Correlations: Obesity, Geography, Race and Cancer

MSN published the newest state-by-state obesity figures on their web site today.  Here are the states with the highest and lowest rankings.  (Note:  in this table and all other tables, I’m skipping Washington, D.C.  It’s a city, for Pete’s sake, not a state.)

1 Mississippi
2 Alabama
3 Tennessee
4 West Virginia
5 Louisiana
6 Oklahoma
7 Kentucky
8 Arkansas
9 South Carolina
10 North Carolina
41 California
42 New Jersey
43 Montana
44 Utah
45 Rhode Island
46 Vermont
47 Hawaii
48 Massachusetts
49 Connecticut
50 Colorado

As you can see, southern states dominate the obesity list.  Quite a few articles have been published that attempt to answer the question Why Are Southerners So Fat? — that was actually the title of one.  I believe we’re looking at two factors, which are in part related:  poverty and race.  Cheaper foods tend to be high-carb foods, so it would make sense that poor people tend to eat high-carbohydrate diets.

Tennessee has the third-highest obesity rate.  And yet as I’ve mentioned before, I don’t see many fat people in our area.  In fact, in a list of county-by-county statistics, I saw that the average adult BMI in our county is 25.  Our county also has the highest per-capita income in the state. 

But I also believe race plays somewhat of a role apart from poverty.  Here’s the same list of states again, with another column showing each state’s rank by the African-American proportion of the population:

State Obesity Blacks/Capita
Mississippi 1 1
Alabama 2 6
Tennessee 3 10
West Virginia 4 37
Louisiana 5 2
Oklahoma 6 24
Kentucky 7 23
Arkansas 8 12
South Carolina 9 3
North Carolina 10 7
California 41 26
New Jersey 42 16
Montana 43 50
Utah 44 44
Rhode Island 45 30
Vermont 46 48
Hawaii 47 39
Massachusetts 48 27
Connecticut 49 21
Colorado 50  33

It’s hardly a perfect correlation, but the correlation is there.  The southern states tend to have a higher proportion of African-Americans than the northern states. 

As I mentioned in Fat Head, African Americans and Latinos are disproportionately labeled as overweight or obese, and to a large extent, the label isn’t fair.  I’ve had a few pinheads in cyberspace accuse me of being a racist for stating that on average, African-Americans and Latinos are genetically pre-disposed to be thicker than whites.  They may as well yell “racist!” if I say that those same groups are pre-disposed to have darker skin.  We’re talking about physiology here, pure and simple.

If you look up studies on osteoporosis, you’ll see it stated over and over that African-American and Latina women are the least susceptible.  You’ll also find the reason:  denser, thicker bones — African Americans in particular.  You’ll also see it stated that Asians on average have thinner bones than whites. 

Go figure … given the same height, smaller-boned people tend to be lighter, and thicker-boned people tend to be heavier.  The thicker-boned people also tend to carry around more muscle.  In addition, I’ve seen it stated in research papers that African-Americans on average tend to have bigger thigh muscles — some of the largest muscles in the body. 

So if we’re determining who’s overweight or obese by simply (and stupidly) comparing height to weight, African-Americans are going to have the highest BMI, followed by Latinos, whites, and Asians.  Yes, people with the highest BMI scores may also tend to be the fattest overall.  But the statistics are skewed.

As long as I was looking up data and making correlations, I decided to have a little fun with another set of statistics.  Below are the highest and lowest rates of cancer by state.  (#1 = highest rate of cancer, #50 = lowest.)

1 New Jersey
2 Maine
3 Rhode Island
4 Kentucky
5 Massachusetts
6 Pennsylvania
7 Connecticut
8 Washington
9 Michigan
10 West Virginia
41 Texas
42 California
43 Maryland
44 Montana
45 Hawaii
46 Colorado
47 South Dakota
48 Utah 
49 New Mexico
50 Arizona

Northern states dominate the top 10, while the bottom 10 is made up largely of sunshine states.  Most likely, we’re looking at the effects of vitamin D.  I was curious as to why relatively sunny states like Kentucky and West Virginia have such high rates of cancer, so I looked up smoking statistics.  Sure enough, West Virginia ranks #1 and Kentucky ranks #3, after Indiana.

But here’s what’s really interesting:  New Jersey, with the highest overall cancer rate, ranks #49 in smoking.  Maine ranks #27, and Rhode Island ranks #35.  In those northern states, it’s not smoking that’s causing the high cancer rate.

In previous posts, I’ve tried to pound home the point that correlations don’t tell us much.  Traits that are correlated often have independent causes … and yet researchers and health reporters too often assume that if two traits are related, one must be causing the other.

Look at the table below, where I’ve put the cancer rankings alongside the obesity rankings.  (I’ve removed Kentucky and West Virginia because of the high smoking rate — we’ll call that a confounding variable.)

State Obesity Cancer
Mississippi 1 24
Alabama 2 25
Tennessee 3 21
Louisiana 5 11
Oklahoma 6 31
Arkansas 8 38
South Carolina 9 33
North Carolina 10 35
California 41 42
New Jersey 42 1
Montana 43 44
Utah 44 48
Rhode Island 45 3
Vermont 46 26
Hawaii 47 45
Massachusetts 48 5
Connecticut 49 7
Colorado 50  30

You know what I see there?  I see evidence that obesity reduces your risk of developing cancer.   Or, to adopt the style I often see in health-article headlines, LOWER BODY WEIGHT RAISES CANCER RISK.

Somebody get T. Colin Campbell on the phone …

Share

58 thoughts on “Correlations: Obesity, Geography, Race and Cancer

  1. Ned Kock

    > Interesting. Any difference in the amount secreted by subcutaneous or visceral fat? Visceral fat seems to be more metabolically active, and not in a good way.

    I think that loss of visceral fat will have a more pronounced effect on adiponectin levels than loss of subcutaneous fat. The flip side is that abnormal visceral fat gain will have more of a negative effect. And this will not be only in adiponectin, but also various pro-inflammatory hormones. Here is a post:

    http://healthcorrelator.blogspot.com/2010/07/subcutaneous-versus-visceral-fat-how-to.html

    In other words, excess subcutaneous fat is not nearly as detrimental as excess visceral fat, as you noted. Excess visceral fat typically happens together with excess subcutaneous fat; but not necessarily the other way around. For instance, sumo wrestlers frequently have excess subcutaneous fat, but little or no visceral fat.

    The more health-detrimental effect of excess visceral fat is probably related to its proximity to the portal vein, which amplifies the negative health effects of excessive pro-inflammatory hormone secretion. Those hormones reach a major transport “highway” rather quickly.

    That makes sense.

    Reply
  2. Laurie

    “Ketosis-Prone Diabetes”. I had to search around and look this up. Just a quick and cursory look and I found a few journal articles that go on to say this is KETOACIDOSIS – in the finer print.
    I will delve further, but as far as I know there is definitely a large difference between ketosis and ketoacidosis.
    I had a request to look up and find an article about low cholesterol and increased cancer risk. I found one that was particular to the cancer in question (hairy cell leukemia). The study pretty much supported that LOW blood cholesterol was correlated with INCREASED cancer cell proliferation. The Abstract to this article noted this and actually concluded with something like, ‘but we all know cholesterol in the diet is bad so we recommend further reducing dietary cholesterol and further drugging down cholesterol’!. You just can’t invent some of this stuff.

    Amazing, isn’t it? It’s just like the ADA’s advice to eat carbs, even though their own literature says carbs raise blood sugar.

    Reply
  3. Laurie

    “Ketosis-Prone Diabetes”. I had to search around and look this up. Just a quick and cursory look and I found a few journal articles that go on to say this is KETOACIDOSIS – in the finer print.
    I will delve further, but as far as I know there is definitely a large difference between ketosis and ketoacidosis.
    I had a request to look up and find an article about low cholesterol and increased cancer risk. I found one that was particular to the cancer in question (hairy cell leukemia). The study pretty much supported that LOW blood cholesterol was correlated with INCREASED cancer cell proliferation. The Abstract to this article noted this and actually concluded with something like, ‘but we all know cholesterol in the diet is bad so we recommend further reducing dietary cholesterol and further drugging down cholesterol’!. You just can’t invent some of this stuff.

    Amazing, isn’t it? It’s just like the ADA’s advice to eat carbs, even though their own literature says carbs raise blood sugar.

    Reply
  4. Vin

    As for my calculations , a medium serving of chips has 40g carbs (-fiber). A Lager has 10g and a cola 40g. Depending on what he liked in his fry up he might have had some baked beans and adding in the carbs from his hamburger bun puts me at about 280g.

    Just keep repeating: it was the burgers that killed him. It was the burgers that killed him. It was the burgers that killed him.

    Reply
  5. Vin

    As for my calculations , a medium serving of chips has 40g carbs (-fiber). A Lager has 10g and a cola 40g. Depending on what he liked in his fry up he might have had some baked beans and adding in the carbs from his hamburger bun puts me at about 280g.

    Just keep repeating: it was the burgers that killed him. It was the burgers that killed him. It was the burgers that killed him.

    Reply
  6. Walter B

    One factor linking obesity and income, is that high carb diets causes low energy states which is bad for staying employed. I lost at least one job for not being able to stay awake at my desk. After going low carb, I stopped having that problem. Also is the effect of obesity induced sleep problems.

    Reply
  7. Walter B

    One factor linking obesity and income, is that high carb diets causes low energy states which is bad for staying employed. I lost at least one job for not being able to stay awake at my desk. After going low carb, I stopped having that problem. Also is the effect of obesity induced sleep problems.

    Reply
  8. chuck

    Although politics is a subject to be avoided in a lot of cases. I saw that list yesterday and thought to myself “Red states, Blue states???”. There is a pretty interesting correlation. Democratic states tend to be less obese. I have thoughts but this is not the forum.

    Reply

Leave a Reply to Dave Wilson Cancel reply

Your email address will not be published.