Science has a history of ignoring some of the most important and astonishing work for a long time, then coming to its senses and celebrating it. It is time now to celebrate the work of Dr Joseph Kraft MD, a pathologist now in his nineties. Dr Kraft is the author of “Diabetes Epidemic and You” which is still available, at least in hard copy. He defines what he calls “Diabetes in situ” which is borrowed from his cancer pathology background. He has written several good papers, none of which have ended up being cited that highly. A nice, free summary of his work is available here. Kraft has carried out more than 14,000 oral glucose tolerance tests over a few decades. Normally we measure the glucose response to drinking glucose. This response can tell us the degree to which we metabolize and remove glucose from our blood. Very important for diabetes diagnosis and other metabolic issues. Kraft’s test are different though. He is way more thorough than normal. First, rather than monitoring glucose for 2 hours post test, he monitors it for at least 5 hours. Second, he also measures insulin, as well as glucose, over the course of the test. From his test results and the other literature, as well as his pathology and direct autopsy observations, he concludes that:
- We may be able to diagnose diabetes much much earlier than we do
- Abnormal insulin levels (high) are directly and indirectly damaging to the vascular system, and therefore almost every organ in the body
- This high insulin (hyperinsulinemia) is a condition in its own right and really the causal mechanism behind most of the metabolic and chronic diseases we experience today.
He calls the abnormal insulin response, with normal glucose response “Diabetes In Situ” What is Diabetes In Situ? Here’s the deal – you go to your doctor. You get a series of tests to see how well you are functioning metabolically. This is a really important series of tests because metabolic functioning is what will determine almost all of our quantity and quality of life. I would go as far as to at least hypothesize (there’s enough evidence) that problems in carbohydrate metabolism are implicated in virtually every chronic disease from head to foot, and most of the organs in between. This includes Alzheimer’s and vascular dementia, peripheral vascular disease and everything in between, from common obesity to diabetes to fatty liver disease to cardiovascular disease…and the list goes on. The trouble is that your doctor will look for metabolic markers which are either flawed as predictors, or mostly are markers of end stage dysfunction. Things like elevated blood glucose, high blood pressure, vascular blockages (arterial sclerosis) etc. If your doctor is really interested in your health and suspects you aren’t doing that well, they might order an oral glucose tolerance test. Kraft shows us that there are five typical patterns of insulin response to the glucose which characterize disease state. As mentioned earlier, these are really what we should be considering in early identification of metabolic problems. Unfortunately, the glucose tolerance test is rarely run long enough, nor is insulin (key to this) measured concurrently. In fact, in New Zealand, insulin measurement can only be ordered and covered under the public hospital system by a handful of specialist endocrinologists. Insulin is not considered as useful in the diagnosis of early stage chronic disease. That’s a shame really because Dr Kraft shows it certainly is useful. Let’s look at Kraft’s five patterns….. Here’s what we consider a normal insulin response to a glucose bolus in an oral glucose tolerance test. Kraft calls this “Pattern 1”. Pattern 1 insulin peaks after a 75 or 100 g glucose load after 30 to 60 min at 50-70 units. Everything is almost back to baseline after 2 hours, certainly after 3 hours. This is what we regard as a normal insulin response to a glucose load. It’s also likely that slower release carbs (lower GI) like beans might provoke an even lower area under the curve for the same total CHO load.
Pattern 2 (below) is the first sign of insulin resistance, but things will likely look fine if you simply look at the glucose responses. We see a similar time to peak, but a much higher peak. Insulin hasn’t returned to close to baseline until 4 hours or so after the initial load. The area under the insulin curve is very large. The total stress on pancreatic beta cells is high simply because of the overall demand for insulin. Many (most) people with this pattern present normal glucose curves and are told they are doing “just fine”.
Pattern 3 (below) is a delayed and high peak. This is indicative of some beta cell function, but some functional loss. In essence we are seeing the beginning of beta cell burn out. People can eventually move glucose into their cells, but they take a long time to produce enough insulin.
Pattern 4 (below) is the same as Pattern 3 except the fasting level of insulin is much higher.
Pattern 5 (below) is beta cell failure. In other words, there is little or no ability to move the glucose into cells because the pancreas can’t produce insulin. Patients will need exogenous insulin injected at that point to get the same result. Interestingly, Kraft also shows that sometimes people on low carb diets can show this pattern temporarily. It’s not that they are unable to produce insulin, it’s just more likely that they can take up glucose into cells (eg, brain) without insulin in the first instance. These people return to Pattern 1 after a period on a higher carb diet. I think this is also consistent with the finding that people on LCHF diets typically have a temporary diagnosis of peripheral insulin resistance which is resolved after a large carb meal. It’s likely that the body adapts to preferentially shunt the available glucose to the brain first. So Pattern 5 with a low carb diet is OK, its jus the same physiology as Pattern 1 but adapted to low carb.
Take-home messages, implications for practice, and questions
- Dr Kraft was onto it decades ago – that we should measure glucose and insulin concurrently for an extended period in the glucose tolerance test to identify those people who are at risk early.
- This early detection is both ethical and necessary.
- Instead current best practice sends people away who are insulin resistant until they present with end stage disease markers. Only then will we take them seriously.
- Why haven’t we taken Dr Kraft’s excellent work more seriously?
Finally here is a link to a letter I received from Dr Kraft recently. He outlines is overall approach and ideas from his results of several thousand OGTTs over several decades. It’s interesting stuff – page 1 click here – page 2 click here
As usual, for all my blog thanks so much to Helen Kilding who tirelessly edits and corrects my terrible typing!