Hello from the home of The Beatles! Diabetes UK 2019 kicked off in Liverpool today, and all eyes were on two-year data from DiRECT – a smashing success, in our view. We also heard a fascinating debate on the definition of diabetes remission, encouraging new data from the UK’s national diabetes prevention program, and more. Check out our preview for a look at what’s to come.

Diabetes Therapy Highlights

1. Two Year DiRECT Data Show 70% Maintenance of Type 2 Remission from Year One (36% Overall); Durable Remission “All About Weight Loss”; Long Term Outcomes?

Two-year data from DiRECT (n=298) show that 70% of participants who achieved type 2 diabetes remission at one year maintained remission through 24 months, giving an overall 36% rate of remission in the intervention arm (vs. 3% of controls, p<0.001; see publication in The Lancet). At two years, 11% of the intervention arm achieved ≥15 kg weight loss (p<0.001 vs. 2% of controls). This compares to 46% remission (A1c <6.5% with no diabetes medications) in the intervention arm at one year (vs. 4% in control group), driven by 24% of participants achieving ≥15 kg (~33 lbs) weight loss – the two primary outcomes of the trial. Mean weight loss in the intervention group was 10 kg (~22 lbs) at 12 months and 7.6 kg (~17 lbs) at 24 months, compared to 1 kg (~2.2 lbs) and 2.3 kg (~5.1 lbs) in the control group, respectively. Remission and weight loss, and their maintenance, remain closely linked: At year two, 64% of participants who had lost ≥10 kg were in remission. Those who achieved remission at both time points (n=48) did experience mild weight regain from 12 to 24 months, while those who achieved but then “lost” remission had more serious weight regain (see figure below). From where we stand, it’s hard to over-emphasize the significance of these results: Not only does DiRECT achieve an impressive rate of diabetes remission, but it shows that can be done in the primary care setting through nurses and dieticians with fairly minimal training. We are thrilled to think about this moving to standard of care and all the people that could do so much better with appropriate investment in this intervention. 

In the entire population, 24-month remission was linked to greater weight loss from baseline (aOR=1.2 per kg lost) and from 12-24 months (aOR=1.11 per kg lost), and there were slight associations with older age (aOR=1.08 per year) and sex (aOR=0.44 female vs. male). Of particular note, 24-month remission did not vary by baseline BMI or duration of diabetes – the latter is particularly interesting, as data presented by Prof. Roy Taylor at ADA 2018 suggested diabetes duration plays an important role in beta cell recovery and likelihood of remission. However, in one-year data presented by Prof. Mike Lean, older age, lower A1c, and fewer diabetes medications predicted remission, while male gender, shorter diabetes duration, and higher baseline A1c and BMI predicted weight loss ≥15 kg. That said, we’re looking forward to more specific subgroup analyses and anticipate far more insight over the coming year(s). Only 16 of 298 randomized participants withdrew from the intervention during year 2, and 86% of those 298 attended the 24-month review (24 months is the end of the RCT, though we understand that additional follow-up is planned).

• Secondary outcomes indicate improvements in long-term risk factors, and data suggest a potential impact on CV events and cancer. The DiRECT intervention is associated with a significant, ~4 mmHg reduction in systolic blood pressure (p=0.04 vs. controls), despite lower use of antihypertensives (p=0.006 vs. controls) at 24 months. And despite little impact on cholesterol, the intervention does drive an improvement in triglycerides. Among participants in remission at 24 months, Q-RISK (a measure of 10-year CVD risk) fell from 16.1% to 8.2%, heart age from 70.4 to 60.8 years, and A1c from 7.4% to 5.9%. Unfortunately, the field of weight loss and obesity treatment as a whole still lacks evidence that treating weight – and improving cardiometabolic health through weight loss – translates to an improvement in long term outcomes. On this front, DiRECT is promising: From 12-24 months, the intervention group saw fewer serious adverse events than the control group (p=0.029, NS from 0-12 months and 0-24 months). While this is a small post-hoc analysis, there is a suggestion of a positive impact on both CV events and cancer.

• These highly-anticipated results follow one-year results from IDF 2017 and the November announcement that NHS would conduct a 5,000-person pilot on diabetes remission informed by the design of DiRECT, to be focused on recently-diagnosed patients. At a high level, we’re impressed by the UK’s clear willingness to invest in both prevention and remission in a huge way – DiRECT (funded by Diabetes UK) cost ~$25,000/person in the intervention arm, though we imagine NHS can likely push this down through scale. The UK seems to have the message about early investment in health and about interventions that are high-investment and high value.

• While weight regain did occur in DiRECT, presenters emphasized the beneficial impact of “rescue therapy” and that 24-month weight loss was still superior to most lifestyle interventions. Relapse management was used when a participant regained >2 kg or “relapsed” into type 2, and involved some combination of (i) reviewing causes of weight regain, (ii) reviewing behavioral strategies, (iii) brief use of diet replacement/food reintroduction, and (iv) the offer of orlistat (only three participants used this, all in the latter 12 months). While 50% of the intervention group used one or more rescue plans during the trial, this group actually achieved an equivalent rate of remission to the group that didn’t use rescue strategies (despite slightly lower weight loss).

• Only 40% of intervention participants were taking diabetes medications at 24 months (down from 75% at baseline), compared to 84% of control participants (77% at baseline). That is, ~25% of participants in the intervention arm did not achieve remission but were not taking diabetes meds. To be sure, it’s possible some of these people could have benefitted from taking medications, but this is also an endpoint that’s highly important to patients. If those 25% were managing diabetes successfully, but without quite reaching an A1c <6.5%, we consider that a win. Indeed, on average, the intervention group saw a ~0.55% drop in A1c (6 mmol/mol; p=0.0063 vs. baseline), compared to no change in the control group. We’d like to see further discussion, of course, of medicine designed to reduce cardiovascular and renal complications – we also see this as an investment.

2. How to Define Diabetes Remission? Debate Over 6.0% vs. 6.5% Emerges as Professional Societies Aim to Publish Consensus

During a late afternoon session, Prof. Roy Taylor (pro-6.5%) and Dr. Chirag Bakhai (pro-6.0%) debated the merits of these differing definitions of diabetes remission. For context, Diabetes UK, EASD, and ADA are currently working to develop an international consensus on remission of type 2 diabetes, and Diabetes UK recently published an interim position statement on the topic (without naming a definite A1c cutoff). And Prof. Taylor, during his presentation, announced that the Association of British Clinical Diabetologists and the Primary Care Diabetes Society just published their own position statement, stipulating (i) weight loss, (ii) A1c <6.5% or FPG <126 mg/dl, and (iii) attainment of these following cessation of all glucose lowering therapies for the achievement of remission. However, we expect the eventual DUK/EASD/ADA consensus to serve as the most definitive statement in this area; the illustrious working group includes Drs. Will Cefalu, Philip Evans, Roy Taylor, Douglas Twenefour, Francesco Rubino, Philip Schauer, Matthew Riddle, Hertzel Gerstein, Carel le Roux, and Michael Nauck. Of note, the audience was asked to vote on 6.0% vs. 6.5% before and after the debate took place; from start to finish, 6.0% climbed from the minority to the majority, though both votes were on the close side. Find the primary arguments in favor of each below – while we aren’t sure of the “best” answer, we do think this is a highly interesting and increasingly important area of discussion.

• Prof. Taylor argued in favor of an A1c <6.5% definition with “five good reasons”, stipulating that these must be associated with weight loss, per the statement he co-authored:

  • It is a motivating, achievable, patient-centered definition;

  • normalizing lipid metabolism gives remission of diabetes plus normalization of macrovascular risk;

  • A1c has been shown to remain constant ≥2 years in this range;

  • it’s consistent with the just-released ABCD/PCDS position statement; and

  • DiRECT defined it this way – he implored the audience not to “throw away DiRECT.”

Prof. Taylor also argued strongly in factor of a new term – “post-diabetes” – to describe those in remission: In his assessment, putting type 2 into remission involves a reversal of the pathophysiology underlying diabetes and is actually a lower-risk clinical stage worthy of recognition. He pointed to DiRECT data (those above) suggesting a “clinically significant” reduction in adverse events with remission, and explained that mislabeling remission as prediabetes erroneously implies that adverse consequences are still a present threat.

• In the counterargument, Dr. Bakhai argued that a cutoff of <6.5% is dangerous for patients and took philosophical issue with the idea that miniscule differences in glycemia could bring a patient in or out of remission. He rephrased the question: Should remission involve a return to normoglycemia, or is it okay to return to “impaired glucose regulation”? Dr. Bakhai noted that ADA’s 2009 statement defined complete remission as <6.0% (vs. “partial remission” at <6.5%) before outlining a potentially problematic scenario: Patients between 6.0% and 6.5% who need to come off diabetes medications to meet the definition of remission – and realize the financial benefits this has in the UK – will feel pressure to do so. In the real, non-DiRECT world, where weight loss and medication use occur simultaneously, this is a problem: It could result in many patients who are doing very well managing their diabetes to creep upward while technically still achieving “remission.” (Or they rebound above 6.5% and have to restart medication.) Dr. Bakhai questioned whether this is really the best outcome for patients and argued that, while microvascular risk is still low at A1c <6.5%, it does exist. Going one step further, he said, some people will get the “cure message,” and people who achieve “cure” and even “remission” will be less likely to come for follow-up – but with a stricter threshold, there would be less worry.

3. Two-Year DiRECT Mechanistic Study Data Further Support Reduction in Hepatic + Pancreatic Fat, Restoration of Beta Cell Function as Drivers of Remission; Differentiate Responders/Non-Responders/Relapsers

The twin cycle hypothesis, developed by Prof. Roy Taylor, posits that hepatic and pancreatic fat directly cause the underlying pathophysiology of type 2 diabetes – namely insulin resistance and beta cell failure. Today, Dr. Ahmad Al-Mrabeh presented extensive and compelling data from the DiRECT Mechanistic Study (a DiRECT substudy), designed to investigate the twin cycle hypothesis by taking a subgroup of participants (n=64) and dividing them into responders (n=29 at the end of the trial), relapsers and non-responders. The findings strongly supported the twin cycle hypothesis – in fact, the data were so good that Dr. Taylor was moved to comment, somewhat light-heartedly: “This hopefully explains the simplicity of type 2 diabetes.” Not only was liver and pancreatic fat shown to differ greatly in responders, non-responders, and those who relapsed, beta cell function (first phase and maximal insulin response) was restored in the responders, but not the other groups. It was clear that weight loss was the driving factor for these changes.

• Liver fat, quantified by MRI, showed a dramatic 80% decline from baseline for responders, appearing to be influenced rapidly by the low energy diet. In contrast, the non-responders gained back liver fat more rapidly after the initial period. Unsurprisingly, the liver fat change was shown to be a function of body weight loss. VLDL triglyceride production from the liver (measured by an isotopic method) showed analogous changes.

• In type 2 diabetes, the pancreas is one third smaller and irregular in shape, compared to the healthy pancreas. Using an imaging technique, it was shown that there is a three-fold larger intra-pancreatic fat reduction in the responder group compared to the non-responder group.

• Using a glucose challenge technique, responders were shown to restore their first phase insulin response quickly after initial weight loss, and their maximum insulin secretion improved over 12 months – showing a clear restoration of their beta cell function. By contrast, non-responders never restored either their first phase insulin response or a better maximal insulin response. The technique used was SISTA (Stepped Insulin Secretion Test with Arginine).

• Comparing responders and relapsers, it was clear that weight loss was the major factor. Relapsers had a similar initial weight loss, but regained much weight. In doing so, they lost their first phase insulin response, in contrast to the responders. We also saw relapsers increasing plasma VLDL triglycerides, and not reducing their pancreatic fat.

• Per the twin cycle hypothesis, excess obesity (fat exceeding a personal weight threshold) leads to fat deposition in the liver, which in turn leads to increased plasma triglycerides. In susceptible people, this leads to fat in the pancreas and beta cell dysfunction, reducing insulin secretion and exacerbating insulin resistance in a vicious cycle. The concept behind DiRECT is that by reducing weight, this cycle can be broken and hepatic and pancreatic function restored.

Questions and Answers

Q: How do you measure pancreatic fat correctly? How do you explain the triglyceride drop?

A: Visceral fat invasion is problematic in measuring pancreatic fat, but we’ve developed new image measurement techniques that we standardized and published.

A: Fatty liver induces both insulin resistance and coronary artery disease. Many people have fatty liver, but not all have diabetes. That’s because only some people have beta cells that are susceptible to fat.

4. Banting Memorial Lecture: Prof. Simon Heller Sheds New Light on Hypoglycemia a Century After the Discovery of Insulin

For the prestigious Banting Lecture, Sheffield’s Dr. Simon Heller took us through the history of insulin-induced hypoglycemia from Banting and Best up to the latest CGM devices. He noted that although today’s tools are better than those available in Toronto 100 years ago, they are still inadequate if patients are to meet ever tightening glucose targets without the risk of hypoglycemia. He also strongly suggested that hypoglycemia causes cardiac dysfunction in susceptible people, leading to the inconsistent results of cardiovascular trials in type 2 diabetes and the ‘dead in bed’ syndrome in type 1 diabetes.

• In summer 1921 Banting and Best were attempting to isolate insulin from the pancreas, although progress was difficult. Their collaborator, James Collip, is hardly known, but he was a very important contributor to the production of insulin in volume. He was also the first person to identify insulin induced hypoglycemia. Originally, 1 unit of insulin was chosen to be the amount required to in induce a hypoglycemic seizure in a rabbit.

• Banting described insulin induced hypoglycemia in a 1923 paper. In 1941 RD Lawrence first described reduced awareness of hypoglycemia. It was only in the late 1970’s that we learned that during hypoglycemia, the adrenaline and glucagon response was impaired with increasing duration of diabetes. After five years duration, glucagon was significantly reduced and after 15 years, both hormones were impaired.

• In 1985 it was demonstrated that strict glycemic control was associated with reduced counter-regulatory response. In 1987 it was established that tight control sets the threshold for hormonal response below the glucose level needed for normal cognitive function. So by the time patients develop a counter-regulatory response they are already impaired and need the help of another person.

• It became apparent that hypoglycemia itself causes the protective response to diminish. In 1991, a study showed that exposure to hypoglycemia caused a lessening of counter-regulatory responses in the same subjects on the following day. A few years later the same effect was demonstrated in patients who had avoided any hypoglycemia for three weeks.

• Dr. Heller remarked that after all this time, it’s remarkable that there is no treatment to avoid hypoglycemia unawareness. People have suggested caffeine or modafinil, but nothing has entered large scale trials. Instead we have to avoid hypoglycemia itself. CGM and automated insulin delivery (AID) are both becoming more helpful in this regard.

• A meta-analysis of major CV trials in type 2 diabetes, (UKPDS, ADVANCE, VADT, ACCORD) showed no average impact on all-cause mortality, or cardiovascular mortality. It’s notable that in the UKPDS follow up there was a clear reduction in all-cause mortality (participants were all recently diagnosed). VADT and ADVANCE had contradictory trends but neither were significant. Famously, ACCORD had to be stopped because of unacceptable mortality in the intensively treated group. Why did the trials give different results? Dr. Heller asserted that severe hypoglycemia predicted mortality in three of the trials, although he accepted that not everyone in the room might agree. It’s possible that severe hypoglycemia is associated with patients for whom there is another primary risk factor.

• Dr. Heller presented a lot of evidence suggesting that hypoglycemia increases cardiovascular risk. He started by elucidating numerous basic mechanisms by which hypoglycemia could influence cardiac function. His work investigated thrombosis and clotting which showed adverse changes many days beyond a hypoglycemia episode. He also compared ECG and CGM data, which demonstrated cardiac arrhythmia during extended nocturnal asymptomatic hypoglycemia in some (but not all) people with type 2 diabetes. He also noted an eightfold increased risk of bradycardia in some (but not all) patients

“Dead in Bed” syndrome is the leading cause of sudden death in people with type 1 diabetes – and might be caused by hypoglycemia. It’s believed that hypoglycemia may cause an abnormally long and alarming QT, which could lead to cardiac arrest. Dr. Heller supported this with a somber CGM record of a patient who died during nocturnal severe hypoglycemia. But it appears that bradycardia and sudden death are perhaps confined to susceptible individuals - suggesting that some kind of screening might be possible.

5. Prof. John Todd Updates on GPPAD (Global Platform for the Prevention of Autoimmune Diabetes) in the UK: POInT (oral insulin) Enrollment Progressing

Oxford’s Prof. John Todd gave an update on the UK arm of GPPAD’s POInT trial, announcing that four participants in the UK have been enrolled and 68,587 newborns across all five participant countries have been screened; In the UK screening began in October 2018.  According to Prof. Todd, POInT aims to screen 330,000to enroll 1,040 newborns (estimating that 1 in 400 will be at-risk and participate).Oxford (Chief Investigator, Dr Matthew Snape) itself plans to screen 30,000 babies in three years. So far in the UK, 34 newborns have been confirmed eligible for the study (>10% genetic risk of developing autoantibodies by age six, among other criteria), 28 in-person consultations have taken place, and 14 of these families wish to participate in POInT, (and four are officially enrolled. Prof. Todd characterized this progress as on-track, and he was particularly enthusiastic about the infrastructure developed for screening. Screening for POInT is actually taking place through a separate study, INGR1D (INvestigating Genetic Risks of type 1 Diabetes; Chief Investigator, Dr Manu Vatish): researchers approach women for neonatal screening starting at ~20 weeks gestation, then test the baby and discussion randomization in the first week of life. Currently, INGR1D is only active in four maternity hospitals in the Thames Valley, and Prof. Todd expressed positivity about the level of interest parents have had in the trial – as well as a strong desire to expand geographically: 6,004 have been consented in six months, and on average 66% of those approached consent to take part (4,005 babies have been born and had blood processed). Additionally, he was very excited about Public Health England’s agreement to allow researchers to use newborn Guthrie cards (heel prick blood tests) for the DNA testing, making the test much less cumbersome for the investigators. Not only, he said, will this serve its purpose for type 1 risk screening, but it also creates an infrastructure and sets a precedent for future preventative research across other childhood diseases.

• As a reminder, children in POInT will be given daily oral insulin (titrated to 67.5 mg daily) or placebo until age three, starting at 4-7 months old, then they’ll be followed up to 6 years. The Helmsley Charitable Trust awarded five grants totaling $52 million to GPPAD, led by the Institute of Diabetes Research, Helmholtz Zentrum München, in December 2017, its largest-ever investment in type 1 (and one of its largest ever, period). The higher dose of oral insulin and earlier intervention (genetic risk, as opposed to after autoantibody development) have been cited as reasons for optimism despite lukewarm oral insulin results to date. Pre-Point, as Prof. Todd pointed out, certainly gives reason to think the higher dose of oral insulin could have a stronger effect on the immune system and tolerability.

6. Opening Plenary Features Patient Perspectives on NHS DPP, DiRECT Trial

Three patients offered their perspectives on the UK’s Diabetes Prevention Programme and the DiRECT trial throughout the day. We loved seeing patient perspectives incorporated so prominently into the program, including two during the conference’s opening plenary – we’ve summarized their comments below.

• Harry Matharu, aged 57, was part of the UK’s NHS Diabetes Prevention Programme. His mantra is “if it swims it slims”. But he also does not eat any simple carbs or added sugar.

  • After a routine checkup, Harry’s GP referred him to the NDPP because he was borderline obese and prediabetic. His weight was 190lbs and A1c 6.5%. One year later, he had lost 50lbs to reach a weight of 140lbs and his A1c was 5.2%

  • The program consisted of education sessions, which also served to inspire and motivate Harry. There were a lot of myths about dieting.  ‘Healthier You” handbooks helped set targets and record progress.

  • What was most interesting about Harry’s talk was that it was aimed at motivating diabetologists and MDs to believe that obesity can be successfully tackled within the normal NHS clinical setting.

• David Paul was part of the DiRECT trial and made a memorable presentation  at last year’s Diabetes UK. He reminisced that his four shakes a day “was hellish but absolutely worth it”. He initially lost 56lbs in three months. He was motivated to address his type 2 diabetes. He quoted a famous UK playwright as saying “having type 2 diabetes and not doing anything about it is equivalent to deciding to commit suicide slowly”.

  • The key question is whether participants will ultimately regain weight. In Mr. Paul’s case it is the sheer fear of getting diabetes again that makes him keep fighting. He noted that trying to keep weight off is a “daily battle that is far harder than taking it off in the first place”. So he signed up to train for a 1000 mile bike ride. From the perspective of a middle aged overweight man, “Lycra is an unforgiving fabric”, he said in passing. Exercise seems to be the key to weight maintenance.

  • The role of practice nurses is important in mentoring patients during weight loss and weight maintenance. Mr. Paul felt accountable to his “lovely but scary” practice nurse Maureen. It seems that keeping his weight down is done day by day and meal by meal.

• Joseph McSorley was a DiRECT participant. His experience was that once you get past the first couple of days, the low calorie phase wasn’t too difficult. At the end of 12 weeks, he had lost 20kg, which made a huge difference to his appearance.

  • Food reintroduction was the hardest part of his experience. Getting expert support for food reintroduction was critical to maintaining weight loss. This was when the trial really started for him. He minimized alcohol intake, managed portions. Exercise was also key for weight maintenance. There was also a contingency plan discussed should weight start to creep up again.

  • Mr. McSorley has benefitted in many ways. In addition to stopping medications, he experiences better sleep, has stopped snoring, and his allergies, psoriasis and dandruff have all resolved as a result of the trial.

7. Patient & Provider Interviews Reveal Early Apprehension Toward DiRECT Protocol, Allayed into High Satisfaction; Importance of General Practitioners Highlighted

In a psychological study as part of DiRECT, participants and HCPs shared their fears in advance of the trial and their great satisfaction afterwards. HCPs were concerned about stopping medication and the initial dietary (‘non-food’) intervention, as well as the burden on their practice (delivering dietary intervention). After the fact, HCPs were very satisfied by participating and found that the execution wasn’t too challenging. Patients were concerned about the food reintroduction phase and were hugely motivated by going off medications and lessening the stigma of having diabetes, but they needed support to reintroduce food and keep the weight off.

To be sure, GPs played an important role in DiRECT, although at signup they didn’t know if they would be randomized to the control or intervention group. Notably, GPs had to review potential participants, facilitate the delivery of intervention (by a nurse or dietitian), and withdraw/recommence medications as appropriate.

• The DiRECT team interviewed participants and HCPs at key points throughout the trial, in order to understand the practicality of implementing the intervention. Engagement, beliefs, relationship and trust were all interlinked.

• Dietitians were initially concerned by rapid weight loss and many expected it to fail. HCPs were worried about taking people off medications and also thought that a low calorie diet would make them unwell. HCPs were reassured by the dietician support and also felt that it enabled a low impact on the practice.

• Food reintroduction is difficult in comparison to the initial weight loss phase, because it is more difficult and also led to weight regain. The weight maintenance phase is also problematic.

• Patients commented that they have changed their attitude towards food, and the fear of going back on medication is a strong motivation to keep weight off.

• Participants and HCPs needed the right amount of support to continue motivation and to implement DiRECT.

• All GPs said that they were extremely satisfied by participating. They found implementation and stopping/starting medications easy, and they spent only a medium effort keeping patients engaged (this was the most difficult aspect). Many GPs have now modified their practice to focus on earlier weight loss in diabetes.

• The DiRECT protocol should be an option for all practices for the treatment of diabetes. But first, more dietetic support will be required. There also needs to be a therapeutic trial of medicine withdrawal, and the confidence to target >15kg weight loss. There also needs to be a focus on weight maintenance for the long term. In England and Scotland, more dietitians are coming, and a broader trial of 5,000 patients is planned.

Questions and Answers

Q: Many people have dieted many times before and they haven’t succeeded. Is this group somehow different?

A: We do have the data on previous diet attempts, but haven’t analyzed it yet. But we suspect that most of our participants have tried numerous times to lose weight. There is nothing different about this group of people.

Q: Was there variation in attitudes by age cohorts?  What happened to the anxiety of HCPs over the diet?

A: Most respondents were in their 40s and 50s. There wasn’t much age difference but there was a difference between those who were working versus those retired. The HCPs were persuaded after the fact, because the protocol didn’t lead to any adverse effects.

Q: Our GPs seem reluctant to stop metformin, regardless of A1c. They say “keep taking the tablets.”

A: We are all familiar with trials of introducing medications. But we should think about trials where we withdraw medication, such as DiRECT. It’s potentially threatening for doctors. They were actually most worried about withdrawing blood pressure medication. In the earlier trials, they had problems with hypotension, so they removed all drugs at the outset, although maybe we would remove them sequentially today.

Q: What about withdrawing medication – this drives cost/benefit because of the cost of drugs. Which are the best ones to withdraw first? What about insulin?

A: The best ones to withdraw first are the ones with side effects. There is some early work published (in posters) in taking this approach on patients who take insulin.

Comment: For withdrawal of medication, we just used NICE guidelines in reverse. Removing those first that we added last, as their condition improved.

Q: Will GPs feel confident in telling their patients that diabetes can be reversed?

A: On day one of diagnosis, we have to tell patients that diabetes is serious, but that it can be sent into remission. We are developing a remission information prescription that can be integrated into GP systems.

Q: What do we do with patients who have had diabetes for 10-15 years?

A: It does seem reversible up to even 20 years in a few patients, but we don’t know how to characterize those people. But the benefits of weight loss are so great, we shouldn’t withhold this treatment for anyone.

Q: Can we use real food in the low calorie intervention?

A: Using non-food products initially has a behavioral advantage because it breaks the participants relationship to food, so it can be redefined later in the process.

Q: How many patients were served by each practice nurse?

A: Between five and ten patients per practice nurse. Diabetes remission hubs (a specialist service) might be a good idea for smaller patient numbers.

Q: Can you deliver DiRECT in a group setting or via technology?

A: We should indeed be considering groups, and there is some work underway in this area. Many groups are looking at digital support, but the evidence is not in yet.

Q; Did you have to recruit extra dietitians? Presumably the benefits will allow us to afford more resources.

A: We were able to work within the available resources. There was enthusiasm to divert dietitians to a new and exciting area. But we will need a lot more in the future and funding will have to be provided. Cost per QALY will be favorable to making this happen. In future, we need to think about supporting remission as well as treatment for type 2 diabetes.

Q: Shouldn’t peer support be included in the program?

A: Some participants said they would have valued some kind of group forum for those taking part. It wasn’t in the trial but should be considered.

Prof. Taylor: When we did the interviews, I suddenly realized that the input of the spouse is absolutely critical. We might be missing a trick by not directly involving the spouse in the program. It’s not easy being a spouse during the weight loss.

-- by Ann Carracher, John Close, and Kelly Close