This study looked at roughly 14,000 NHANES III participants that were not diabetic. I think this is an important distinction because it looks at differences in normal range, "free", if you will, of what the frank hyperglycemia associated with metabolic dysfunction might do to mean and/or median values. Here is the hazard ratio plot:
To repeat from the caption -- HbA1c of 4.3% corresponds with the 2.5th percentile meaning only 2.5% of non-diabetics have levels this low (this percentage would be even lower for the entire population including diabetics). I would note that the 50th percentile (aka the median) is at 5.3%, but you go all the way up to the 90th percentile and reach a whopping 5.9%. Conversely you go all the way down to the 10th percentile at only 4.7%. It looks like the distribution is pretty uniform so I estimated the other major percentiles below (exact values shaded light gray):
The middle 80% of non-diabetics have a range of HbA1c from 4.7 to 5.9% ... a difference of a mere 1.2% ... the interquartile range is a mere 3/4ths of a percentage point!! The first reality about HbA1c, therefore, is that there isn't much difference in the normal range.
... Participants with a low HbA1c (< 4.0%) had the highest levels of mean red blood cell volume, ferritin, and liver enzymes and the lowest levels of mean total cholesterol and diastolic blood pressure compared with their counterparts with HbA1c levels between 4.0% and 6.4%. An HbA1c < 4.0% versus 5.0% to 5.4% was associated with an increased risk of all-cause mortality (HR, 3.73; 95% CI, 1.45 to 9.63) after adjustment for age, race-ethnicity, and sex. This association was attenuated but remained statistically significant after further multivariable adjustment for lifestyle, cardiovascular factors, metabolic factors, red blood cell indices, iron storage indices, and liver function indices (HR, 2.90; 95% CI, 1.25 to 6.76).
Here's the mortality rate and hazard ratios (HR's)... The range of HbA1c values was 2.8% to 6.4%, and 97 participants had an HbA1c 4.0%. Participants with an HbA1c between 4.0% and 4.4% were the youngest; those with an HbA1c of 6.0 to 6.4% were the oldest (Table 1). A greater proportion of participants with HbA1c 5.0% were women. J-shaped or U-shaped associations were present for age, nonHispanic black race-ethnicity, less than a high school education, and annual household income $20 000. Mean cell volume, ferritin, transferrin saturation, and liver function measures were elevated in the HbA1c 4.0% group, whereas mean systolic blood pressure, diastolic blood pressure, total cholesterol, and median triglyceride levels were lowest among those with an HbA1c 4.0%. Also, the proportion of participants with a history of CVD disease was lowest among those with an HbA1c 4.0%, whereas the proportion of participants with hepatitis C was the highest in this group. ...
So HbA1c under 4 has HR twice that of the diabetics in this other study. Those with HbA1c between 4 and 4.5 had similar HR's to those with HbA1c's between 6.0 and 6.4.
If I had a very low HbA1c, I'd spend a bit more time looking at the different models, but note that the 97 with HbA1c < 4.0 represented only ~ 0.7% of people. This is rare. And it's associated with a significantly higher risk of all cause mortality compared to median HbA1c of 5.0-5.4%.
That low an HbA1c is an anomaly. Some seem to believe that the lower the better, right? Might brush off these HRs and mortality rates. Well, have a gander at the table below (this is from a study I discussed in considerable detail here). The numbers in parentheses in the population line represent the percent of the population in the study. Of particular note was that we don't see increased mortality (all cause or stroke) for the bottom 90% despite a significant range in postprandial glucose levels (note that the absolute values here do not translate well due to differences in analysis and glucose challenge). They expanded the top 10% across the right four columns and even then we're talking ~1.15, 1.3 and 1.6 for hazard ratios.
Let's not forget this study that found lowest all-cause mortality centered about a fasting glucose of 99 mg/dL. The plot at right might look familiar, those dots show the actual death rates. Below is the table of HR's for fasting glucose of 7.8 (140 mg/dL) or 7.0 (126 mg/dL) vs. 6.0 mmol (108 mg/dL) as reference, or 2 hour OGTT of 11.1 (200 mg/dL) vs. 7.7 (~140 mg/dL) as reference. Even compared to the higher unadjusted HR's, that HR for HbA1c below 4% is not looking so hot.
So what brought this up? Well, Fred Hahn asked a question on FB the other day, basically how a diabetic that resolves BG with a VLC diet can still be diabetic. This is somewhat of a topic (perhaps) for another day -- I've actually covered it quite a bit here. Basically I brought up pancreatic beta cell function and that basal hyperinsulinemia and the postprandial GSIS are not related (as I've discussed on here quite a bit. Some 800+ comments later, playing the "diabetes is a disease of glucose intolerance" card, along with the addict and Inuit cards, and Fred started another post:
...The one thing I found quite interesting in that thread was Evelyn suggesting that even though a low carb diet will help in the short term, the problem is finding a way to make the T2D able to handle carbohydrate. She said:
"It is a disease of relative insulin deficiency. Managing the hyperglycemia by lowering dietary carb simply manages the symptoms but does not fix the problem."
What problem is Evelyn talking about? The high carb eating T2D has *relative* insulin deficiency. Should the alcoholic try and find a way for his body to metabolize alcohol better so he can drink like every body else? Or should he stop drinking? Should the parents of a child with a severe peanut allergy try and find a way that their child can eat peanuts?
Why would a person need to find a way to be able to tolerate a large amount of carbohydrate in the diet? For what purpose?
Yeah, carbs, just like aflatoxin and alcohol ... but I digress. Dana Carpender entered the discussion: "The question is not are a low carb diet and exercise the only course of action that would benefit the patient, the question is do they offer the *most* benefit with the fewest risks." Now this gets frustrating because Dana is among the long term low carbers that have developed metabolic issues of late, and she does sport a list of health maladies that is not inconsequential. Though they can't all be blamed on her diet, there's simply no acknowledgement that her 75% on up fat diet that she aims might be a contributing factor. But her doctor was "startled and dazzled by my blood work." Not that there has ever been any proven benefit to an HDL over 100 or very low fasting triglycerides other than normal levels. She (and her doctor apparently) attribute her prior need for Metformin and Victoza and elevated FBG to gluconeogenesis from excess dietary protein .... which she has "fixed" by cutting that
chocolate cake* protein down to 60-80 g/day (likely sufficient but at 20-30g carb per day that includes veggie carbs I wonder if lean mass will suffer over the long run and only she knows). (*a reference to Jimmy equating chocolate cake and steak in lectures)
But here was the boast that caught my eye:
And the impetus for me exhuming this study from my Mendeley files and publishing this up. I'm not sure I'd celebrate, but what do I know. Her low carb doctors tell her everything's stellar ... I'm sure they'll find something magical about ketones that renders this irrelevant.
And here is where these prospective studies and the FACT that there are no ... zero, zilch, none ever -- human populations consuming a diet remotely like hers or any of these other low carbers becomes important. We only have their reports to go by, and they aren't even encouraging. And yet they still promote this diet as healthy. And the doctors too. It's shameful, really.
As a dietary intervention LC has a great place. It's promoters harm that reputation by claiming the diet is superior or mandated for even a portion of the population. As a therapeutic measure it is generally used only in the short term, with supplementation I would add to prevent things like kidney stones that are otherwise extremely rare in the target population (e.g. epileptic children). That there are no known side effects other than constipation but let's compare to failed diabetes drugs and ignore that we have no long term data on LC diets (and the Inuit don't count (unless and until a new version centered on seal meat is implemented and promoted.)