Vocal changes appear to occur in individuals with type 2 diabetes mellitus (T2DM) versus those without diabetes, according to a study published online in the December issue of Mayo Clinic Proceedings: Digital Health.
Jaycee M. Kaufman, from Klick Inc. in Toronto, and colleagues investigated the potential of voice analysis as a prescreening or monitoring tool for T2DM. Voices were compared for 267 participants without diabetes (79 women and 113 men) and those with diabetes (18 women and 57 men).
The researchers observed significant differences between voice recordings of men and women with and without diabetes, both for the entire dataset and in an age-matched and body mass index (BMI)-matched sample. Overall, pitch and pitch standard deviation achieved the highest predictive accuracy. For women, relative average pertubation jitter was also significant, as were intensity and 11-point amplitude perturbation quotient shimmer for men. When combining these features with age and BMI, the optimal prediction models achieved accuracies of 0.75 for women and 0.70 for men in the age-matched and BMI-matched sample.
