Diabetic retinopathy, the number one cause of blindness in working age Americans, requires early identification and treatment in high-risk eyes. Because T2D accounts for 90% – 95% of new cases of diabetes, the international diabetes federation predicts DR to increase significantly as one in nine adults may develop T2D by 2045.¹⁹ Markers of poor cardiometabolic health, a risk factor for diabetes, CV disease, obesity and chronic diseases, was evident in 93.1% of the US adult population between 1999 and 2018.²⁰ Assisting T2D patients towards remission would be a powerful intervention.

Diabetic patients at greatest risk for development or progression of DR include type 1 diabetes (T1D) after 5 years, large hormonal swings such as puberty and pregnancy, uncontrolled hypertension, elevated haemoglobin A1c [HgA1c] levels, dyslipidaemia and large variations in blood glucose.^1,2,17 Primary care physicians are well positioned to manage, and lower risks, of the systemic associated diseases predicted by DR including neuropathy, nephropathy, CV disease, stroke and peripheral arterial disease.¹⁸ Risk factors, duration of disease and diabetes type, T1D versus T2D, form the basis of screening recommendations as per the AAO outlined in Table 1.

Given the growing epidemic of T2D, increasing prevalence of DR is anticipated. Therefore, greater screening at the time of diagnosis may identify DR at earlier stages. The pathophysiology of DR suggests that both neural networks and microvasculature are under assault for years prior to identification of haemorrhages, exudates, new blood vessel growth or ischemic changes.¹ Specialised testing of the retinal cells, including electrophysiological testing, psychophysical tests, dark adaptation, contrast sensitivity, and visual fields reveals abnormalities years in advance of visible microangiopathy changes clearly associated with DR.¹ Early neurodegeneration in DR may contribute to microangiopathy²¹; therefore, the American Diabetes Association (ADA) statement of 2017 was modified to include this understanding: ‘Diabetic retinopathy is a highly specific neurovascular complication of both type 1 and type 2 diabetes’.²² Hyperglycaemia has numerous pleiotropic effects leading to advanced glycation end products, glycocalyx damage, local activation of the renin-angiotensin system, protein kinase C, hexokinase and polyol pathways all adding to increased free radicals and oxidative stress.^{23,24,25,26,27,28} Damage to neural retinal, supportive cells and endothelial microvasculature results in a breakdown of the blood-retinal barrier, a proinflammatory milieu, production of vascular endothelial growth factor (VEGF) and retinal dysfunction with apoptosis, haemorrhage, macular edema and retinal detachment.^29,30

If screening determines retinopathy is present, a basic understanding of DR is helpful to guide patients in need of a referral. Although AI systems in primary care offices accurately determine the need for ophthalmological referral, discussion with the patient affords another opportunity to approach the topic of diabetes remission.¹⁶ Stages of DR are outlined in Table 2, while DR pathophysiology with classical fundus changes is illustrated in Figure 1.

Because T2D has long been viewed as a chronic progressive disease, slowing or preventing DR through the mechanism of T2D remission was not considered until recently. Lifestyle-driven therapies to reduce the insulin resistance and hyperinsulinemia are now having global impact on how one can reframe T2D from a chronic progressive illness to one which can be placed into remission.

Type 2 diabetes remission is happening with frequency and, out of necessity to define T2D remission, three major societies recently published accurately and consistently on the topic. Diabetes Australia Position Paper T2 Diabetes Remission,³¹ the ADA Consensus Report: Definition and Interpretation of Remission in Type 2 Diabetes,³² and Diabetes UK Position Statement for Healthcare Professionals – Remission in Adults With Type 2 Diabetes³³ all have aligned to define ‘remission’ as the most appropriate descriptive term. Haemoglobin A1c (HbA1c) < 6.5% (48 mmol mol⁻¹) measured at least 3 months after cessation of glucose-lowering pharmacotherapy is defined as ‘remission’.^31,32,33

The three primary approaches to achieve this are bariatric surgery, very low calorie diets and significant carbohydrate reduction without intentional calorie restriction.^31,32,33 Active management of medications is essential with all approaches. The authors feel option three is the most sustainable and safest. The primary care physician is positioned to assist and direct this approach.

Detailed articles and reviews on the current landscape of T2D remission have been published by Hallberg et al.,³⁴ Brown et al.³⁵ and Shabib et al.³⁶ A review of medication reduction with diet and lifestyle change was recently published by Cucuzzella, Riley, Isaacs and International Working Group on Remission of Type 2 Diabetes.³⁷ The clinical features best predicting remission in the three reviews^34,35,36 were younger age, early-stage disease, lower number of diabetes medications and greater sustained weight loss. Substantial weight loss of 15% is associated with reduced liver and pancreatic fat and with restoration β-cell insulin function.³⁸ A challenge to sustaining remission is the maintenance of weight loss and lifestyle change. Therefore, we make the case for reducing and eliminating insulin and sulfonylureas, which both promote weight gain.

The data on bariatric surgery and T2D remission is heterogeneous depending on the type of procedure and the patient population. An international consensus statement endorsed by 45 international diabetes associations including the ADA cited numerous RCTs with postoperative follow-up ranging between 1 and 5 years have consistently documented diabetes remission in 30% – 63% of patients.³⁹

The UK DiRECT study used a low-calorie meal replacement (about 825 kcal/day) for 3–5 months followed by a stepped food reintroduction with supportive follow by primary care. The remission rates were 46% after a year and 36% after 2 years.⁴⁰ Remission in the control group was 4% and 3%, respectively. An informative subset was 1 year remission rate of 86% in participants who lost 15 kg or more.⁴¹

The primary care-based DIADEM-1 study from Qatar which replicated the DiRECT approach with an intensive lifestyle therapy in people with T2D achieved 61% diabetes remission after a year compared with 12% for the control.⁴² The study was a cohort of Middle Eastern and North African origin mostly male and younger than the DiRECT participants.

There is a growing body of published literature discussing the clinical application of a low carbohydrate diet (LCD) for T2D. An RCT comparing a low carb Mediterranean diet to a low fat diet reported a remission rate of 15% after a year and 4% at 6 years follow-up compared to 1% and 0% at year one and year six in the low-fat diet group.⁴³

Virta Health (https://www.virtahealth.com/) implements a digitally monitored continuous care intervention (CCI) aiming for nutritional ketosis through an LCD and active management of medication reduction. The non-randomised study of 349 patients with T2D were enrolled, with 262 self-selecting the ketogenic diet, reported T2D remission rates of 25% at year one and 18% at year two.^44,45 Virta Health protocol is to recommend continuing metformin for those patients who can tolerate it, even when remission criteria are otherwise achieved. When including the participants taking no diabetes medication or only Metformin, 60.3% (123/204) of participants achieved HbA_1c below 48 mmol mol⁻¹ (< 6.5%) at 1 year with 54% of completers maintaining HbA1c below 6.5% at 2 years. Adherence to the protocol was good. One hundred ninety-four participants (74% of 262) remained enrolled in the CCI at 2 years as did 68 in the usual care group (78% of 87).

David Unwin published data from an audit of over 6 years in his primary care cohort in the United Kingdom.⁴⁶ This study reflects a real-world practice and therefore is not randomised and does not have a control group but provides important insights into what is possible in primary care when T2D remission options are given to patients and when results are tracked on an ongoing basis. Important findings from the study include: (1) for those choosing an LCD approach nearly half achieved drug-free T2D remission as well as improvements in weight, blood pressure and lipid profiles; (2) an LCD approach reduced HbA1c to below prediabetes threshold in 93% who presented with that condition and (3) patients with the highest HbA1c saw the greatest improvements in their T2D control.

Recent Phase 3 trials show promise that high dose GLP-1 analogues (i.e. semaglutide)⁴⁷ or GLP-1/GIP dual analogues (i.e. tirzepatide)⁴⁸ can be powerful assists in type 2 diabetes mellitus (T2DM) reversal and weight loss although not meeting the definition as stated above by the three consensus groups because this is, of course, added medication. When titrated up to the maximum doses, these medications can promote weight loss and A1c reductions to the levels of bariatric surgery. The question remains though if these medications can ever be removed.⁴⁹

For many years the ADA has been somewhat silent on carbohydrate reduction often suggesting ‘individualized therapy’. But in 2021, their standards of care document clearly made the case for carbohydrate reduction:

The report builds on their Nutrition Therapy for Adults with Diabetes or Prediabetes: A Consensus Report 2019 which states:

The European Association for the Study of Diabetes (EASD)⁵² and Diabetes Canada⁵³ have also acknowledged carbohydrate reduction as a powerful method in diabetes care.

The often cited UK Prospective Diabetes Study (UKPDS) demonstrated higher HbA1c is associated with more diabetes complications, morbidity and mortality.⁵⁴ Yet globally we are failing to achieve these goals even with new and often expensive medications. An important trial, the Action to Control Cardiovascular Risk in Diabetes (ACCORD) was initiated with a goal of intense pharmacologic therapy to lower HbA1C. The trail was stopped early due to adverse outcomes, thus giving caution to this approach. Intensive medical treatment carried an increased risk of all-cause mortality, a 35% increased risk of CV mortality, and a greater risk of hypoglycaemic events and weight gain of 10 kg compared to those on standard insulin therapy.⁵⁵

The ACCORD is not alone. Other high-quality large trials using medications to achieve tight glycaemic targets did not demonstrate reductions in heart disease or in overall mortality.^{56,57,58,59,60,61} These consistent findings from highly funded and well-conducted trials make the case for an alternative approach to treating people with T2D with a focus on addressing insulin resistance.

Type 2 diabetes is, by definition, an inability to tolerate carbohydrate. The literature and clinical experience support therapeutic carbohydrate reduction; lifestyle changes to include exercise, stress reduction and sleep; and accompanying medication reduction. Large practice gaps exist in bringing the science and the experience to clinicians and diabetes educators. Expert panel reports such as the recent Lancet Commission on Diabetes do not make effort to close the gaps as the only mention of carbohydrate in the 63-page report has to do with dosing insulin.⁶² Primary care physicians must bridge this gap.

Defining LCDs has also been an ongoing challenge as described in the article Adapting Medication for Type 2 Diabetes to a Low Carbohydrate Diet:³⁷

In some reported studies, an LCD has included up to 45% of daily calories from carbohydrates. For an individual consuming 2500 calories a day, this would be 280 g of carbohydrates. An LCD by some definitions comprises less than 130 g of digestible carbohydrates per day which is less than 50% of the average daily intake in the United Kingdom and the United States. To fully address insulin resistance and promote T2D remission reduction to levels below 50 g of digestible carbohydrates a day are often needed. Digestible carbohydrate is defined as simple sugars and complex carbohydrates such as starch, which is digested to glucose; this contrasts with fibre, which is a carbohydrate that is not digested or is only partly digested with the aid of intestinal bacteria. Recommended food choices on a lower carbohydrate meal plan include (1) non-starchy vegetables, (2) protein-containing foods such as fish, meat, poultry and eggs, (3) natural fats such as olive oil and butter and (4) foods that naturally contain fats, fibre and/or protein such as nuts, olives, and avocado. Sugar and refined, starchy carbohydrates should be eliminated.