Around 86 million people in the US have prediabetes, putting them at higher risk of being diagnosed with diabetes in the future. But a research team led by investigators from the University of Michigan Medical School has created a “precision medicine” model that they say could aid diabetes prevention in high-risk individuals by identifying the best possible treatment strategies.
Lead author Dr. Jeremy Sussman, an assistant professor of general medicine at the U-M Medical School, and his team say the model could also be adapted for use in individuals at high risk of other diseases.
To create their model – the details of which are published in The BMJ – the researchers assessed data of 3,060 participants who were part of the Diabetes Prevention Program (DPP).
The DPP is a gold-standard clinical trial that involved the random assignment of people at high risk of diabetes to treatment with a placebo, a lifestyle-modification program or metformin – a drug commonly used to treat people with type 2 diabetes.
All the participants studied had prediabetes – determined by having at least two abnormal results on fasting blood sugar tests – and a high body mass index (BMI). The majority of participants also had a family history of diabetes, while more than a third were African-American or Latino – ethnicities known to be at increased risk for diabetes.
The team assessed 17 risk factors for diabetes and identified seven that appeared to be most useful for predicting an individual’s risk of developing the condition:
- More than 29 million people in the US have diabetes
- Diabetes is the seventh leading cause of death in the US
- Diabetes costs the US around $245 billion annually.
- Fasting blood sugar
- Long-term blood sugar (A1c levels)
- Total triglyceride level
- Family history of high blood sugar
- Waist measurement
- Waist-to-hip ratio.
The team created a scoring scale by allocating points to each of the seven risk factors identified and applied this scale to the study participants.
They found that more than half of participants who received scores in the highest quarter were at risk of developing diabetes over the next 3 years, while less than 1 in 10 participants in the lowest quarter were at risk of the condition during the following 3 years.
The researchers then used the model to identify which treatments – metformin or lifestyle modifications, including exercise and weight loss – were most effective for reducing the risk of diabetes among participants.
They found that for people with the highest risk of diabetes, metformin was very effective, reducing their risk of the condition by 21%. However, metformin was revealed to have no benefit at all for the remaining participants.
Lifestyle modifications were also found to be highly effective for patients at the highest risk of diabetes, reducing their risk of developing the condition over the next 3 years by 28%. Lifestyle modifications were less effective for patients at lower risk of diabetes, but they still reduced their likelihood of the condition by 5%.
Co-author Dr. Rob Hayward, a professor of medicine and public health at U-M, says these findings suggest that, although the overall benefit in a clinical trial might be moderate, the response to treatment in patients at high risk of disease is likely to vary significantly. He adds:
“In this instance, a more rigorous analysis of this important trial found that three quarters of patients took a drug with non-trivial side effects without receiving any benefit, but that the average benefit found in the trial also greatly underestimated the benefits for those at very high risk of developing diabetes in the next 3-5 years.”
The researchers note that, although exercise and weight loss did not offer large benefits for participants at lowest risk of diabetes in the DPP study, such lifestyle modifications are important for other areas of health.
The researchers believe the newly created model could be an effective tool for doctors treating patients who are at high risk of diabetes, and it can even be modified to help patients at high risk of other conditions.
“We think this approach should be broadly applicable, since one of the main determinants of any patient’s likelihood of benefiting from a therapy is their risk of having the bad outcome that we are trying to prevent,” says study co-author Dr. David Kent, a professor at Tufts University in Medford, MA.
“It is poorly appreciated how many patients receive treatments unnecessarily – when the possibility of benefit is very low, and may well be exceeded by the burdens of treatment,” he adds, noting that the model could also be useful for forthcoming clinical trials. “If these types of analyses were incorporated routinely into trial design, we believe we would have a much clearer understanding of this issue.”
While a primary strength of the team’s findings is the use of data from the gold-standard DPP trial, there are some limitations to be considered.
For example, the researchers note that their model was not validated externally. In addition, their study would have carried over the limitations of the DPP study, which include a short-term follow-up of patients.
Last month, Medical News Today reported on a study in the journal Interface, which detailed the creation of a model that predicts public response to disease outbreaks.