Some thoughts on the recent "Type 2 diabetes reversal" paper

"Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol"

Shortened URL for this post is http://is.gd/nbqdYX

Let me first of all make it very clear that the following thoughts are mine alone - I do work for Diabetes UK but I'm not a spokesperson so in the unlikely event that anyone wants to attribute any of the following to someone, please make sure it's to me, Jo Brodie - thank you :)

I'm also going to dual-post this to my ScienceDiabetes blog.

Other important things to note are that (a) I'm not medically or dietetically trained and therefore in no fit state to give anyone medical, or dietary advice and (b) I'm writing this at home without full access to the full 'diabetesphere' that I can log in to from work. So I might come back to this.

I do not think anyone should try this drastic diet and certainly not without full medical support (don't forget these people were in a trial which meant that the researchers had to take a great deal of responsibility for their wellbeing). I also think it's rather too soon to be talking about 'cures' and that we've no idea what effect this diet will have on people who've had Type 2 diabetes for longer than four years, and who have complications.

What does reversal of Type 2 diabetes mean
Forgetting for a minute that this was a tiny study (11 people) I'm always a bit wary of the terms 'cure' or even 'reversal' as it's not really what I'm used to in discussions and readings about diabetes. I've heard endocrinologists refer to people having their diabetes "pushed back along the continuum / curve" if they've undergone bariatric surgery (eg stomach stapling) and it seems to be fairly well accepted that losing weight (often a combination of losing weight and increasing physical activity) can help delay progression to Type 2 diabetes in those at risk. And people making dramatic lifestyle changes / improvements do seem to be able to come off tablets (I suspect some people with Type 2 diabetes using insulin may also be able to come off that too).

But Type 2 diabetes is generally thought to be a progressive condition (however the authors of the latest study suggest that this mightn't be the case). Progressive in this means that the way in which Type 2 diabetes is treated tends to change over time, in a way that it doesn't really in Type 1 diabetes (people start on insulin and remain on it, whereas people with Type 2 might go through the stages from diet plus physical activity controlled to tablets plus diet plus physical activity controlled to insulin plus diet plus physical activity controlled. Note the constant through each of these (diet and physical activity!) - it's not quite diet > tablets > insulin.

When someone comes off medication I never think that they no longer have diabetes. I think they've jumped back a stage, perhaps even two - but I don't know if they're necessarily 'off the conveyor belt'. The progression of Type 2 diabetes involves a combination of insulin resistance (when the body is less able to respond to insulin produced - and in the earlier stages this can lead to hyperinsulinaemia because the pancreas secretes more insulin to compensate) and 'beta cell dysfunction' - basically the insulin-producing cells begin to struggle to produce enough insulin.

Type 2 diabetes takes a while to develop and during its development the person might have no symptoms whatever. It's estimated that, at diagnosis of Type 2, someone will already have had the condition for several years (~8-12, however with better awareness this is probably dropping). That's why doctors, and my employers, generally recommend a focus on risk factors in preference to symptoms as symptoms are notoriously unreliable anyway.

An important aspect of this could be that damage done to tissues and organs may, or may not (I don't know) be irreversible - even if glucose levels are brought back to normal later on. If someone has been hyperglycaemic (too much glucose in the blood) for some time, before diagnosis or a few or many years afterwards, then there's scope for damage.

Glycaemic memory / legacy effect and later complications
There is this idea of a "glycaemic memory" in which the body somehow retains an imprint of whatever the glucose status is now, later down the line.

This seems appears to work both ways (or I've misunderstood it!):

High blood glucose levels early on, and for a long time, can be 'remembered', so that even when glucose levels are normalised, complications can still arise several years later -

"... studies in type 1 diabetes (e.g. DCCT) and type 2 diabetes (e.g. UKPDS) have shown that a period of poor glycaemic control earlier in the course of the disease is associated with an increased burden of complications much later in the course of the disease, even when glycaemic control is latterly improved. The Veterans Affairs Diabetes Trial suggested that more than 12—15 years of poor control in older type 2 patients minimised the benefits of subsequently improved glycaemic control. The delayed adverse effects of hyperglycaemia emphasise the importance of effective early glycaemic control." Source: Bailey CJ and Day C (2008) Glycaemic memory British Journal of Diabetes & Vascular Disease 8 (5): 242-247. doi: 10.1177/1474651408098784

Or, a period of good control might have a protective effect down the line

"Furthermore, there appears to be a beneficial ‘legacy effect’ or ‘glycaemic memory effect’ following a period of intensive glucose control that has also been observed in other studies.(5)" Source: Younis N, Soran H and Hassanein M (2009) Cardiovascular disease and intensive glucose lowering in type 2 diabetes Quarterly Journal of Medicine 102 (4): 293-296. doi: 10.1093/qjmed/hcp001

It seems pretty clear to me that good blood glucose control is a good idea, but importantly it's not just glucose that needs to be considered, but blood pressure and blood fats (blood lipids) too.

If someone already has diabetic complications then it's not always a done deal that dramatic weight loss or improvements in glucose control is necessarily going to make that much difference to those complications (it depends on the complications though).

There are exceptions of course - sensory neuropathy can resolve over time by itself anyway, I think it can also be improved with better glucose levels too. Some eye damage can actually get temporarily worse with improving glucose levels although I think there can be some improvements there too. Not sure about kidney or cardiovascular damage though (yeah I should probably find out!). What I want to emphasise is that 'resolution' of diabetes doesn't automatically mean resolution of all accompanying complications. A point put forward much more pithily by @EvidenceMatters

For me, a cure would have to mean resolution of all threats, glycaemic (blood glucose), lipidaemic (blood lipids), hypertensive (blood pressure) and any complications. I don't think we know yet what the long-term effects or benefits are of the various gastric surgeries that resolve cases of Type 2 diabetes. It should be noted that these procedures are not risk free and one effect is poorer absorption of some nutrients (obviously this can lead to weight loss).

Drastic diets
There are potentially very serious risks to health from following a drastic diet without medical supervision or support. I have read a couple of anecdotal horror stories of young women going on extreme diets and dying from heart failure - bit grim (hat tip @landtimforgot). But in general, the body just needs energy to function. I didn't spot any mention in the paper suggesting the participants had an electrocardiogram / EKG / ECG.

The 11 people on this trial were not just sent off with diet packs but given access to telephone support and regularly monitored every four weeks (in weeks 1, 4, 8 and then 12 but they stopped the diet after 8 weeks). It's also worth pointing out that three trial participants dropped out because they didn't stick with the diet in the first month and 1 dropped out for other reasons (so 3 / 15 = 20% drop out). The diet was about 46% carbohydrate but this shouldn't be taken to mean that it was medium-high carbs because the overall amount of all nutrients taken in was so low.

A diet of 600 calories a day is tough going and not really something anyone should attempt without proper support from someone like a dietitian who knows what they're talking about (this usually excludes anyone calling themselves a 'nutritionist' as anyone can call themselves that). I'd suspect that such a low calorie diet would be 'prescribed' by a doctor first as it's pretty extreme. Most of the reports I've seen have stressed this which is good.

Enough people have tried to tell me, while hoping to get me to promote their miracle cure, that being off medication means they're cured of Type 2 diabetes but that's an oversimplification I think. Additionally, 'coming off medication' doesn't mean that that was actually the appropriate thing to do!

Having said that I've seen a couple of commenters on news sites raising the issue about relative risks - devastating complications from diabetes are rather serious too... I don't have a snappy answer I'm afraid.

Insulin physiology
The paper talks about the dual problems in Type 2 diabetes of beta cell failure and insulin resistance (it also distinguishes between hepatic (liver) insulin resistance and peripheral (I assume this refers to skeletal muscle) insulin resistance.

The liver produces its own glucose and can do so in two ways: gluconeogenesis (or de novo synthesis) which means creating from scratch from precursor molecules, and glycogenolysis which means freeing glucose previously stored as glycogen.

If the body gets the message that there's not enough glucose it takes steps to get some released from the liver. Unfortunately, in diabetes the bloodstream may be swimming with glucose but the cells aren't getting it and so they're 'reporting' a lack of glucose and the liver is chucking more into the bloodstream. This has been described as "starving in the land of plenty" as the cells go hungry but there's plenty of glucose available, but inaccessible without insulin (Type 1) or the ability to respond appropriately and to dwindling supplies of insulin (Type 2).

Insulin, when all is working well, suppresses this hepatic production of glucose. When all isn't working well, it doesn't.

Fasting glucose levels (ie, not after a meal) tell you what's going on in the liver in terms of it just getting on and chucking out some glucose into the bloodstream. It does this because it's failing to respond to the insulin signal which should be telling the liver to suppress its glucose production. But after following this diet for just a week, fasting glucose levels plummeted, as did hepatic glucose production.

Observations on the people studied in the paper
There were eleven people studied, two women, and they all had been diagnosed with Type 2 diabetes for less than four years. I thought it was interesting that no-one appeared to have an HbA1c of more than 7.7%. This is a reasonably high HbA1c (I think normal values are lower than 6% and people with diabetes are generally recommended to keep it lower than 6.5 per cent) but some poeple with diabetes will have really much higher HbA1c values than this.

They lost a lot of weight, very quickly - an average of 3.9kg in the first week (of which 61% was fat loss) and 5.7kg from weeks 1-4 (I assume they mean a total of 5.7kg over the four weeks and not 5.7kg each week which would be a bit alarming) of which 86% was fat. In the final stretch they los another 5.7kg over the last month (assumed) of which 94% was fat. Gosh. They shifted a rather large 15% of their bodyweight, an amount which is generally achievable (for most people) but over a much longer time.

But the rapidity with which the weight was lost is suggested to be behind the disappeareance of triacylglycerol (TAG or triglycerides) from the liver and pancreas. Triacylglycerol consists of a glycerol molecule (the 'backbone') to which is bonded, by an ester linkage, three fatty acids of varying length. Un-attached fatty acids (called free fatty acids FFAs, also known as non-esterified fatty acids NEFAs) can interfere with an organs ability to do its job and apparently impede the normal production of insulin from the pancreas.

I'll probably come back to this as I've written a few notes from my first reading of the paper, but I expect it will bear further reading and digestion.

Please let me know if you spot any mistakes or omissions, thank you.
Please don't go cutting your diet dramatically without speaking with your doctor.

Lim EL, Hollingsworth KG, Aribisala BS, Chen MJ, Mathers JC and Taylor R (2011) Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol Diabetologia

[Published online 9 June 2011] doi: 10.1007/s00125-011-2204-7 (html) (PDF)

This was funded by Diabetes UK who also happen to employ me, but my blog post and thoughts are my own :)

Further reading

• Can diet alone reverse Type 2 diabetes? Diabetes UK website 24 June 2011
• Type 2 diabetes in newly diagnosed 'can be reversed' BBC News website 24 June 2011
• ‘Reversing’ type 2 diabetes? Quality in Care Programme blog 24 June 2011
• Is this a ‘wonder cure’ for type 2? DiabetesChoices website 24 June 2011
• 'Crash diets' studied for type 2 diabetes NHS Choices "Behind the Headlines" website 24 June 2011