.  Kristof writes:

ONE delusion common among America’s successful people is that they triumphed just because of hard work and intelligence. In fact, their big break came when they were conceived in middle-class American families who loved them, read them stories, and nurtured them with Little League sports, library cards and music lessons. They were programmed for success by the time they were zygotes. Yet many are oblivious of their own advantages, and of other people’s disadvantages. The result is a meanspiritedness in the political world or, at best, a lack of empathy toward those struggling…

That lack of empathy leads to a lack of action; we are willing to accept people living in conditions that we would never accept for our family and neighbors, not only across the globe but across town, or even across a street. From the point of view of health, our priorities and investments are misplaced when we do not address the social determinants of health as well as cures for disease. When we do not try to change the known factors of zip code that impact our health as we investigate those of the genetic code.

If there are to be “cures” that come from our understanding of genetics, there is every reason to expect that they will be one more thing that is available to the people on the south side of Delmar Boulevard in St. Louis long before they are to those on the north side of the street.

---

[1] Jones CP, Jones CY, Perry GS, “Addressing the social determinants of children’s health: a cliff analogy”, Journal of Health Care for the Poor and Underserved, 2009Nov;20(4):supplement pp 1-12. DOI: 10.1353/hpu.0.0228

Weight and class: who is obese and why should we care?

One of the things that people are most fixated upon, in both the health arena and in society at large is weight. In popular culture, weight is a major issue. Celebrities are (mostly) thin; when they are not, and look like more of the regular people who are around us, they are seen as unusual. Diet books and “fad” diets abound as do classes to help us exercise. Issues of body image are major stressors for adolescents in particular, and health problems like anorexia are all too common. And, yet, an increasing number of Americans are obese, and health problems that are certainly associated with obesity – notably, but not only, Type II diabetes – are rapidly growing.

There is a major class association with weight; as income and class go down, prevalence of obesity goes up. Perhaps ironic compared to earlier centuries, when being heavy was associated with money – that is, the ability of the person to afford all that food – and poor people were starving. But if ironic, it is serious; the abundance of cheap, high-calorie foods in our society mean that poor people are not denied the opportunity to have lots of calories, but the stressors of poverty that affect all aspects of social life are still there, creating obesity as just one more problem to be confronted (or not).

In this context, a recent article in JAMA by Katherine M. Flegal and colleagues has garnered a lot of attention. “Association of All-cause mortality with overweight and obesity using standard Body Mass Index Categories: A systematic review and meta-analysis”[1], reviewed 97 studies with over 2.8 million people and encompassing 270,000 deaths, and performed a meta-analysis (a set of statistical techniques that allows accounting for studies that are somewhat different in design and have different numbers of people). The results were that people whose body mass index (BMI, a ratio of weight to height) was in the “overweight” range (25-30) had lower all-cause mortality than those in the “normal” range (18.5-25). In fact, the all-cause mortality rate was no higher for those in the range of “grade 1 obesity” (30-35) than for those in “normal” weight range. However, it was higher for those with grade 2 obesity (35-40), grade 3 obesity (>40) and for all obesity taken together (>30). In addition, unsurprisingly, the “hazard ratios” for mortality were greater for the same BMI when heights and weights were self-reported rather than measured (suggesting people under-estimate or under-report their weight, which would mean their BMIs are actually higher than reported).

This is not, of course, really new news, since all of the studies reviewed had been previously published. There was already a sense among many in the medical field that people at the “low end” of overweight (say 26-27) might be as least as healthy (have as low a mortality risk) as those at the low end of “normal” (18.5-25).  Heymsfield and Cefalu, in their editorial commenting on this study, “Does body mass index adequately convey a patient’s mortality risk?”,[2] say “Persons with a BMI between 18.5 and 22 have higher mortality than those with a BMI between 22 and 25. Placing these persons in a single group raises the mortality rate for the normal weight group. The average resulting from combining persons in the lowest mortality category (BMI of 22-25) with those who have greater mortality (BMI of 18.5-22) might explain why the NHLBI category of normal weight has an observed mortality similar to class 1 obesity (BMI of 25-30).”

If people with a BMI of 18.5-22 have a higher mortality rate than those with a BMI of 25-30, why, for goodness sakes, is 18.5-25 considered “normal”. For reference, a 5’4” person with a BMI of 18.5 would weigh about 108lbs, at a BMI of 22 it would be 128lbs, at 25, 145lbs, and at 30, 175lbs. For a person who is 5’10”, the weights at the same BMIs would be about 129, 152, 174, and 207. I think most people would not think that the lower range was normal except for models and marathon runners (hey, I’m overweight and would like to lose about 10-15 pounds of fat, but I think I’d be pretty sick before I lost the 60 lbs needed to get me to 22! My son was heavy when he developed Type I diabetes as a young adult, and went from a BMI of about 32.5 to about 21; let me tell you, he looked bad!), but the real question is “what are the healthiest ranges to be at”? This is what official recommendations should be based on, and it is clear from the work reported by Flegal and colleagues that this is not the case for the current numbers.

The relationship between adiposity (presence of significant amounts of excess fat) and risk for many diseases is well-established; the relationship between adiposity and BMI less well so. Variables include amount of muscle mass (not a risk factor but leading to greater weight-for-height), sickness (people who lose weight as a result of disease), and overall body structure. I tried to find out where these ranges come from, but have, so far, been unsuccessful. I found the references to the “Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults”, put out by an expert panel in 1998, but at least their “Executive Summary”^[3] does not reveal the source of how the “normals” were derived; they are just asserted. 

Obviously, this is going to be controversial. Paul Campos’ op-ed piece in the NY Times on January 3, 2013, “Our absurd fear of fat”, makes many of the same points I have, but letters generated in response range from those lauding it and saying people (especially children) should be taught to be proud of their bodies, to those arguing it minimizes the health dangers of obesity. What is clear, though, is that the fixation on “ever thinner” that exists in much popular culture has no place in health discussions. The JAMA article strongly suggests that our standards for “normal”, “overweight” and “obese” BMI are too low, although not irrelevant. It calls our attention to a tendency to other similar areas in which health professionals have adopted uni-dimensional disease markers and driven them even lower, to result in poor health outcomes for many. Recent examples include blood sugar (or its related value, hemoglobin A1c), blood pressure, and cholesterol. Studies that held everything else equal found benefit in lower values, so experts kept driving down the definitions of normal and desirable for these tests. Unfortunately, not everything else is equal. Pushing the desirable hemoglobin A1c level of people with diabetes to 5% instead of 6% led to a lot of morbidity from hypoglycemia; lowering cholesterol goals led to toxicities from drugs; lowering blood pressure goals to poorer functioning and greater mortality in some populations, especially the elderly. Most people don’t exercise regularly, but rather than lauding all efforts to exercise, “experts” keep raising the bar for how often, how long, and how intense exercise should be.

So let’s get back to class, and its associated characteristics. It is time for health professionals to recognize that they are also social service professionals and members of a society whose broad policies have a much more profound impact upon health than small numbers variation in BMI, blood pressure, cholesterol, and blood sugar. We need to treat, as well as support and encourage, people at the extremes whose health is at risk, but we shouldn’t fall prey to definitions that name more people as diseased and needing interventions and distract us from the real business at hand.

Which is creating a more just, fair, equitable and safe society.

---

[1] Flegal KM, et al., “Association of All-cause mortality with overweight and obesity using standard BMI Categories: A systematic review and meta-analysis, JAMA Jan 2 2013;309(1):71-82

[2] Heymsfield SB, Cefalu WT, Does body mass index adequately convey a patient’s mortality risk?” JAMA Jan2 2013;309(1):87-88.

[3] Expert panel, “Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary”, Am J Clin Nutr, 1998;68:899-917.

Health Outcomes: The interaction of class and health behaviors

.
I have recently discussed (Poverty, Primary Care and the Cost of Medical Care, February 10, 2010) the “Whitehall Studies” conducted by Sir Michael Marmot and colleagues that “demonstrate that there is a more or less linear correlation between health (including longevity) and increasing social class". That piece discussed the report of a panel headed by Marmot, “Fair Society, Healthy Lives”, that shows that these problems have not been resolved. A new paper from the follow-up “Whitehall II” study, conducted by Silvia Stringhini and colleagues from both Britain and France, “Association of socioeconomic position with health behavior and mortality”, (JAMA Mar24/31,2010;303(12):1159-66), examined the role of alcohol, tobacco, diet, and physical activity in accounting for these differences over an extraordinarily long 24-year follow-up period.

Stringhini, et. al., found that in fact adverse health behaviors accounted for about 42% of the increase in mortality in lower socioeconomic groups (which was about 1.6 times as high in lowest than in the highest socioeconomic group). Smoking was the most powerful negative factor, with the others contributing a smaller amount. “There was a marked social gradient in health behaviors at baseline. Participants in the lower socioeconomic positions were more likely to smoke, abstain from alcohol consumption, follow an unhealthy diet, and be physically inactive and less likely to consume heavy amounts of alcohol.” Most of this is consistent with the observations of physicians and epidemiologists in the US, with the surprising exception of alcohol use being lower in lower income groups. This may be a difference between the US and Britain; in Britain, in the 20th century, cirrhosis was a disease largely of the upper class who could afford the highly taxed, and high alcohol content, distilled spirits. Another possibility (and this is my speculation, not data) is that the lower socioeconomic group studied by Whitehall II in England may have a large component of Muslims, who do not drink. In any case, the impact of smoking, poor diet, and physical inactivity accounted for a significant part of the class difference in mortality, although it did not account for even the majority of that difference.

Thus, this study supports two well-established assumptions: 1) that adverse health behaviors are a significant contributor to ill health and higher age-adjusted mortality rate, and 2) that people in lower socioeconomic groups have worse health and higher mortality rates, much, but not all, of which can be associated with their higher rates of adverse health behaviors. Previous work on the results of Whitehall have suggested, and demonstrated evidentiary support for, the hypothesis that stress in daily life (of worrying about how you will pay the rent and feed your family, whether you are going to lose your job, or, particularly in the case of ethnic and racial minorities, not only whether you will be arrested or harassed by the authorities but the indignities of ongoing discrimination), mediated through only partially understood neurochemical pathways, account for much of this effect. However, to the extent that people can divest themselves of risky health behaviors, they can decrease, if not eliminate, their higher risk for adverse health outcomes.

In the same issue of JAMA, James R. Dunn of McMaster University in Canada, has a very insightful editorial commenting on the Stringhini article, “Health behavior vs the stress of low socioeconomic status and health outcomes” (JAMA, Mar24/31, 2010;303(12):1199-1200). He repeats the caution of the Whitehall authors that the population studied in the Whitehall cohort may not be representative of the British population overall (and, by extension, of the US or Canadian population). Indeed, the cohort was originally selected by Marmot and colleagues to reduce the confounding that might come from general studies of people in different classes because of occupational risks. Dunn points to the association of the stress of low socioeconomic status and the prevalence of adverse health behaviors: “…it is possible to consider both factors [stress and behavior] as part of the same pathway between relatively low socioeconomic status and health. Unhealthy behaviors are more common among individuals with low socioeconomic status because of the stress of low socioeconomic status. Accordingly, there is a direct causal pathway between low socioeconomic status and poor health as well as an indirect causal pathway through health behavior, which reinforce one another over the lifecourse.” That is, the stress of being poor makes you more likely to do unhealthful things that we know about (smoking, poor diet, low physical activity) that make you less healthy, and also makes you less healthy through a pathway that we don’t completely understand.

Dunn notes that while changing health behaviors in lower socioeconomic populations would be a good thing, “The problem is that traditional individually oriented health behavior education interventions are not very effective, and individuals with low socioeconomic status have been notoriously difficult to reach with such programs”. He discusses a variety of early childhood developmental characteristics, especially “executive function” and “self regulation” which might increase the probability of not adopting or stopping adverse health behaviors, which are on average less well developed in those growing up in lower socioeconomic groups, presumably also as a result of the stress impacting them as young children.

The relatively good news from the Stringhini study is that the prevalence of many adverse health behaviors did decrease over the time period studied. For smoking, the prevalence decrease from 10.1% to 4.8% in the highest, and from 29.7% to 16.5% in the lowest socioeconomic groups and unhealthy diet from 5.8% to 1.0% and 14.9% to 5.2% respectively diet; on the other hand, sedentary behavior increased from 6.6% to 21.4% in the highest and from 35.4% to 41.6% in the lowest socioeconomic groups. Again, extending this to the whole British population is uncertain, and in the US the prevalence of obesity (a combination of both poor diet and physical inactivity) is growing at a staggering rate in all age groups, and especially in low socioeconomic groups.

The take-home message is that all people should be encouraged and supported to adopt healthful and eschew unhealthful behaviors, particularly related to smoking, diet and exercise, and the degree to which any programs can be demonstrated to be successful for large numbers of individuals or, better yet, groups, they should be promulgated and replicated. However, to have greater success, programs will have to strike closer at the etiologies of these behaviors. A lower level, achievable (and achieved in some jurisdictions) by legislation, exemplified by indoor smoking bans, calorie and fat content labeling of foods, especially fast foods, and banning the use of toys as gifts in fast-food meals (as recently done in Santa Clara County, CA), can have much more significant impact (see “Promoting health through tobacco taxation” by Ali and Koplan from JAMA, and “Cardiovascular effect of bans on smoking in public places: a systematic review and meta-analysis” by Meyer, et. al., in JACC, cited in The Public’s Health: Smoking and Salt, February 6, 2010).

The greatest changes, however, involve even more significant societal changes: the elimination of the wide disparity in income and opportunity, thus socioeconomic status, and of racism. Health-focused, as well as social justice focused, policies should try to achieve this end, but in the US it will be a long time coming. In the meantime, it remains a good idea to choose your parents wisely; being born white and rich still significantly enhances your health status.
.