The internet and the media are filled with discussions of risk and phrases such as “high risk” and “low risk”. People with diabetes are described as being at high risk for serious long-term health problems. In addition, diabetes care is all about using therapies to reduce the risk of these problems. Unfortunately, concepts like risk, high risk, low risk and risk reduction are often misunderstood.
When discussing risk in a medical setting, the word usually means incidence: the percentage of people who will develop a problem over time. For example, about 2-4 out of 100 (i.e., 2-4%) middle-aged people with diabetes will have a heart attack or stroke every year. The risk or incidence of heart attack or stroke is therefore 2-4% per year, and over a 10-year period, it would be somewhere between 20-40%. This doesn’t mean that every single middle-aged person with diabetes has this risk of a heart attack or stroke. Rather, it means that the average middle-aged person with diabetes is expected to have this risk. In comparison, the incidence of a heart attack or stroke in an average middle-aged person without diabetes is about half of this: 1-2% per year or 10-20% over 10 years. Diabetes, therefore, more or less doubles the risk of a heart attack or stroke.
The actual risk or incidence of heart attack or stroke clearly depends on other factors, such as age or whether someone has had a previous heart attack or stroke. The younger and the healthier a person with diabetes is, the lower their expected risk or incidence. Medicine conventionally considers an annual risk of incidence of more than 2% per year (20% over 10 years) to be high risk.
Based on the above, therapies that have been proven to reduce the risk of an outcome actually reduce its expected incidence. For example, GLP-1 receptor agonist drugs have been clearly proven to reduce the incidence of heart attacks or strokes by about 15%. This means that if a drug like semaglutide is prescribed to someone whose expected incidence of a heart attack or stroke is 40% over 10 years, it would change their expected incidence from 40% to 34% over 10 years (which is 15% lower than 40%). Once again, this doesn’t mean that every person getting this drug will have the same benefit. It means that the average person who is prescribed a GLP-1 receptor agonist will, on average, have this benefit.
Unfortunately, medicine has no way of predicting whether an individual person will or will not benefit from a drug or have a particular side effect. The optimal therapies for one person will never be identical to those for another. Fortunately, there is a growing list of well-studied, effective therapies. That means that patients and health care providers can together choose the therapies that are best for them, based on the best evidence and personal needs.
About the author
Hertzel Gerstein
Hertzel is an endocrinologist and professor at McMaster University who is in high demand as a speaker, advocate, and educator on diabetes-related topics. His research focuses on using large, international randomized trials to identify and test new ways of preventing type 2 diabetes, reducing serious health outcomes like strokes and death, and achieving type 2 diabetes remissions.