The Evolution of Diabetes Treatments

Up until the medieval ages, diabetes was considered a death sentence as there was no cure for it. People who displayed signs of excessive hunger, thirst, and urination got their urine tested with the help of ants. A diabetic patient’s urine would attract ants and this test was employed to confirm the diagnosis of diabetes.

In 1889, Oscar Minkowski and Joseph Von Mering, famously known for their experiment on a dog by extracting its pancreas were the pioneers in the discovery of the correlation between the pancreas and diabetes. About 20 years later, Sir Edward Albert Sharpey-Schafer discovered that the lack of insulin was the prime cause of diabetes. This ground-breaking discovery impelled the unfolding of treatments for diabetes.


Another 10 years fast-forward, Frederick Banting and his student, Charles Best isolated insulin from the pancreas of healthy dogs and administered it to dogs from which pancreas was previously extracted. Thereby, insulin started to be used in human patients with diabetes. There have been myriad advancements in the delivery of insulin and its formulations since then. Today, insulin is extensively used among patients with type I and type II diabetes.

The next consequential discovery in the field of diabetic medications was the antihyperglycemic agent isolated from the French lilac: ‘Phenformin’. This agent belonged to the class of Biguanides. However, the increased susceptibility of lactic acidosis which accompanied the use of Phenformin led to its withdrawal from the market. Later, ‘Metformin’ replaced this agent and has since then been the most widely used antihyperglycemic agent in the treatment of diabetes mellitus.

Approximately 5 years after the discovery of Phenformin, Marcel Janbon observed the hypoglycemic activity of the antibiotic para-amino-sulfonamide-isopropyl-thiadiazole while treating his typhoid patients. This led to the identification of sulphonylureas as another class of antihyperglycemic agents which stimulated the pancreas for release of insulin. Thus arrived, tolbutamide, chlorpropamide, glyburide, glipizide, and glimepiride.

The US pharmaceutical market launched ‘Troglitazone’ - a thiazolidinedione in 1996. These agents enhanced skeletal muscle sensitivity towards insulin and decreased glucose production. Yet, it was cast out of the market owing to its predisposition towards hepatic failure. In due course, pioglitazone and rosiglitazone were introduced. Being associated with fluid retention; it is still used with caution in patients with congestive heart failure.

Down the road, more antihyperglycemic agents like the alpha-glucosidase inhibitors, meglitinides, GLP-1 agonists, DPP-4 inhibitors, amylin agonists, bromocriptine, colesevelam, and the SGLT-2 inhibitors have spawned. Over the past 90 years, 11 different categories of antihyperglycemic agents to treat diabetes have been founded.


As researchers in the field of diabetic care worked towards reversing and decelerating the disease progression, advancements in technology focused on helping patients with diabetes improve their lives. Today, patients have continuous blood glucose monitors handy that help keep track of their blood glucose levels whenever required. Innovations like the artificial pancreas system, virgin beta cells that may restore pancreas insulin production function, and automated insulin pumps are some major leaps by technology towards progress in medicine.

Dr Jothydev Kesavadev in 2018 during an interview with the RSSDI said, “If you look at technological advancements in diabetes of the last 50 years, more advancements in diabetes technologies have happened in the last five years compared to the past five decades and artificial intelligence plays a pivotal role in it.”

Evidence suggests that AI today is revolutionizing diabetic care in four areas: decision-support tools for clinicians, patient self-management aids, predictive population risk stratification, automated screening and detection of diabetic retinopathy and macular oedema. These technologies ensure accuracy, efficiency, satisfaction and ease of use. Diabetic Retinopathy is the major cause of secondary blindness. However, if screened and detected early retinopathy can be treated. With the help of deep learning-based Artificial Intelligence, such cases can be detected easily with a sensitivity and specificity over 90%.

Using AI, physicians can now predict medication adherence threshold and hence customize diabetic medications to optimize health outcomes. Several breakthroughs in deep learning algorithms also have helped diabetic patients achieve better blood glucose control, minimised hypoglycemic episodes and improved glycated haemoglobin levels. (Mullan et al - 2019)


The Apple Watch can now flag signs of diabetes with the help of AI. The nerve damage complication caused by diabetes results in irregular heart rate. And this less obvious link is the secret to this signature wearable prediction of diabetes.

As diabetics are prone to develop hypoglycemia due to the administered insulin, these patients have to constantly do blood tests, analyze every meal and live in constant fear of seizures. The Bigfoot Biomedicals uses quantitative trading algorithms that predict blood sugar and automatically delivers insulin in the right amount at the right time. They created an effective automated insulin delivery system.

The Virta Application works by guiding patients by providing nutritional recommendations. It helps rely on fat as a source of energy which simultaneously brings down blood glucose levels. Virta’s nutrition therapy works by collecting data points from several patients each day and then deciding the appropriate nutrition that will keep each patient on track. A highly complex protocol which is a combination of different variables and permutations helps Virta deliver outstanding outcomes that last.


Diabetic patients frequently visit their healthcare practitioners pertaining to complications in their illness or for regular checkups. A comprehensive history for such patients ensures that nothing is missed out and that the best care is provided.

Arintra is an AI-enabled history-taking tool that was launched in March 2019 developed by doctors for the doctors. It helps capture comprehensive patient history before consultation and hence reduces the doctor’s workload. With the help of its clinical natural language processing abilities, raw text like the doctor’s notes and patient’s messages can be processed and extracted as clinical entities. Arintra’s AI inference engine analyses patient data to generate lab recommendations, procedure recommendations and clinical diagnosis.

From remote monitoring tools to wearable devices with inbuilt data processing abilities, technology is certainly improving healthcare at a rapid pace. The improvements brought to patient care by using these fundamentally new tools in such a short time period only paints a glowing picture of future of healthcare. These technologies promise enormous benefits and drive healthcare towards improved accuracy and efficacy.