ROCKVILLE, Md. — Smoking has long been proven to negatively affect people’s overall health in multiple ways.
A study by scientists at Insilico Medicine, a biotechnology firm, set out to determine biological age differences between smokers and non-smokers, and to evaluate the impact of smoking using blood biochemistry and recent advances in artificial intelligence.
According to the study, smokers demonstrated a higher aging ratio, and both male and female smokers were predicted to be twice as old as their chronological age as compared to nonsmokers. The results were carried out based on the blood profiles of 149,000 adults.
By employing age-prediction models developed by supervised deep learning techniques, the study analyzed a number of biochemical markers, including measures based on glycated hemoglobin, urea, fasting glucose and ferritin.
Other findings suggested that deep learning analysis of routine blood tests could replace the current unreliable method of self-reporting of smoking status and evaluate the influence that other lifestyle and environmental factors have on aging.