In the Part (1), I explained the three primary markers we use to track blood sugar: fasting blood glucose (FBG), oral glucose tolerance test (OGTT) and hemoglobin A1c (A1c). We also looked at what the medical establishment considers as “normal” for these markers. The table below summarizes those values.
| Marker | Normal | Pre-diabetes | Diabetes |
| Fasting blood glucose (mg/dL) | <99 | 100-125 | >126 |
| OGGT / post-meal (mg/dL after 2 hours) | <140 | 140-199 | >200 |
| Hemoglobin A1c (%) | <6 | 6-6.4 | >6.4 |
In this article, we’re going to look at just how “normal” those normal levels are – according to the scientific literature. We’ll also consider which of these three markers is most important in preventing diabetes and cardiovascular disease. But before we do that, I’d like to make an important point: context is everything.
In my work with patients, I never use any single marker alone to determine whether someone has a blood sugar issue. I run a full blood panel that includes fasting glucose, A1c, fructosamine, uric acid and triglycerides (along with other lipids), and I also have them do post-meal testing at home over a period of 3 days with a range of foods.
If they have a few post-meal spikes and all other markers or normal, I’m not concerned. If their fasting BG, A1c and fructosamine are all elevated, and they’re having spikes, then I’m concerned and I will investigate further.
On a similar note, I’ve written that A1c is not a reliable marker for individuals [2] because of context: there are many non-blood sugar-related conditions that can make A1c appear high or low. So if someone is normal on all of the other blood sugar markers, but has high A1c, I’m usually not concerned.
With all of that said, let’s take a look at some of the research.
Fasting blood sugar
According to continuous glucose monitoring studies of healthy people, a normal fasting blood sugar is 83 mg/dL or less. Many normal people have fasting blood sugar in the mid-to-high 70s.
While most doctors will tell you that anything under 100 mg/dL is normal, it may not be. In this study [3], people with FBG levels above 95 had more than 3x the risk of developing future diabetes than people with FBG levels below 90. This study [4] showed progressively increasing risk of heart disease in men with FBG levels above 85 mg/dL, as compared to those with FBG levels of 81 mg/dL or lower.
What’s even more important to understand about FBG is that it’s the least sensitive marker for predicting future diabetes and heart disease. Several studies show that a “normal” FBG level in the mid-90s predicts diabetes diagnosed a decade later.
Far more important than a single fasting blood glucose reading is the number of hours a day our blood sugar spends elevated over the level known to cause complications, which is roughly 140 mg/dl (7.7 mmol/L). I’ll discuss this in more detail in the OGGT section.
One caveat here is that very low-carb diets will produce elevated fasting blood glucose levels. Why? Because low-carb diets induce insulin resistance. Restricting carbohydrates produces a natural drop in insulin levels, which in turn activates hormone sensitive lipase. Fat tissue is then broken down, and non-esterified fatty acids (a.k.a. “free fatty acids” or NEFA) are released into the bloodstream. These NEFA are taken up by the muscles, which use them as fuel. And since the muscle’s needs for fuel has been met, it decreases sensitivity to insulin. You can read more about this at Hyperlipid [5].
So, if you eat a low-carb diet and have borderline high FBG (i.e. 90-105), it may not be cause for concern. Your post-meal blood sugars and A1c levels are more important.
Hemoglobin A1c
In spite of what the American Diabetes Association (ADA) tells us, a truly normal A1c is between 4.6% and 5.3%.
But while A1c is a good way to measure blood sugar in large population studies, it’s not as accurate for individuals. An A1c of 5.1% maps to an average blood sugar of about 100 mg/dL. But some people’s A1c results are always a little higher than their FBG and OGTT numbers would predict, and other people’s are always a little lower.
This is probably due to the fact that several factors can influence red blood cells. Remember, A1c is a measure of how much hemoglobin in red blood cells is bonded (glycated) to glucose. Anything that affects red blood cells and hemoglobin – such as anemia, dehydration and genetic disorders – will skew A1c results.
A number of studies show that A1c levels below the diabetic range are associated with cardiovascular disease. This study showed that A1c levels lower than 5% had the lowest rates of cardiovascular disease [6] (CVD) and that a 1% increase (to 6%) significantly increased CVD risk. Another study showed an even tighter correlation between A1c and CVD, indicating a linear increase in CVD as A1c rose above 4.6% [7] – a level that corresponds to a fasting blood glucose of just 86 mg/dL. Finally, this study showed that the risk of heart disease in people without diabetes doubles for every percentage point increase above 4.6% [8].
Studies also consistently show that A1c levels considered “normal” by the ADA fail to predict future diabetes. This study [9] found that using the ADA criteria of an A1c of 6% as normal missed 70% of individuals with diabetes, 71-84% with dysglycemia, and 82-94% with pre-diabetes. How’s that for accuracy?
What we’ve learned so far, then, is that the fasting blood glucose and A1c levels recommended by the ADA are not reliable cut-offs for predicting or preventing future diabetes and heart disease. This is problematic, to say the least, because the A1c and FBG are the only glucose tests the vast majority of people get from their doctors.
OGTT / post-meal blood sugars
If you recall, the oral glucose tolerance test (OGTT) measures how our blood sugar responds to drinking a challenge solution of 75 grams of glucose. I don’t recommend this test, because A) it’s not realistic (no one every drinks 75 grams of pure glucose), and B) it can produce horrible side effects for people with poor glucose control.
However, there’s another more realistic and convenient way to achieve a similar measurement, and that is simply using a glucometer to test your blood sugar one and two hours after you eat a meal. This is called post-prandial (post-meal) blood sugar testing. As we go through this section, the numbers I use apply to both OGTT and post-meal testing.
As the table at the beginning of this article indicates, the ADA considers OGTT of between 140 – 199 two hours after the challenge to be pre-diabetic, and levels above 200 to be diabetic.
But once again, continuous glucose monitoring studies suggest that the ADA levels are far too high. Most people’s blood sugar drops below 120 mg/dL two hours after a meal, and many healthy people drop below 100 mg/dL or return to baseline.
This study showed that even after a high-carb meal, normal people’s blood sugar rises to about 125 mg/dL for a brief period, with the peak blood sugar being measured at 45 minutes after eating, and then drops back under 100 mg/dL by the two hour mark [10].
Another continuous glucose monitoring study confirmed these results. Sensor glucose concentrations were between 71 – 120 mg/dL for 91% of the day [11]. Sensor values were less than or equal to 60 or 140 mg/dL for only 0.2% and 0.4% of the day, respectively.
On the other hand, some studies suggest [12] that even healthy people with no known blood sugar problems can experience post-meal spikes above 140 mg/dL at one hour. As I said in the beginning of the article, context is everything and all of the markers for blood sugar must be interpreted together.
If post-meal blood sugars do rise above 140 mg/dL and stay there for a significant period of time, the consequences are severe. Prolonged exposure to blood sugars above 140 mg/dL causes irreversible beta cell loss [13] (the beta cells produce insulin) and nerve damage [14]. 1 in 2 “pre-diabetics” get retinopathy [15], a serious diabetic complication. Cancer rates increase [16] as post-meal blood sugars rise above 160 mg/dL. This study showed stroke risk increased by 25% [16] for every 18 mg/dL rise in post-meal blood sugars. Finally, 1-hour OGTT readings above 155 mg/dL correlate strongly with increased CVD risk [17].
What does it all mean?
Let’s take a look again at what the ADA thinks is “normal” blood sugar:
| Marker | Normal | Pre-diabetes | Diabetes |
| Fasting blood glucose (mg/dL) | <99 | 100-125 | >126 |
| OGGT / post-meal (mg/dL after 2 hours) | <140 | 140-199 | >200 |
| Hemoglobin A1c (%) | <6 | 6-6.4 | >6.4 |
But as we’ve seen in this article, these levels depend highly on context and whether all markers are elevated, or just a few of them.
If you’re interested in health and longevity – instead of just slowing the onset of serious disease by a few years – you might consider shooting for these targets. But remember to interpret the numbers together, and also remember that blood sugar is highly variable. If you wake up one morning and have a fasting blood sugar of 95, but your A1c and post-meal numbers are still normal, that’s usually no cause for concern. Likewise, if you see a one-hour post-meal spike of 145 mg/dL, but all of your other numbers are normal, that is also usually no cause for concern.
| Marker | Ideal |
| Fasting blood glucose (mg/dL) | <86* |
| OGGT / post-meal (mg/dL after 2 hours) | <120 |
| Hemoglobin A1c (%) | <5.3 |
*If you’re following a low-carb diet, fasting blood sugars in the 90s and even low 100s may not be a problem, provided your A1c and post-meal blood sugars are within the normal range.
Another key takeaway from this article is that fasting blood glucose and A1 are not often reliable for predicting diabetes or CVD risk. Post-meal blood sugars are a more accurate marker for this purpose. And the good news is that this can be done cheaply, safely and conveniently at home, without a doctor’s order and without subjecting yourself to the brutality of an OGTT.
I’ll describe exactly how to do this in the next article.
