The sugar hypothesis

I’m always amazed at how the lipid hypothesis of heart disease has wormed its way deep into the souls of physicians, scientists, and medical/nutritional writers the world over. The flimsiest piece of research appearing to confirm this entrenched bias is not only accepted uncritically (not by yours truly, of course), but is shouted from the rooftops by the likes of USA Today, the New York Times, Reuters, and other general publications.

Then along comes a study with some true value that, although published in a prestigious journal, is vigorously ignored. Such seems to be the fate of a tremendous piece of research appearing in the current issue of Diabetes Care. Since this journal, like JAMA, makes available to the public the full text of articles its editors deem of significant public health importance, you can get the article in its entirety here.
Researchers at University College in London looked at the mortality data generated over the past 33 years from the Whitehall Study, which, like the Women’s Health Initiative, is a large, government funded study. In the Whitehall Study

18,403 nonindustrial London-based male civil servants aged 40-64 years were examined between September 1967 and January 1970. In brief, measurements included height, weight, blood pressure, six-lead electrocardiogram, lung function (forced expiratory volume in 1 s and forced vital capacity), plasma cholesterol concentration, and a glucose tolerance test. A self-administered questionnaire was completed regarding employment grade, smoking habits, health status, and physical activity.

After an overnight fast the participants of this study underwent a glucose tolerance test in which they drank a preparation containing 50 grams of glucose. Two hours later they had their blood sugars measured. In a normal glucose tolerance test blood sugars typically return to normal within two hours after having consumed a glucose drink. The researchers wanted to record these 2-hr post glucose load figures and see how they correlated with mortality over the ensuing years.
The researchers divided the study participants according to their 2-hr blood glucose values into three groups. Those with 2-hr glucose levels below 95 mg/dl were considered normal; those who measured 96-199 mg/dl were deemed glucose intolerant; and those with 2-hr blood glucose levels at 200 mg/dl or greater were classified as diabetic.

As might be expected, after 33 years the diabetic group fared poorly compared to the normal and even the glucose intolerant group. Having a diabetic 2-hr post load glucose profile increased the odds of all-cause mortality by 2.37 and of death from coronary heart disease by 3.70. Clearly having an abnormal glucose tolerance test is bad news for prospects of a long life, much more so, in fact, than a “bad” cholesterol test.

But, the most interesting aspect of this study is what the researchers found in the normal group first at 10 years then at 33 years down the road. Everyone knows that a markedly elevated blood sugar level or a diabetic glucose tolerance test bodes poorly for long term mortality, but what about the lower end of the curve? Is there a difference there? Is there a difference in long term mortality between a blood glucose level of, say, 95 mg/dl and one of 85 mg/dl, both of which are considered normal.
At the 10 year followup of the 18, 403 men the researchers found heart disease mortality started to increase at blood sugar levels above 95 mg/dl, but that those subjects with blood sugar levels below 95 mg/dl showed no increased risk for death from heart disease.

After 33 years, however, the picture changes. The cutoff level drops to 83 mg/dl. In other words, in terms of cardiovascular mortality, the risk starts to rise as 2-hr post load blood sugars reach 83 mg/dl and there is a linear increase in risk between 83 mg/dl and 95 mg/dl. I’m sure that after 40 years, the minimum level will drop a few points further.

Why this increase in risk? According to the authors:

Several mechanisms may account for the link between elevated glucose level and CHD risk. A raised glucose level at baseline may indicate emerging insulin resistance and a downward trajectory in glycemic control, with increased risk of glucose intolerance, diabetes, and CHD in subsequent years. Related risk factors, particularly hypertension and an atherogenic lipoprotein profile [even these folks are in the grip of the lipid hypothesis], tend to develop in parallel with insulin resistance. Other pathways include oxidative stress and formation of advanced glycation end products that accelerate atherosclerosis when blood glucose is only slightly raised. A further indirect mechanism involves common antecedents of hyperglycemia and CHD such as poor early growth. Adjusting for conventional risk factors, existing baseline CHD and socioeconomic position in the present study reduced the size of the glucose-CHD association by 45%. This is evidence of a causal role for glycemia in CHD etiology. It is also consistent with an element of confounding and with risk-factor clustering of the metabolic syndrome type.

What needs to be grasped about this study is that there is no comparable study showing mortality related to cholesterol levels, despite all the ignorant blather to the contrary. In fact, in this very Whitehall Study researchers measured cholesterol levels and found no significant differences between them in those subjects with normal blood sugar, those who were glucose intolerant, and those with outright diabetes (see Table 1 in the study).

This is all important stuff. It’s much more important for your long term health to work on keeping your blood sugar down than it is to work to keep your cholesterol down. It’s especially important when you realize that most people try to keep their cholesterol levels at bay be consuming a low-fat, high-carbohydrate diet, which is a diet containing, in many cases, a cup and a half to two cups of sugar equivalents per day. Now, remember, a normal blood sugar equates to only a teaspoon of glucose dissolved in the blood. So, it shouldn’t take a rocket scientist to realize that a couple of cups of sugar added to this teaspoon will rapidly run blood sugar levels up, up, up. This misguided effort to focus on cholesterol levels can only end up increasing the risk for early mortality from heart disease, the very thing reducing cholesterol is supposed to prevent.

I’ve got to point out that these studies are observational studies and the correlations shown do not necessarily mean causation.  The higher blood sugars are correlated to increased rates of death from heart disease.  This kind of evidence doesn’t prove the increased blood sugar levels are causative.  But there are observational studies and then there are observational studies.  Although observational studies can only show correlation, the stronger that correlation is, especially if it is combined with a plausible mechanism, the more likely the correlation may indeed be an indication of causation.  A famous correlation with no plausible mechanism at all was the Super Bowl victor and the stock market.  For 25 or 30 years if a team from the original NFL won the Super Bowl the stock market went up that year, and when a team from the original AFL won it, the market dropped.  There was indeed a strong statistical correlation between the two events but absolutely no credible mechanism for the Super Bowl winner to have anything whatsoever to do with the direction of the stock market.  Let’s look at an obvious example of a plausible mechanism and a correlation.  Excessive alcohol consumption is correlated with increased risk of death from an automobile accident.  Alcohol impairs mental and motor function, and a fair amount of mental and motor function are needed to safely drive a car.  This would appear to be a no brainer, but there is probably less statistical correlation between drinking and dying in a car wreck than there is between the Super Bowl winners and the stock market.

Several mechanisms may account for the link between elevated glucose level and CHD risk. A raised glucose level at baseline may indicate emerging insulin resistance and a downward trajectory in glycemic control, with increased risk of glucose intolerance, diabetes, and CHD in subsequent years. Related risk factors, particularly hypertension and an atherogenic lipoprotein profile, tend to develop in parallel with insulin resistance. Other pathways include oxidative stress and formation of advanced glycation end products that accelerate atherosclerosis when blood glucose is only slightly raised. A further indirect mechanism involves common antecedents of hyperglycemia and CHD such as poor early growth. Adjusting for conventional risk factors, existing baseline CHD and socioeconomic position in the present study reduced the size of the glucose-CHD association by 45%. This is evidence of a causal role for glycemia in CHD etiology. It is also consistent with an element of confounding and with risk-factor clustering of the metabolic syndrome type.

The maddening thing about all this is that you will probably never read about this tremendously important study in your local paper. But the next time some boob publishes anything that points a finger at cholesterol or, God forbid, fat in the diet, you will read all about it.
Oh, by the way, what’s the best way to keep your blood sugar low? How about by not eating a lot of it and letting the body make the sugar as it needs it. In my clinical experience, most patients on low-carbohydrate diets reduce their blood sugar levels into the 75-85 mg/dl range.