Comment by AlbertCory | Hacker News Reader

Comment by AlbertCory | Hacker News Reader

> Over 100 million Americans have diabetes or prediabetes

These two are not the same. It is dishonest to combine the two.

https://www.chicagotribune.com/2016/07/29/prediabetes-the-ep...

I'm not quite ready to publish my hypothesis that prediabetes is a scam. I need to research it a little more. But a statistician friend of mine responded:

A handful of years ago, I looked into the National Health and Nutrition Evaluation Survey, NHANES. To first order, the 1AC level defining pre-diabetes, 6.5, is rather close to the median level. So that's scam-adjacent. Every once in a while, my doctor thanks me for giving him this NHANES table.

https://docs.google.com/spreadsheets/d/1g3Icgu0ixLtYCjscYoiC...

As soon as you say this, someone will respond "My father has diabetes and he had his foot amputated!" This is not minimizing diabetes; it's questioning whether prediabetes is a thing.

How is it different from saying "Men over 60 have pre-prostate cancer?" Or, "we all have pre-death?"

You should have a regular blood test and your doctor should be monitoring a lot of things, including blood sugar. If the level is close to diabetes, he or she should warn you. But that's different from saying, "you have a disease."

I have no credentials here, but I'm often working in this space and your take is the polar opposite of what I normally hear from the endos and diabetes researchers that I've worked with. More accurate terms might be diabetes and morbid diabetes.

That the cutoff for prediabetes is close to the median level is a statement that much of the population is actually unhealthy in this regard.

No, it's a statement that the "cutoff" was chosen on questionable grounds. Defend it.

Did you check the litterature?

https://pubmed.ncbi.nlm.nih.gov/20697688/

Hba1c 6.0 to 6.5 (so called prediabetic range) has 2.5 times the hasard ratio of below 6.0.

Increasing Hba1c is the leading indicator of future health issues and by far.

This comes closer to a cogent answer, which is what I was after.

> future health issues

like?

As I said, I'm not ready to publish this, but I wanted to hear some arguments that made sense. So thanks.

To me, "probability of progression from pre-diabetes to diabetes" is the only reason to say it's a real condition rather than just a risk factor.

Furthermore, it would be the derivative of the level, rather than the level itself.

And finally, it would have to be separated from other unhealthy conditions. Meaning, if they drink too much AND have pre-diabetic A1C, ERROR! Otherwise, how do you know it's not the alcohol?

Get peer review and publish; let’s see what happens.

With all due respect, if you're "working on a paper" on this, but the hazard ratios for differing HbA1c values is a novel discussion point to you, I am wary of what else you "don't know".

According to what I understand from researches in this area is that even a slight increase in A1C is highly correlated with later progression. You may not like the term, but it effectively communicates the situation.

If you're going to try an appeal to authority, then at least come up with a plausible authority instead of some random statistician who likely doesn't know the basics of physiology. The first thing you need to understand is that normal ≠ healthy. For example, a resting heart rate of up to 100 bpm is considered "normal" in the sense that it doesn't require urgent medical intervention. But of course anything over 60 bpm for an adult usually indicates some underlying pathology with a risk of premature morbidity and mortality.

The essence of type 2 diabetes is insulin resistance. Like many medical conditions it exists on a spectrum. The specific HbA1c thresholds of 5.7% for pre-diabetes and 6.5% for diabetes are inherently arbitrary and serve mainly to make communication easier. But there is a clear correlation between elevated HbA1c levels and higher all-cause mortality.

https://doi.org/10.1136/bmjopen-2017-015949

Also note that HbA1c tests aren't perfect for diagnosing type 2 diabetes and can have false positives or false negatives. If there is reason to suspect diabetes then it would be wise to conduct additional tests to get a better understanding of the patient's metabolic condition.

https://peterattiamd.com/ama15/

> some random statistician who likely doesn't know the basics of physiology.

since you know nothing about him, that's hardly called for. I could doxx him but he didn't consent to that. And he said his doctor thanked him for it.

And you have fuzzed the difference between "a disease" and "something to watch for." If ~50% of the US population needs to be watched, the doctor learns nothing by having a label put on their HbA1c level. The word added nothing to their understanding.

> If there is reason to suspect diabetes then it would be wise to conduct additional tests to get a better understanding of the patient's metabolic condition.

And finally, you just confirmed what I said. It's not "a disease" -- it's a risk factor. Like smoking, drinking, obesity, or sedentariness.

The more everyone objects without any logical argument, the more it's confirmed:

"if that biomarker, all by itself, predicts type 2 diabetes better than random chance, in the absence of any other risk factor, we're entitled to call it a disease."

Labeling insulin resistance as a "risk factor" is correct only in a narrow, technical sense. That's not helpful for patients and clinicians who need to make treatment decisions in the real world. The terminology is irrelevant. Instead of getting hung up on semantics you need to spend some time learning the basics of human metabolism and the endocrine system.

Insulin resistance is a risk factor in the same way that a growing malignant tumor is a risk factor: unless you expect to die soon from something else you should take urgent measures to fix it. While a HbA1c test isn't perfect, it's pretty good as a cheap and easy initial screening for insulin resistance. If you want to believe that it's some kind of "scam" then I don't know how to help you.

The average US person today has at least some level of metabolic problems. Playing games with statistics won't change that reality or the negative long-term outcomes.

> Also note that HbA1c tests aren't perfect for diagnosing type 2 diabetes and can > have false positives or false negatives.

When we do workups on pts suspected of diabetes, we use the following criteria.

- Iron deficiency anemia workup to confirm accuracy of HbA1C

- HbA1C ≥6.5%

- FPG ≥126mg/dL (7.0 mmol/L). Fasting is defined as no caloric intake for at least eight hours

- Two-hour plasma glucose ≥200mg/dL (11.1 mmol/L) during an oral glucose tolerance test

- In a patient with classic symptoms of hyperglycemia or hyperglycemic crisis, a random plasma glucose ≥200mg/dL (11.1 mmol/L)

I would like to redo the diagnostic criteria to include their BMI.

Excellent. Thanks.

Is an A1C level sufficient, all by itself, sufficient to classify them as "pre-diabetic" or are other symptoms required? Or is that even a thing?

Usually, yes.

However, some PCPs / endo will order up A1c as well as glucose challenge, depending on certain factors (eg, family history, BMI, symptoms). If I were you, order up the following:

- A1C - 12 hour Fasting blood glucose levels - Two-hour glucose tolerance test: this test will measure blood glucose levels before and after ingestion of 75 g of glucose solution (if the results show blood levels that fall between 140 mg/dL to 199 mg/dL, it is diagnostic of prediabetes. - A random plasma glucose test

Personally, if you have a family history of diabetes/obesity, you should get checked often. Some providers have been writing rx's of GLP-1's to aid in weight loss, which reduces the risk of developing T2DM.

> pre-diabetes [...] is rather close to the median level

but according to the NIH, 30.7% of Americans are overweight, and 42.4% are obese. as the median American is overweight, it doesn't seem a stretch to claim that the median American is also pre-diabetic? I don't know whether it's true or not, but your evidence seems a bit thin.

> it doesn't seem a stretch to claim that the median American is also pre-diabetic

No, it's your evidence that's thin. You seem to have started from the premise ("Americans are unhealthy") and derived a pre-diabetes level from that.

You've literally provided zero evidence beyond hearsay from an alleged statistician friend.

The prediabetes level is drawn from where you see a huge uptick in the risks for developing diabetes mellitus.

You want data? There is plenty of it out there. If you're gonna try to dispute the established understanding, you need to bring evidence or you are just wasting people's time.

According to the CDC, 38.0% of the adult US population is pre-diabetic. 11.6% of the total population have actual diabetes. It's very close to half.

https://www.cdc.gov/diabetes/data/statistics-report/index.ht...

I switched health providers and got a physical recently. One of my labs came back with "pre-diabetic" check marked. I asked my doctor about this and I was .1% into the metric. It could have been either side of the line. I'm very fit because I enjoy hiking and exploring.

I already avoid sugar and bread. I work out nearly every day. I'm very active and my BMI is on point. My blood sugar was fine. It was another metric that I can't recall at the moment - but it seemed weird. Regardless, a very obese nurse sat me down and gave me tips for changing my diet.

This was likely your A1C, or a metric of the last three months or so of your sugar levels.

Just because you feel healthy doesn’t necessarily mean you are healthy. You may have, for example, a pancreas condition. Or it may be nothing at all, and just something to track, which is why you never base any medical decisions off a single test.

Your last sentence is nonsense. People can provide exceptional (and correct) advice without necessarily following it themselves, and for all you know this nurse has spent the last year actively trying to improve their health.

It's strictly a question of his likelihood P for developing diabetes. If his P is the same as that of other people with lower A1C, then "prediabetes" is nonsense.

> Or it may be nothing at all, and just something to track, which is why you never base any medical decisions off a single test.

Reply on news.ycombinator.com