Binary classification of users of electronic cigarettes and smokeless tobacco through biomarkers to assess similarity with current and former smokers: machine learning applied to the population assessment of tobacco and health study.

Introduction:
This article presents a study that aimed to evaluate the population-wide effects of reduced-exposure tobacco products, such as electronic cigarettes and smokeless tobacco, on health. The study utilized data from the Population assessment of Tobacco and health (PaTh) study, a large-scale longitudinal study in the United States. Machine learning techniques were applied to classify participants based on their biomarkers of exposure and potential harm, and to investigate whether users of electronic cigarettes and smokeless tobacco were classified as current or former smokers.

Key Points:

* The study used data from the Population assessment of Tobacco and health (PaTh) study, a large-scale longitudinal study in the United States.
* Machine learning techniques were applied to classify participants based on their biomarkers of exposure and potential harm.
* Users of electronic cigarettes and smokeless tobacco were classified as current or former smokers.
* The classification models for biomarkers of exposure and potential harm both had high model accuracy.
* More than 60% of participants who used either one of electronic cigarettes or smokeless tobacco were classified as former smokers.
* Fewer than 15% of current smokers and dual users were classified as former smokers.
* Participants who used electronic cigarettes or smokeless tobacco had a lower prevalence of cardiovascular and respiratory diseases compared to current smokers.

Main Message:
The study suggests that users of electronic cigarettes and smokeless tobacco are likely to be similar to former smokers in their biomarkers of exposure and potential harm. This finding suggests that using these products helps to reduce exposure to the harmful constituents of cigarettes and are potentially less harmful than conventional cigarettes. however, further studies are needed to verify the risk reduction achieved by switching to these products and to evaluate the long-term effects of their use on health.

Ohara h, Ito S, Takanami Y. Binary classification of users of electronic cigarettes and smokeless tobacco through biomarkers to assess similarity with current and former smokers: machine learning applied to the population assessment of tobacco and health study. BMC public health. 2023;23(1):589. doi:10.1186/s12889-023-15511-3