Prediction of an individual's smoking habits using saliva microbial signatures
Principal Investigator
Name
Manfred Kayser
Institution
Erasmus MC Medical Center University Rotterdam
Position Title
Professor, Head Department
Email
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCO-647
Initial CDAS Request Approval
Jul 17, 2020
Title
Prediction of an individual's smoking habits using saliva microbial signatures
Summary
The impact of smoking in human health together with its prevalence in the general population necessitates the inclusion of smoking habits as a cofounder in medical and public health research. However, information on the study participants’ smoking habits is not always available, and where it is, it is not necessarily reliable since individuals may under report smoking habits. An approach for predicting an individual’s smoking habits can overcome these limitations.
Previous studies have reported models using differently human methylated DNA CpG markers in whole blood for the prediction of an individual’s smoking habits. However, blood is not always a suitable sample-type in research surveys due to the level of invasiveness during sample collection. Saliva is an easily accessible alternative and is in direct contact with the compounds resulting from tobacco burning. However, available human epigenome-wide methylation data in saliva are insufficient for the discovery of smoking-related markers in the general population to be used in a prediction model.
The saliva microbiota is known to be a good indicator of tobacco smoking habits, as previously showed in various studies. Smoking habits are associated with changes in the oral cavity such as impaired host immune response, decrease of oxygen tension, lower buffer capacity of saliva (resulting in lower saliva pH), change in saliva proteins concentrations, and increase virulence of certain bacterial species, among others. These alterations apply selective pressure on the growth and proliferation of the saliva microbiota which ultimately differ among individuals with different smoking habits.
In this study, we aim to build a validated prediction model based on a large saliva microbiome dataset that can be applied to the general population. For that, we will use publically available 16S rRNA gene saliva microbiome data from different studies.
Previous studies have reported models using differently human methylated DNA CpG markers in whole blood for the prediction of an individual’s smoking habits. However, blood is not always a suitable sample-type in research surveys due to the level of invasiveness during sample collection. Saliva is an easily accessible alternative and is in direct contact with the compounds resulting from tobacco burning. However, available human epigenome-wide methylation data in saliva are insufficient for the discovery of smoking-related markers in the general population to be used in a prediction model.
The saliva microbiota is known to be a good indicator of tobacco smoking habits, as previously showed in various studies. Smoking habits are associated with changes in the oral cavity such as impaired host immune response, decrease of oxygen tension, lower buffer capacity of saliva (resulting in lower saliva pH), change in saliva proteins concentrations, and increase virulence of certain bacterial species, among others. These alterations apply selective pressure on the growth and proliferation of the saliva microbiota which ultimately differ among individuals with different smoking habits.
In this study, we aim to build a validated prediction model based on a large saliva microbiome dataset that can be applied to the general population. For that, we will use publically available 16S rRNA gene saliva microbiome data from different studies.
Aims
- Obtain publically available 16S rRNA gene saliva microbiome data and associated metadata. Particularly: sex; age; ethnicity; smoking-related metadata (cigarettes per day, years smoking, years since quitting for former smokers); and cotinine levels in blood.
- Perform diversity analysis between different smoking groups to investigate whether significant differences in the saliva microbiota exit.
- Perform differential abundance analysis between different smoking groups to identify particular taxa associated with a specific smoking category.
- Build a prediction model of an individual's smoking habits using saliva microbial signtaures.
Collaborators
No external collaborators.