Comedian/documentarian Tom Naughton recently made a highly intriguing post about the “Spanish Paradox”; that is, the low rate of cardiovascular illness among Spaniards despite their apparently poor markers of heart health.
This post was made on the discussion site of Naughton’s 2009 documentary Fat Head. This movie (which I have yet to see, but plan to soon) criticizes the conventional wisdom on weight, fat intake, and health. Particularly, it was meant to be a response to Morgan Spurlock’s 2004 documentary Super Size Me, in which Spurlock went on an all McDonald’s diet for 30 days. Naughton found that the conventional wisdom about heart health and diet was deeply flawed, as he demonstrated in the film.
That conventional medical wisdom is highly suspect is not too surprising, as this is in line with John Ioannidis’s finding that most published research findings are false. This problem is particularly pervasive in the human sciences, especially medicine, (and of course, psychology, something with which regular readers are all too familiar).
In keeping with his theme critical of the exact role of diet and fat in health and cardiovascular illness, Naughton cited this chart from the American Heart Association on the incidence of cardiovascular illness and death in selected countries:
The first thing that jumped out at me was the very tight association of related groups. Roughly, we see a pattern that goes like this (in descending order of illness rates):
- Slavs and Eastern Europeans
- Celtic groups
- Germanic and Anglo groups
- Mediterraneans, tied with East Asians
Edit, 1/26/13: See this map of age-adjusted cardiovascular death rates across much of Europe, by region. This is of males, age 45-74, year 2000.
Notice the apparent Northeast to Southwest cline.
While there are some irregularities (i.e., China, Australia, and Greece), the general pattern is strong. This strongly suggests that genetic factors are involved. Indeed, Naughton brilliantly highlights the problems with the traditional markers of heart health:
Surveying a cross-section of Spanish adults, Dr Auxiliadora Graciani (Universidad Autonoma de Madrid, Spain) and colleagues measured “ideal cardiovascular health”–as described by the AHA–and found that only 0.2% of the 11, 408 subjects attained ideal values for all seven CVD health metrics: nonsmoker, body-mass index (BMI) <25 kg/m2, physical activity at goal, diet consistent with recommendations, untreated cholesterol <200 mg/dL, untreated BP <120/80 mm Hg, and untreated fasting glucose <100 mg/dL in the absence of clinical CVD and diabetes. Of those surveyed, 3.4% attained ideal values for at least six of the metrics and 15.4% for five, they note in their report published online January 8, 2013 in Circulation: Cardiovascular Quality and Outcomes.
Wow, that’s just awful. Barely one-fifth of one percent of the Spanish adults meet the American Heart Association’s ideal values for cardiovascular health. And I thought the people I saw falling down in Barcelona were just party animals who’d had too much sangria. Now I realize they were having heart attacks. In retrospect, I feel guilty for clapping.
“This is the first study to report information on cardiovascular health from Spain, a European country with low coronary heart disease mortality compared with many Western countries,” observe the authors. However, the level of cardiovascular health in Spain “is as low as in the United States, primarily due to poor lifestyles, especially lack of ideal diet,” they note.
Um … uh … but … did I read that correctly? The Spanish have a low level of cardiovascular health but also low coronary heart disease mortality? Isn’t that a bit like saying they’re in poor physical condition, but can run for miles and bench-press their own body weights? I don’t know about these researchers, but I define “good cardiovascular health” as “not dying from cardiovascular disease.”
Although the CHD mortality rate is low in Spain, recent research shows the prevalence of angina there is high, suggesting that atheromatous plaques are stable, the doctors say. Research is needed as to why these plaques do not rupture, translating into a high incidence of acute MI, as is seen in other countries, they state.
Oh, okay. Got it. The researchers are speculating that thanks to their stubborn refusal to adopt the American Heart Association’s definition of good health and a good diet, the Spanish actually do have lots of plaque buildup … but for some reason, plaques don’t rupture in Spain. Must be the weather. I’m surprised they didn’t say it’s because the Spanish drink wine.
This “paradox” of low CHD mortality and poor cardiovascular health in Spain could be explained by several things other than traditional CV risk factors, such as sedentary behavior, specific dietary components (eg, wine consumption and the Mediterranean diet), psychosocial factors (such as family support), and quality of medical care, among others, say Graciani et al.
Head. Bang. On. Desk.
I see. So the Spanish suffer from a “lack of ideal diet,” they have poor cardiovascular health because they don’t meet the American Heart Association’s criteria, but they don’t die from heart attacks because they eat a Mediterranean diet. They have a lousy diet, but their diet saves them from heart disease. Oh, and because they drink wine. And because of the quality of medical care. I guess that means they get treated for their high cholesterol.
And because lifestyles appear to contribute far more heavily than biological factors to poor cardiovascular health in Spain, there is a great need to strengthen the role of public-health efforts in the management of CVD there, they observe. Health services also need to improve, they add, noting that five out of every 10 people with elevated cholesterol are not being treated, and half of those with BP >140/90 mm Hg are unaware of this.
Head. Bang. On. Desk.
So, to sum up … The Spanish are in poor cardiovascular health because of their lifestyle and lack of an ideal diet, which means there’s a great need to strengthen government efforts in the management of cardiovascular disease. But they don’t die of heart attacks because of their diet and high-quality medical care.
I want to know how any researcher can make those arguments in a paper without reading what he wrote and thinking, “Wait a minute … I sound like a flippin’ moron here.”
Naughton continues, noticing that the connection between cholesterol levels and cardiovascular death rates do not seem to cross national lines:
A couple of years ago, I downloaded data from the World Health Organization’s MONICA study, which tracks cholesterol levels and cardiovascular death rates around the world. The average cholesterol level in Russia is 189. In France, it’s 210. Romania (near the top of the chart) and Spain (near the bottom) have the same average cholesterol level: 197. When I ran the correlation function in Excel on all the countries and their rates of cardiovascular deaths, the result was -0.25. In other words, there’s almost no correlation, and the slight correlation that exists points to cardiovascular deaths going up as cholesterol levels go down. (emphasis added)
But then, why should it? Cholesterol levels may be predictive within a group, but may mean nothing across groups. This is because (to quote the venerable HBD Chick) “different peoples is different.” What may be a perfectly acceptable level of cholesterol for individuals in one group may be highly detrimental to individuals in another (this is assuming that cholesterol levels matter at all). Indeed, failure to recognize this fact stymies research in this field.
Naughton continues in a newer post, this time looking at Norwegians:
…the researchers report on the results of applying them to data collected from several thousand Norwegians. Here’s what they found:
At age 40, 22.5% of women and 85.9% of men were at high risk of cardiovascular disease. Corresponding numbers at age 50 were 39.5% and 88.7%, and at age 65 were 84.0% and 91.6%. At age 40, one out of 10 women and no men would be classified at low risk for cardiovascular disease.
Hmmm … people in Norway must be dropping like flies from heart disease, at least according to the prevailing guidelines for estimating heart-disease risk … you know, cholesterol levels and all that stuff.
But as we see from the above post, they’re not.
So if heritable group differences are involved in heart health, what could those heritable factors be? One clue is digestive function itself. A Scientific American guest blog post by Rob Dunn, which discussed some of the complexities in measuring the effective calories delivered by food touched on one potential factor:
A Body is Not a Body—Amazingly, there are more ways in which a calorie is not a calorie. Even if two people were to somehow eat the same sweet potato cooked the same way they would not get the same number of calories. Carmody and colleagues studied a single strain of heavily inbred lab mice such that their mice were as similar to each other as possible. Yet the mice still varied in terms of how much they grew or shrank on a given diet, thanks presumably to subtle differences in their behavior or bodies. Humans vary in nearly all traits, whether height, skin color, or our guts. Back when it was the craze to measure such variety European scientists discovered that Russian intestines are about five feet longer than those of, say, Italians. This means that those Russians eating the same amount of food as the Italians likely get more out of it. Just why the Russians had (or have) longer intestines is an open question. Surely other peoples differ in their intestines too; intestines need more study, though I am not going to volunteer to do the dirty work. We also vary in terms of how much of particular enzymes we produce; the descendents of peoples who consumed lots of starchy food tend to produce more amylase, the enzyme that breaks down starch. Then there is the enzyme our bodies use to digest the lactose in milk, lactase. Many (some say most) adults are lactose deficient; they do not produce lactase and so do not break down the lactose in milk. As a result, even if they drink milk they receive far fewer calories from doing so than do individuals who produce lactase. Each of us gets a different number of calories out of identical foods because of who we are and who our ancestors were. (emphasis added)
It’s only too bad that the “craze” of measuring human differences went out of fashion, but regular readers know how that is.
Human bodies vary in both their efficiency in absorbing calories and nutrients as well as how well they metabolize them. This is to be expected, as recent evolutionary forces – particularly agriculture – have had a huge impact on human evolution, one of the most salient of these which is significant to diet is the distribution of lactose intolerance:
Additionally, there are heritable differences in how the body responds to certain chemical levels, such as cholesterol or fat levels, as Naughton notes about the Spanish and the Norwegians. (Indeed, as Greg Cochran and Henry Harpending note in The 10,000 Year Explosion, Africans are more retentive of salt, which is involved in the higher rates of hypertension among Blacks.)
But that’s not the only area where heredity could make a difference. Heritable behaviors could be at play (e.g., smoking or alcohol consumption), including societal structure (which may impact stress levels). Dietary preferences, which are clearly heritable between individuals within a group, likely plays a role in the preferred foods of each culture (not for the least reason that the same foods taste different to different people). The effects of heredity could range from the most direct (metabolism) to the most indirect (levels of societal stress).
My Twitter followers might have noticed a recent and somewhat heated discussion between myself and Naughton over a potential partial genetic explanation for the “French (and Spanish) Paradox”, particularly over the genetic distinctiveness between the English and the French. Of course, while there’s no question that the French and the English are genetically distinct today, to test the extent of the genetic contribution to the observed differences, we could look at some of the offshoots of said groups. To that end, I decided to take a look at the Québécois.
Here are the rates of “cardiovascular disease factors” across the various Canadian provinces:
|Prevalence of risk factors for cardiovascular disease in Canadians ages 12 or over by province or territory|
|Province||Current smoker (%)||Hypertensive (%)||Diabetic (%)||Obesity (BMI ≥30) (%)||Physically Inactive||Low Income (%)|
|Price Edward Island||27.9||14.0||5.0||18.5||56.4||13.0|
And here are death rates from cardiovascular diseases in the various Canadian provinces:
|Cardiovascular disease death rate per 100,000 population in Canada|
|Province||Death Rate||Province||Death Rate|
|Prince Edward Island||264||Manitoba||257|
Quebec has significantly a significantly lower obesity rate that its fellow Canadian provinces. As well, it has a somewhat lower cardiovascular mortality rate, though the difference here isn’t as dramatic as the difference between France proper and the Anglosphere. While not extremely so, this is consistent with the genetic explanation. (However, it’s worth noting that the Québécois aren’t genetically representative of all French, having descended from about 2,600 colonists from a few specific regions of France. As well, the Québécois have undergone subsequent evolution).
As Naughton noted, there are outliers, such as Australia, indicating that genetics isn’t the whole story. Nonetheless, heredity must be a significant factor, as it is clear that both heart health and one’s lifespan are significantly heritable (also here and here).
Heredity clearly isn’t the sole explanation, but it’s obviously a significant part of it. If only research into human differences didn’t become taboo, we may have had these answers long ago.