by Jason S. Brinkley, PhD, MA, MS

On the Brink addresses topics related to data, analytics, and visualizations on personal health and public health research. This column explores current practices in the health arena and how both the data and mathematical sciences have an impact.

Jason S. Brinkley, PhD, MA, MS

It’s no secret that Americans are divided. Not just geographically, though geographic differences certainly determine many things about our lives. There’s also religious, financial, and cultural diversity impacting everything from where we eat, to where we shop, to what we do for entertainment.

Even what kills us.

The Centers for Disease Control and Prevention estimates that 2.6 million Americans die every year from various causes, with heart disease (614,348 deaths) and cancer (591,699 deaths) leading the way. While 2.6 million deaths is certainly a large number of people, a population of almost 320 million turns the overall yearly death rate from all causes to about 0.8%.

The percentages get even smaller when we start talking about specific causes of death. But that is where the numbers can begin to be misleading. Indeed, it is difficult to really get a handle on things that are pretty rare when they become represented by such small numbers. Scientists have tried to deal with the problem by putting death rates and other rare events on alternative terms, and mathematics has even developed whole new fields and measures just for dealing with rare events.

So instead of using a number like 0.8%, we say that there are about 822 deaths per 100,000 people. Even that gets hard to deal with when you consider the “living rate” is 99,178 per 100,000 people, but “death rate per 100,000” gives us a good starting place for comparing and estimating risk.

So, in a country as diverse as America, what happens when we look at the death rate for different events and how they impact various groups? How does breast cancer stack up to gun homicides to Sudden Infant Death Syndrome?

For the purposes of this examination, I will stick to only publicly reported sources and I will try to lean toward sources that are considered “mainstream”: charities, news outlets, or government sources that can be found through simple Google searches. There can be controversy surrounding calculations for death counts and rates, and I will say up front that I am side-stepping that entire conversation. Simply put, I’m going with just reporting what other people have found. Also, I only report statistics that are denoted in the “per 100,000” individuals in the population of interest.

So, what kills us? Beginning with breast cancer, the Komen Foundation reports that the death rate for breast cancer among all women is 20.7 per 100,000. For men, the death rate is 0.3 per 100,000, so women are 69 times more likely to die from breast cancer than men. Comparing that to gun homicides, the FiveThirtyEight blog recently released a controversial interactive graphic titled “Gun Deaths in America.” They report an annual tally of 9,935 gun homicides of males in America, which gives a death rate of 6.4 per 100,000. For women, the rate was reported as 1.1 homicides per 100,000. Digging deeper, however, we find that the reported death rate from homicide for black males aged 15 to 34 is an astounding 73.5 per 100,000.

How about children? The CDC maintains data on Sudden Infant Death Syndrome (SIDS) and finds the recent death rate to be 38.7 per 100,000 live births, which is in stark contrast to the rate of 130.3 deaths per 100,000 live births in 1990.  The CDC reports these types of national vital statistics yearly, with death rates for the top 10 leading causes of death across all ages, races, genders, and combinations therein. They created a seminal, 95-page report on this, which you can dive into online by going here. That report finds things such as the accident death rate for all children aged 1 to 4 is the same as the suicide rate among all 15- to 19-year-olds at 3.8 per 100,000.

Turning toward the elderly, the same report finds the Alzheimer’s disease death rate among white women aged 65 and older is 248 per 100,000, while the diabetes mellitus death rate among black men aged 65 and older is 217.4 per 100,000. But in black females aged 65 and older, the Alzheimer’s death rate is 166 per 100,000, and the diabetes mellitus death rate is 196.1 per 100,000. Those rates are even lower among Hispanics of both sexes (121.9 for Alzheimer’s and 153 for diabetes).

The picture we’re painting here is incomplete, and I’ve cherry-picked a variety of different sources and events in order to explore an interesting theme of diversity. It is certainly true that age and race play a large role in death rates, and there are many places where researchers may want an adjustment made to make better comparisons. We have also made great strides in medical and public health arenas where the death rate from different events has fallen dramatically, like we see in SIDS. A combination of research, policy interventions, and public engagement has created great strides in early detection and creation of life-saving best practices for even some of the things we’ve discussed today.

What we are looking at here are purely outcomes, counts, and causes of death, which for most intents and purposes is the end of a person’s story. To me, the most interesting idea stems from exactly how we got to this diversity of causes of death. Science says that genetics and family history play a key role in cause of death, but environment is important as well. And not just where you live, but HOW you live.

There is ongoing research into the idea of “allostatic load,” which looks at how the human body responds to stress. Humans deal with chronic or repeated stress in different ways, and some argue that just living different lifestyles can make us more or less eligible for certain outcomes simply by what we are exposed to and what options we have for coping. There is a lot of ongoing effort to sort out the cause and effect of geography, income, lifestyle, and even cultural impacts on long-term health and death rates. For now, the best take-home message is that the picture is not fully complete, but there do seem to be differences across race, gender, and age lines on how we live and how we eventually die.

But is that the sum total of the story? The central message seems to be that causes of death vary and some causes disproportionately affect some groups more than others. So as these different groups vie for attention and resources, should we just be content to know that here, like in so many other aspects of American life, competing interests will continue to polarize Americans? What about common ground and what about the initial question, what kills us?

Make no mistake that the death rates compared here all pale in comparison to heart disease and cancer deaths. The counts listed at the beginning roughly translate into 193 heart disease-related deaths and 185 cancer-related deaths from the 822 deaths per 100,000 total American population. At almost 46% of all deaths, these are the common enemy.  Heart disease and cancer are the top two causes of American deaths for whites, blacks, Hispanics, Asians, men, women, people aged 45+, and all combinations therein. Preventing and lowering the death rates in many of these other areas is certainly needed, but there is still a continued need to unite around these two.

That’s all for today. Welcome to On the Brink. In the coming months we will explore topics like this as we look at (and potentially abuse) publicly reported statistics, data, visualization and methods for research. Whenever possible, there will be some jokes and practical lessons thrown in, though this week it’s hard to have a good laugh when talking about death rates. But let’s end with a joke anyway, one that the Guardian recently reported as winner of a local gag contest in Europe.

Submitted by Masai Graham, the joke goes, “My dad has suggested that I register for a donor card. He’s a man after my own heart.” I hear that joke killed in Edinburgh, although I expect the death rate is very low.

Jason S. Brinkley, PhD, MS, MA is a Senior Researcher and Biostatistician at Abt Associates Inc. where he works on a wide variety of data for health services, policy, and disparities research. He maintains a research affiliation with the North Carolina Agromedicine Institute and serves on the executive committee for the NC Chapter of the American Statistical Association and the Southeast SAS Users Group. Follow him on Twitter. [Full Bio]

jphmpdirect