You don't get owt for nowt.....

[Docb]

Begiining to get a correlation between end of day blood BG and carbs consumed during the day. A lot more data is needed to make sure it is real and reproducible but it is beginning to look like I would need to get to well under 50g carbs a day to get back into a "normal" range.

On the basis that you don't get owt for nowt, can anybody point out the downsides (other than having to forgo some favourite foods) of a very low carb diet, especially if you are not overweight by normal standards?

 

[Eddy Edson]

Obvious comment, but it depends a lot on what you replace them with. Given that T2D is very much a cardiovascular disease, heart-healthiness should be a big factor in the choice, IMO. And making sure you get enough fibre.

 

[Eddy Edson]

You might also want to do some research into alternatives to an aggressively low-carb approach. Despite what the Internet says, it's not the only way or necessarily the best way of skinning the cat.

Eg: Even though you say that you're "normal" weight, losing 10kg+ might get you to "remission". Check out the "Newcastle" stuff and note that even though an 800 cal per day short-term intervention is usually associated with it, it isn't an essential component. Prof Roy Taylor's FAQ is a useful starting point: https://www.ncl.ac.uk/media/wwwncla...ecentre/files/2018 Diabetes reversal info.pdf

Anyway, it seems to have worked for me.

 

[Docb]

Eddy, the FAQ you referenced is very informative. If I lost 10kg+ then I would be clinically underweight and that would introduce a load of other issues. I'll keep digging away until I get a good balanced picture. Any more thoughts or info sources anybody?

 

[Eddy Edson]

Eddy, the FAQ you referenced is very informative. If I lost 10kg+ then I would be clinically underweight and that would introduce a load of other issues. I'll keep digging away until I get a good balanced picture. Any more thoughts or info sources anybody?

Understood on weight loss. (Is it pretty certain that you are T2, and not a misdiagnosed T1?)

On other strategies: There's the "high-carb low-fat" approach. I think this is the main hub for it: https://www.pcrm.org/health-topics/diabetes

Essentially vegan. I don't know much about it, but I think it's probably more than simply woo. Might be worth digging into a bit.

Very high levels of fibre are a major factor. FWIW, I think that's a bit interesting, personally, since I think that high fibre levels as well as the weight loss have been important in getting my BG under control. Would hate to have to give up nuts, avocados, olive oil, seeds etc though.

More broadly, this is an interesting recent study attempting to compare different dietary approaches for T2D control: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5871653/ Heroic attempt to squeeze results out of messy data, but worth a look. Generally, though, I'm not sure that it tells you anything much beyond the proposition that whatever is best for sustainable weight loss is best for T2D, at the population level.

 

[Drummer]

It was early in the 1970s when I started low carbing and the only downside has been the reaction of doctors and other HCPs, and to my regret I have caved in to the propaganda which is high carb low fat on numerous occasions right up until I was healthied into type two diabetes. Now I am feeling rather more bolshie about their advice and where they can stick it - and have been practicing a hard stare and 'I could not possibly comment' just in case I was offered the same advice as I got two years ago - that baked potato and beans was a good choice for lunch.

 

[Docb]

Thanks again Eddy, and in answer to your question about T1 or T2 I have been wondering about that myself. First diagnosed T2 10 years ago, and comfortably under contol until the end of last year when things changed a bit dramatically suggests to me that there something a bit atypical going on. Won't be content until I have an explanation and next step is to get the GP to come up with the idea that things should be looked at a bit more closely. Need to work on how to make that happen! GP is not experienced so maybe that should help.

 

[Docb]

Your ref www.nbic........ was quite interesting - at least on reading the anstract - and well summarised by your final comment. Its the best way of accounting for the observation that one diet is best for HBa1c reduction and another is best for fasting glucose glucose reduction.

 

[Eddy Edson]

Your ref www.nbic........ was quite interesting - at least on reading the anstract - and well summarised by your final comment. Its the best way of accounting for the observation that one diet is best for HBa1c reduction and another is best for fasting glucose glucose reduction.

The comparatively high fasting BG levels for very low carb diets is consistent with a lot of other data/anecdotes I've seen and seems to be related to insulin resistance, which low carb doesn't address very effectively by itself. So I see comments from low-carbers indicating big Dawn Phenomenon effects, which is apparently an insulin resistance thing. Then if weight reduces (and muscle mass increases), insulin resistance and so fasting BG also decrease.

 

[Sally W]

I’m trying the Gut friendly diet, which is lower carb but not aggressively low. It focuses On Mediterranean diet and eating right sort of veg, probiotics etc. I’ve found it helpful for weight loss and satisfying.

However, it does involve minimal processed foods, hence more cooking from scratch. I have a large freezer and do batch cooking. Highly recommend Dr Mosley’s Clever Gut Diet book.

 

[Docb]

The comparatively high fasting BG levels for very low carb diets is consistent with a lot of other data/anecdotes I've seen and seems to be related to insulin resistance, which low carb doesn't address very effectively by itself. So I see comments from low-carbers indicating big Dawn Phenomenon effects, which is apparently an insulin resistance thing. Then if weight reduces (and muscle mass increases), insulin resistance and so fasting BG also decrease.

I think I get that but what I really do get is that you share my instincts that the explanations lie in the numbers, the thoughts feelings, opinions and anecdotes are valuable but really only add the necessary colour. The diffiulty is that there are so many variables, some of which will be significant for most whilst others will be significant for a few. There is room for one of these researchers who pop up on here seeking anecdotal information to call out for people who collect data to share it to create an enormous data set. Then maybe the effects of some of the variables can be teased out and statistical methods used to sort out what applies to the broad population and what doesn't.

 

[trophywench]

That's true @Benny G - Alan Shanley has always said each of us has to attempt to become an expert at treating their OWN diabetes - but that doesn't make any of us expert at treating anyone else's. And that's true whether we're Ragnar Hanas, Alan S or Trophywench.

Diabetes consultants in hospital clinics get greater exposure to large numbers of diabetics in the wild than any of us - but the DSNs in clinics get to know folk in greater depth than the consultants in 'real life' situations.

Just an ongoing learning curve for the rest of our lives!

 

[Eddy Edson]

I think I get that but what I really do get is that you share my instincts that the explanations lie in the numbers, the thoughts feelings, opinions and anecdotes are valuable but really only add the necessary colour. The diffiulty is that there are so many variables, some of which will be significant for most whilst others will be significant for a few. There is room for one of these researchers who pop up on here seeking anecdotal information to call out for people who collect data to share it to create an enormous data set. Then maybe the effects of some of the variables can be teased out and statistical methods used to sort out what applies to the broad population and what doesn't.

There's just so much individual variability! Basing T2D dietary recommendations on population means is a losing game. Individual variability depends on any factors which aren't very well understood and/or easy to measure: beta cells, gut microbes, genes etc etc. Right now, there's really no alternative to self-testing if you want to optimise BG.

This is one of my favourite little illustrations: https://forum.diabetes.org.uk/boards/threads/grains-bg-and-gi.75848/#post-861024 There's no way you can come up with a really useful general recommendation for T2's about what kind of bread is best. That's about as basic as you can get ...

 

[Docb]

Having Diabetes removes us from the broad population. Having Type 1 puts me into an even smaller sub population. There are still smaller sub groups like type 3, or people with several comorbidities.
Outside of the lab controlled environment all we have is anecdotal evidence.

Don't agree with you there that all we have is anecdotal evidence. There is shed load of hard evidence about, the trouble is that it is scattered. Many diabetics will be BG monitoring and writing it down in a diary. Most will be in some category or other. Many will be recording carb intake or be on different diets. There are many different control regimes and all that is written down. Could probably think of some more. Pool it all together an you never know what you might find.

Its part of what I used to do albeit in an industrial context. A process produces a variable product and the variability needs to be reduced. Fred tells you to do one thing, Joe says something else, Arthur's got another idea and so on. The only thing for sure is that the plant manager will blame the quality of the feedstock. My approach was to gather all the numbers I could and then use the anecdotal stuff to guide the analysis. The ancecdotal stuff is valuable, there is always a truth in it somewhere. Might take a couple of months to trawl through the numbers but it was rare that something of real value did not turn up. I remember the instance when some fourier analysis showed a periodicity in the data which could be related to deliveries of feedstock. A case where the problem was firmly assigned to a particular part of the process which could not be changed so the variability had to be treated as a fact of life. Probably the most amusing was demonstrating that one shift (which contained all the deadbeats because it had a weak foreman) was responsible for nearly all the rejects.

So I suggest the data is there if only it could be got at and analysed. Not saying it is easy, might even be impossible, but it in principle it would be a way of getting to grips with what affects what in what circumstances in the world of diabetes. Yeh, and I know all about patient confidentiality.

 

[Docb]

Just looked at your reference Eddy and my comment is that they found nothing probably because they tried to do an experiment to test a hypothesis where there are a load of other variables over which they had no control or understanding. Not uncommon in experimental science. The point is that there might or might not be some merit in their hypothesis about bread but that experimental methodology won't enlighten you for the reasons you suggest. My instinct is to find a different methodology, not to stop because it is too hard.

 

[Docb]

Having Diabetes removes us from the broad population. Having Type 1 puts me into an even smaller sub population. There are still smaller sub groups like type 3, or people with several comorbidities.
Outside of the lab controlled environment all we have is anecdotal evidence.

Yup and as far as data goes you have got loads of it and it only applies your circumstances which will be well recorded. Manna to a data analyst.

 

[Docb]

That's true @Benny G - Alan Shanley has always said each of us has to attempt to become an expert at treating their OWN diabetes - but that doesn't make any of us expert at treating anyone else's. And that's true whether we're Ragnar Hanas, Alan S or Trophywench.

Diabetes consultants in hospital clinics get greater exposure to large numbers of diabetics in the wild than any of us - but the DSNs in clinics get to know folk in greater depth than the consultants in 'real life' situations.

Just an ongoing learning curve for the rest of our lives!

Absolutely true Jen but wouldn't it be nice to have a little handbook of cause and effect with some measure of the statistical significance gained from the collective experience and not from the prejudices of some "expert" in a white coat! Bet there isn't a consultant in the land who can do maths or uses a data analyst.

 

[Eddy Edson]

Just looked at your reference Eddy and my comment is that they found nothing probably because they tried to do an experiment to test a hypothesis where there are a load of other variables over which they had no control or understanding. Not uncommon in experimental science. The point is that there might or might not be some merit in their hypothesis about bread but that experimental methodology won't enlighten you for the reasons you suggest. My instinct is to find a different methodology, not to stop because it is too hard.

There's some interesting analytics work on genetic influences: eg https://forum.diabetes.org.uk/board...of-t2d-via-genetic-clustering-analysis.75345/

But when it comes down to simple questions like "What effect will this bread have on my BG?" it's hard to see how anything is going to be superior to just testing yr BG before and after eating.

 

[Docb]

True, but when answering questions like... On balance, what sort of dietary changes might be the best approach for me? ... it's a different matter. The answer is most likely to depend on the prejudices of the respondent rather than the practical experiences of people with D.

 

[Eddy Edson]

You might find this interesting, just published: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2723644?resultClick=3

Assessment of a Personalized Approach to Predicting Postprandial Glycemic Responses to Food Among Individuals Without Diabetes

Applying a proprietary analytics model to various physical indicators and gut microbe samples to predict glycemic responses to food amongst a fairly large cohort of non-diabetics.

This summarises the results:



The blue line is the performance of their model at predicting high glycemic response. It's pretty good: at around 100% specificity it delivers about 60% sensitivity, and it delivers ~90% sensitivity with ~85% specificity. I think that's as good as many widely accepted medical diagnostic protocols.

By contrast, simple carb-counting looks pretty squidgy: at 85% specificity you're only getting 40% sensitivity.

Consistently with all the recent studies, individual variability is very high, which makes the good performance of the model quite impressive, IMO.

This shows the results of eating a bagel+cream cheese on two different days for six different participants:



For each individual, the results of the two bagel meals are fairly similar, but between individuals, there is wide variation. Eg: One in six can go above 10mmol/L at 2 hours; one third above 8.5mmol/L; one half above 8.5mmol/L after 45min.

That's for non-diabetics. IMO, diabetics really shouldn't pay much attention to guidelines setting out 10mmol/L or 8.5mmol/L or whatever targets. They seem very much out of date in light of all the recent CGM work and as far as I can tell, were only ever formulated as crude estimates for what might correspond to daily area-under-curve outcomes delivering target HbA1c's. Anybody with a Libre would be better off trusting its (area-under-curve based) estimate for HbA1c than worrying about "spikes", and even finger pricking if done reasonably frequently will give a fair estimate, in my experience.