• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

DiRECT study 2 years results presented

bulkbiker

BANNED
Messages
19,569
Location
Worthing
Type of diabetes
Type 2
Treatment type
Diet only
Looks like the 2 year results from DiRECT have just been announced at the DUK Professional conference in Liverpool



Some of our fears about the sustainability of the program seem to have been correct. I'm rather pleased to see of the 6 in the control group that achieved remission there are still 5 going strong.. I still have a sneaky suspicion they found low carb!
 
I have been put into contact with one of the dietitians involved in the trial so hopefully I may get an update on how the 6 control group members achieved remission after 12 months.. I'll let you know.. apparently it was weight loss but I still want to know how they did it.. That 5 are still in remission after 24 months warms my heart...
 
I always struggle to understand these graphs.

So please, anyone correct me if my understanding is wrong.

Anyone know what Fisher's Exact Test is?
https://en.wikipedia.org/wiki/Fisher's_exact_test
Where is @Oldvatr when you need him?

There were 2 groups, each had 149 people in them.
One group was a control, without dietary intervention.
The other group did the Newcastle Diet (shakes, 800 cal restriction for 8 weeks. Or longer?)

Am I correct that these were T2 diabetics selected on the criteria that they have/had fatty livers? And other co-morbidities were excluded?

Weight Loss Graph A:
Of the control group, 2% (3/149 people) had lost 15kg or more by year 2
Of the ND group, 24% (36/149 ppl) had lost 15kg or more by year 1, and 11% (17/149 ppl) had kept it off at year 2.

Remission Graph B:
I understand that Professor Taylor counts remission as an HbA1c of below 48mmol/mol for 6 months. Is that still the case? Where does he stand on Metformin and other medications, nowadays?
Of the control group, 4% (6/149 ppl) achieved remission in year 1, and 3.4% (5/149 ppl) stayed remissed at year 2.
of the ND group, 45.6% (68/149 ppl) achieved remission in year 1, and 35.6% (53/149 ppl) stayed remissed at year 2.

Remission linked to weight loss Graph C:
Of the ND group, the more weight lost, the greater chance of remission. Significantly so.
However, different weight loss groups seemed to have difference rates of continued remission, with the greatest weight loss group also having the most number of people lapse and regain by year 2.

I am completely confused by the n/N numbers on graph C since they do not seem to relate to numbers elsewhere. Possibly due to drop out numbers? Possibly due to statistical number crunching?

As always, my takeaway remains unchanged:
ND is only worth doing if you have a fatty liver, so get tested before embarking on it.
Those who struggle to lose weight will struggle just as much on this regime, and less weight loss = lower remission rate
The real challenge is to keep weight loss off after the end of the ND, as with almost all weight loss regimes.
Extrapolating the regain rate (greatest weight loss, Graph 3) into future years, does not suggest (to me) that the remission will last very long.
 
Correct


I think that the most successful group were those that had lost 15 kg or greater (hence people using 15 kg loss atsthe magic number for the ND) I guess its not surprising that in that group were the most "failures" as keeping that amount of weight off after losing it through "crash dieting" will make it hardest to maintain the loss?
 
I don't like how they defined "remission", as someone could have gone from a A1c of 49 to 47 and be counted as in "remission", even if they have never had an A1c over 49. Yet someone who has gone from 80 to 49 would be counted as having failed.
 
From the DiRECT protocol

"Inclusion criteria

  • Written informed consent
  • Men and women aged 20–65 years
  • T2DM of duration 0–6 years (diagnosis based on 2 recorded diagnostic-level tests, HbA1c and/or blood glucose)
  • HbA1c ≥ 48 mmol/mol at the last routine clinical check, within last 12 months if on diet alone
  • HbA1c ≥ 43 mmol/mol if on treatment with oral hypoglycaemic agents
  • Body Mass Index (BMI) >27 kg/m2 and <45 kg/m2
Exclusion criteria
  • Current insulin use
  • Recent routine HbA1c ≥108 mmol/mol
  • Weight loss of >5 kg within the last 6 months
  • Recent eGFR <30 mls/min/1.73 m2
  • Substance abuse
  • Known cancer
  • Myocardial infarction within previous 6 months
  • Severe heart failure defined as equivalent to the New
  • York Heart Association grade 3 (NYHA
  • Learning difficulties
  • Current treatment with anti-obesity drugs
  • Diagnosed eating disorder or purging
  • Pregnant/ considering pregnancy
  • Patients who have required hospitalisation for depression or are on antipsychotic drugs
  • People currently participating in another clinical research trial
  • People with contraindications for MR scanning"
 
ringi said:
I don't like how they defined "remission", as someone could have gone from a A1c of 49 to 47 and be counted as in "remission", even if they have never had an A1c over 49. Yet someone who has gone from 80 to 49 would be counted as having failed.
Click to expand...
Agree but unfortunately all the studies appear to have adopted that as their criteria.. I think it should beHbA1c lower than 40 mmol/m from a start point over 48 mmol/m but hey what do I know.. I might try and get some anonymised individual numbers from the dietitian I've just contacted and see if I can work out what the drops were per person as well as the starting levels. There seem to be a few comparing DiRECT to the Virta trials but they are two very different beasts..
 
Brunneria said:
Anyone know what Fisher's Exact Test is?
Click to expand...

It is a test to see how likely the results are due to a sampling error. For example, if only 7 people in the "intention to treat" group had got "remission" compared to 5 in the control group, then it is not clear if the result would have been the same with a much larger number of people.

Given how clear cut their results are, you don't need to understand complex stats to know if the results are significant.
 
bulkbiker said:
There seem to be a few comparing DiRECT to the Virta trials
Click to expand...

Virta keeps people on Metformin. I think it is best to think of the Virta trails as being a study to show how much USA health care insurance providers can save on drugs with pt also getting better results than they were without Virta. Let's remember that Virta practically did not exclude anyone with DM2 from their study.
 
What I was saying was its not a fair comparison.. although we get many here too who stay on metformin when their HbA1c results may not "justify" it. When Virta reverses people who were on insulin (who would have been automatically excluded from DiRECT) I can't see it being fair to make comparisons between the two but many seem to.
Obviously the surrounding for Virta are completely different to the UK but from my reading I can't see that they excluded anybody whereas the DiRECT exclusion criteria were fairly rigourous.
Both are of course having a huge impact in T2 world.
 

Hi Ringi,
I put a link to a description of the test in my post - effectively answering my own question.

But I beg to differ about not needing to understand these things.
There has been enough statistical manipulation in studies over the years - and it is ongoing, whether deliberate or not - that I do not take these things on trust. Never will.
The ongoing statins debacle is proof enough of that.
 
Quite interested in some of the exclusion criteria especially those with HbA1c of 108 or above - cherry picking? Many members on here have been diagnosed at that level or above and have had dramatic turnarounds on LCHF to well below the DIRECT remission threshold of 48.

Your point about the Virta study and success with getting people off insulin is also interesting.

I also think that measuring insulin and insulin resistance would be very illuminating in all these studies - suspect it’s not the weight loss per se that leads to most success, but rather that the weight loss reflects improved insulin sensitivity.

A study comparing like for like would be very valuable.

Thanks for posting this.
 
Many of the exclusion criteria are due to them stopping all diabetes drugs on the first day of the diet, so that GPs can't claim they don't have the time to adjust drugs etc. They also need to remove anything thst requires the skilled judgement of a HCF as they need to remove as much variation as possible so other teams can repeat the results.

The large scale reallife data will be out in a few years, and they are being more flexible on who are allowed to use the service being setup at present to surport people.
 
You must log in or register to reply here.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn More.…