Do you want to brainstorm questions for your Kovler interview in a separate thread? Mine would be just about MODY-2, which is somewhat unique, I’m not sure whether you’re planning to address the other MODY flavors, some generalizations can presumably be made but many key things are unique to each.
1 Like
One could structure an entire interview around a core set of questions, which get at the heart of the matter:
- MODY-2 is the only form of diabetes for which the consensus recommendation is to do nothing to treat it. This recommendation appears to be based on two separate premises. One, MODY-2 is not very susceptible to treatment. And two, even if it were susceptible to treatment, MODY-2 entails little if any risk of complications, so it doesn’t need to be treated. But how robust is the evidentiary basis for these premises? How significant are the uncertainties?
There are some additional topics I think it’d be fascinating to explore:
-
Elevated blood sugars appear to correlate with various health conditions, even well down into the non-diabetic end of the spectrum. Is the argument that elevated blood sugars in those with GCK-MODY mutations don’t have negative health repercussions rooted in assumptions about elevated sugars being markers of other, underlying health conditions, rather than elevated sugars playing a causal role in those health conditions? How robust is the science about these issues based on the far larger non-MODY populations?
-
If GCK-MODY mutation carriers really have the same levels of complications as familial relations who lack the mutation, despite the former having meaningfully higher blood sugars than the latter, this suggests that there may actually be something protective about the GCK-MODY condition. Is that correct and what might that mechanism be?
-
Various GCK mutations can lead to varying severities of MODY-2. How well do we understand the resulting spectrum, and how consequential is it for risk assessment and treatment?
-
GCK-MODY mutation carriers appear to have the same risk of more traditional Type 1 and Type 2 diabetes as the general population. Many GCK-MODY carriers are likely to manifest at least some degree of “metabolic syndrome,” and associated insulin resistance, just as the general population does. How well do we understand how GCK-MODY may interact with that condition?
1 Like
I would love to brainstorm. I’m tied up right now, but I love the questions. When I get back maybe we could start a new topic to talk about it. I think it would be approprate to talk about all MODY.
1 Like
Sounds great. If you start a new thread to brainstorm questions, maybe paste my questions into it, or I can do that.
I obviously have a selfish interest in MODY-2, and it’s unique (only form of diabetes for which consensus–warranted or not–is no treatment whatsoever). But one could still ask interesting questions about MODY-2 in the context of a broader MODY or, better, monogenetic, interview.
1 Like
Brian,
To reply to your comment up-thread in the post with the link to the study: Yes, you came up with exactly what I saw. The people with MODY-2 diagnoses I heard from reported deaths from heart attacks in the late 30s of family members. These would not have survived to make it into this study.
The other HUGE issue here is that the controls had a median A1c of 5.8%. This is above the 5.7% my endocrinologist told me she considered to be abnormal. And because that was a MEDIAN, half the “controls” were likely to have had very poor blood sugar control.
I just recently heard from a person apparently in the UK who was told her blood sugars were completely normal though she is seeing readings well over 200 mg/dl at meals every day. So this study suggests that doctors in the UK are not diagnosing people until their diabetes is far more severe than what is diagnosed here in the U.S…
Those “normal” controls have A1cs high enough to have already given them a significantly increased risk of cardiovascular disease over people with truly normal A1cs, especially since with that median A1c there are probably a lot of people in that control group with A1cs just under 6.5%–a level that if I recall correctly doubles cardiovascular risk compared to 5% according to the EPIC-Norfolk data.
And that 30% incidence of retinopathy sticks out like a very sore thumb, too.
The UK NHS rather like Kaiser in the U.S. appears to be bound and determined to do everything it can to reduce the costs of treating people with diabetes by redefining them as not having it, rather than treating it. I would not base any decisions about my health on this particular study.
The pressure to cut costs in diabetes care may well end up making insurers take away resources from anyone who’s A1c isn’t 7% which might be people with blood sugars nearing 300 at every meal. Kaiser makes it very hard to get a diagnosis–I didn’t get one until they went out of business in the NorthEast and my doctor could finally diagnose me so I could get coverage for strips. Apparently readings of 250 mg/dl were normal according to Kaiser as they appear to be in the UK.
1 Like
There’s a lot of “patient-centered health care” rhetoric bouncing around these days at various diabetes conferences and meetings. How can anyone defend this bias to withhold diagnosis based on expense?
“Patient-centered healthcare” is window dressing b-s used to disguise cost-cutting efforts.
So for example, they will come up with the idea that testing blood sugar is unpleasant (patient unfriendly!) and point to the poorly conducted studies where people were told to test their blood sugar fasting and, surprise, surprise, that kind of testing didn’t lower their blood sugar. They put this together and come up with the conclusion that it is patient-friendly not to prescribe test strips.
Every time I think I’ve seen them go as low as they can go, they drop further. I am seeing a lot of stuff dribbling in that is saying it is not worth lowering blood sugar even to 7% for older people. And these conclusions are always drawn on the basis of such horribly badly designed research. But most diabetes research is horribly badly designed.
Since there’s a ton of money in health care, it naturally tends toward corruption. Corporations are amoral profit seeking enterprises. They don’t try to hide it but do seem to want to disguise it a bit for the unwashed masses, the consumers. They can do good, too, but their laser focus is on the money.
@Jenny, this paragraph reflects a fundamental misunderstanding of what studies like this are trying to do: “The other HUGE issue here is that the controls had a median A1c of 5.8%. This is above the 5.7% my endocrinologist told me she considered to be abnormal. And because that was a MEDIAN, half the “controls” were likely to have had very poor blood sugar control.”
I tried to explain this above in this paragraph: “Yes, these folks, both the gene mutation carriers and the non-gene mutation carriers, have a spectrum of blood sugar levels, some of which we’d categorize as pre-diabetic. But that’s the point. What the study is doing is a “natural experiment” where family members who are presumably otherwise similar (other genes, lifestyle, diet, etc.) but who differ on one crucial variable (whether they have the MODY-2 gene mutation or not) are being compared. And the study strongly suggests that having the MODY-2 gene mutation doesn’t increase heart disease risks, despite the life-long elevated blood sugars that gene mutation entails. In fact, if there is a difference in the population, if anything the gene mutation may lower the risk of heart disease. The prevalence of heart disease is too low for the comparison to be statistically significant at these sample sizes, i.e. we can’t be sure that gene mutation carriers have a lower, or an identical, risk of heart disease as non-gene mutation carriers, controlling (via natural experiment) for other relevant factors, but it’s hard not to be at least somewhat reassured by the results.”
Our put differently, the study is trying to isolate the effect of one variable, which is whether an individual has a GCK-MODY gene mutation. So comparing those with and without the gene mutation, and who are otherwise similar (because they’re family members) is what’s relevant. Your idea, that we should compare GCK-MODY gene mutation populations with a subgroup selected to have optimal blood sugars, makes no sense if what we’re trying to isolate is the effect of the GCK mutation. In other words, some of the people who do and don’t have the mutation are overweight, and some eat terrible diets, and some have other genetically-mediated conditions, and we want all that variation in the study because we’re only interested in the effects of the GCK mutation. Your proposed study would only make sense if the hypothesis was that GCK mutations were so protective that GCK people didn’t have to worry about being overweight, or eating crummy food, or getting no exercise, or carrying other genetic risk factors, etc., but that would be a very bizarre hypothesis to postulate.
Does that make more sense? This is pretty fundamental if you want to engage critically with any of the medical studies you’re trying to evaluate.
I may have missed it, @Brian_BSC, but did you get your MODY lab results back yet?
Sorry reading everything on an iphone these days it’s hard to keep up with these conversations.
Not yet. I have one more MODY blog post from the ADA then my plan is to first post my experience asking and eventually being granted the test. Finally after getting the results back, I’ll have a blog post about the results and my feelings about them. So don’t worry you won’t miss it.
Niccolo,
I fully understand how and why the study was set up the way it was. No need to be condescending.
But I also fully understand the forces at play which are telling people with various forms of diabetes that they don’t have to do anything about them. And I have seen what that advice does to people in the real world, and it isn’t pretty.
Obviously, you have decided for yourself that this one study–the details of the data from which are hidden from us–are enough to reassure yourself that you can ignore your A1c.
But I have spent more than 15 years looking more closely at diabetes studies, and what I’ve learned is that they are often very poorly conducted and that the data is often badly analyzed, because doctors are very bad at understanding statistics and even, sadly, at thinking logically enough to eliminate other explanations for the numbers that come out of software they use.
I don’t, for example, see any analysis of important cofactors. Did the people with what they had been told were diabetic blood sugars stop smoking in greater numbers than those family members who thought they had no reason for concern? Were they less sedentary? Either of these might explain the not-very significant heart disease differences without attributing them to the gene.
And to further complicate matters, all we are given here is A1c, my least favorite statistic. I would be very curious to see how high the post-meal blood sugars of all the participants in this study rose. This is because there is some interesting data that finds that ONLY post-meal blood sugars correlate in a significant way with complications, not fasting or A1c if post-meal sugars come down under 140 mg/dl.
The GCK gene alone only affects fasting bg. But the post meal excursions people who have it experience are likely to vary greatly. So does the speed with which these excursions return to a baseline. And that baseline is different from person to person, too.
It is quite possible that enough people in this particular, statistically small, group of people with the mutation had versions that gave them post-meal numbers that didn’t stay high enough to cause complications because they had robust 2nd phase insulin. There should be a huge difference in the outcome of a person with a fbg of 125 who is at 125 2 hours after eating and one who has a fbg of 140 who is at 170 two hours after eating, based on all the research that connects post-meal blood sugar levels with complications.
But buy doctors are blissfully unaware of these kinds of complexities, because they only read the “practice pearls.”. So if you tell them that people with this gene can ignore it, they will not treat the people who are getting that 170, so they will develop retinopathy and very likely cardiovascular damage that will show up in that heart attack at age 57.
When any result from a small study contradicts the data from unrelated large studies (as is the case with A1c and cardiovascular health studies,) you have to look very carefully at the methodology, which is very tough to do because these studies are always published in summary form, even in the “full text” version. And too many times important questions, like what were the post meal numbers and how did they correlate to risks, are not asked.
@Jenny, I still think that suggesting a flaw of the study is that the control group wasn’t selected to have excellent blood sugars reflects a fundamental misunderstanding of what this study, and for that matter all studies like it, are intended to do. I’m sorry if that comes across as condescending.
I agree that some doctors seem to erroneously tell patients they don’t need to address their diabetes. More likely, many seem to be willing to compromise at levels higher than ones that what lots of us here think is both achievable and warranted.
The study we’re discussing has its limitations. It’s far from the only study regarding GCK-MODY, but it seems to be among the best on that topic. I certainly don’t take away from it the message that anyone, myself included, can ignore their A1c.
I agree there’s some shoddy science out there. But, at the risk of being condescending, the argument that “doctors are very bad at understanding statistics” seems a bit rich under the circumstances. In particular, I have a hard time buying the argument that both the Kovler and the Exeter researchers, the two centers of excellence for research on MODY, are both either inept or corrupt.
My assessment is rooted in part in the Kovler folks’ responses to my posing some of the same questions you are posing. But I continue to be interested in understanding these issues better, in part because they have potentially huge implications for my own health, but also because I’m intellectually fascinating by them.
I observe, consistent with other studies, that an intensive insulin regimen (which I did for 13 years) at most very modestly lowers my glucose levels, and it comes along with at least sporadic lows that have their own downsides. I would be very surprised if I went back on insulin in the near-term, absent the development of corresponding more traditional T1 or T2 dynamics on top of my GCK-MODY mutation.
Most oral meds also make little sense given my condition, e.g. I present with high insulin sensitivity. But if my numbers were to worsen down the road and/or depending on what drugs come down the pipeline, there are some interesting options that might help address the specific GCK mutation-mediated pathology I have.
Low carb can curb post-prandial spikes, and I agree with you it’s plausible, and there’s some evidence, that this is highly desirable. Low carb can also have some, albeit limited, effect on reducing my average blood sugars. I’ve never eaten a really high-carb diet, and certainly not since I was diagnosed as diabetic thirteen years ago. I will play around with diet, and I suspect I’ll end up at a low-ish but not super low place. I find severe low carb impacts my athletic performance, which is very important to me, though I wonder whether my body might adapt over time.
I do think you are asking some very good questions and making some good points. Obviously I also think some are less good, but I don’t have to agree with everyone you write to value your contribution, any more than you have to agree with everything I write. And even where I disagree with you, you’ve helped me better develop my own opinions, and I appreciate that.
My experience is that clinicians do not have a good understanding of the science and they are often arrogant about it. When I was at ADA I was just chatting with one MD and mentioned that I many of the studies depended heavily on statistics and I thought a medical education probably didn’t train MDs enough to understand the studies. I was promptly told that of course MDs were taught statistics, that she had a course in statistics and that was quite enough. I’ve taken a half dozen courses in the subject and I still at times struggle to make sure I interpret things correctly. My current endo, although she is highly qualified doesn’t understand the difference between absolute and relative risk.
I am sure there are many clinical physicians who are great scientists and understand science. But for the most part clinical physicians are not trained or competent scientists and while they can read a conclusion or an interpretation of a study, they are have no clue on how to critically evaluate a scientific paper.
And as to the low carb issue, I would like to ask Dr. Naylor as well as others why both Exeter and Kovler have backed off from the low carb recommendations. Just a few years ago Exeter was clearly recommending low carb for MODY-2 but I don’t really hear that anymore.
Yeah, stats illiteracy is scarily widespread, even among physicians.
It seems like a reach to extrapolate that to researchers at the University of Chicago or similar institutions, though. Which is not to suggest they get a blanket pass, but if there are flaws in their analysis, they’re probably more nuanced.
It was my impression that Exeter used to recommend low-carb and sulphonylureas for MODY-2, though perhaps I’m conflating that with another MODY subtype. So I concur that’s a great question to pose.
Even though MODY has no role to play in my diabetes, I am finding this discussion interesting and appreciate the civil tone adopted by the commenters. Thank you. 
1 Like
Well, when I had a discussion about the statistics used in some studies with one of my more approachable doctors, she laughed and said, “I had to take one statistics course and barely got through it!”
I had another discussion with a friend, who is a Ph.D. biologist with her own lab, where I mentioned the regrettable tendency to report statistically insignificant results as “trending” towards a certain conclusion–which is taken to mean that the result does point to a meaningful conclusion. My friend was very surprised, as she told me she always thought “trending” results were important. Not only that, but she did not understand the difference between statistically significant and the way that people use the term “significant” in conversation.
Statistics are brutally abused to convince doctors that things that make almost no difference in actual patient outcomes are important. The way that “risk” is used instead of incidence, is a good example. You can cut risk in half with an intervention that reduces incidence from 2 in a million to 1 in a million. But the impact of an intervention that does is is pretty low, given the very low incidence. If you were to market a drug as eliminating 1 case in a million people who took it, it wouldn’t look that impressive. But advertising it as cutting the risk of the condition in half, and doctors prescribe it.
I’m far from a wiz at statistics, very far. But I know enough to read the actual study and see how the final statistics were collected, and having read hundreds of diabetes and nutrition studies, I can confidently state that most of them are very poorly designed with very little understanding of the factors, other than the one the researcher would like to prove is doing something, that could also produce the same result.
It really goes beyond statistics. It is the lack of what I would call “debugging skills” which engineers learn but doctors don’t that is at fault. This is why severe mistakes of understanding survive for so long among physicians, and why so many people treating diabetes are still telling patients with diabetes that lowering A1c below 6.5% with diet alone is dangerous. (Something I keep hearing from correspondents.) Doctors read only the final paragraph–“those in the intensive intervention group had more heart attacks” and never ask themselves who was in the intervention group? (people taking Avandia), did they lower their blood sugar as the intervention was supposed to do (no, the people with the heart attacks had high A1cs.) Etc.
It’s the same kind of thinking that explains why many of us who have maintained normal A1cs for a decade or more go to doctors who never once ask us HOW we have managed to produce such low A1cs. This never fails to astonish me. As far as I can tell, most of my doctors assume I don’t really have diabetes. But it is the fact they never ask that gets to me.
3 Likes
I think we are in total agreement on this subject. If you haven’t read Richard Feinman’s book “The World Turned Upside Down” you should, I think you would get a kick out of it.
Reviving an old conversation.
I am now suspecting MODY of some type.
Why?
-
I have ‘mild’ diabetes, fairly easily controlled with insulin, and low carb diet. Diagnosed when I was 38 and newly pregnant.
-
My now 8 yo, has what now appears to be very slowly progressing diabetes. Her hba1c was 5.1 3 years ago and is now 5.9 (she is being assessed for Type 1, and is doing pretty well on low carb diet so far, but with bs over 200 if she eats starchy carbs, fruits, or sweets).
-
My nearly 5 yo just tested with hba1c of 5.5. Not terribly high, but not quite normal for a kid. I’ve done spot checks on her after eating meals with a few more carbs than typical and see numbers up to 180 so far.
-
Ms. 8’s new pediatric endo also brought up the idea of MODY, once she heard about me and my situation (she hasn’t yet heard about Ms. Nearly 5).
I hope this is “mody” of some type and more easily managed… time to become a geek and immerse myself in literature… a ‘good’ distraction from time to time…
I never tried a sulfonurea drug, but maybe given this possibility I should try it and see if it works…
Probability of getting to do mody testing is slim to none (and beyond my budget for sure)…
1 Like