New Study: If You Hear This, It Could Be a Diabetes Symptom
Recent research introduces a revolutionary approach to diabetes screening that could mark a significant turn in how it is detected.
Diabetes is a reality for millions worldwide, and the quest for more accessible detection methods has never been more critical. Approximately 537 million adults globally live with this condition, which is projected to rise to an astonishing 643 million by 2030, according to data from the International Diabetes Federation. Traditionally, medical professionals have used tests like the Hemoglobin A1c, which gauges blood sugar levels over a few months, and the fasting blood glucose test, which requires fasting overnight. There’s also the oral glucose tolerance test involving a sugary drink followed by several blood sugar tests. While reliable, these methods can be cumbersome and not easily accessible to all.
However, a fascinating study in the Mayo Clinic Proceedings: Digital Health journal, published in October 2023, is making waves with a unique approach to diabetes screening. This research introduces an artificial intelligence (AI) method that detects subtle indicators of Type 2 diabetes in a way you might not expect: The sound of your voice. This breakthrough could open the door for simpler, quicker screening processes, changing the game in how we monitor and manage this prevalent health challenge.
The details of the study
In this study, researchers recruited 267 individuals, including those with and without Type 2 diabetes, to participate in an experiment using their smartphones. They recorded a specific phrase six times daily for two weeks, accumulating over 18,000 voice recordings. These recordings were then analyzed for subtle acoustic features like pitch and intensity changes—which typically go unnoticed by the human ear.
Jaycee Kaufman, MSc, scientist and lead author of the study from Klick Labs in Toronto, Canada shares in a press release, “Our research highlights significant vocal variations between individuals with and without Type 2 diabetes and could transform how the medical community screens for diabetes.” She emphasizes that conventional methods for detecting diabetes can often be time-consuming and costly, obstacles that this technology aims to overcome.
Using a computer model to examine these vocal differences and including other factors such as age and Body Mass Index (BMI), they achieved impressive success rates in identifying diabetes—89% accuracy in women and 86% in men. This breakthrough in research indicates a significant step forward in using voice patterns for diabetes detection.
So, why do these vocal changes occur?
According to the study authors, long-term elevated blood sugar levels, typical in Type 2 diabetes, can cause conditions known as peripheral neuropathy and myopathy, which essentially mean damage to the body’s nerve and muscle fibers. Such damage, particularly in the larynx—your body’s very own sound-making device—can lead to various voice disorders and even swallowing difficulties. Some individuals with diabetic neuropathy may find themselves dealing with hoarseness or a strained voice.
The study also uncovered that the impact of Type 2 diabetes on vocal characteristics differs between men and women. Men may experience muscle weakness that can change the quality of their voice, while women may develop swelling in their vocal cords, leading to a lower pitch—a finding that supports previous research.
The study, while a significant milestone in diabetes detection, acknowledges its limitations and the need for further research, particularly in broadening the participant base. The researchers are keen to explore how this voice analysis technology can be applied to other health areas, such as prediabetes, hypertension, and women’s health.