Assessing secondhand and thirdhand tobacco smoke exposure in Canadian infants using questionnaires, biomarkers, and machine learning

Abstract

Background: As smoking prevalence has decreased in Canada, particularly during pregnancy and around children, and technological improvements have lowered detection limits, the use of traditional tobacco smoke biomarkers in infant populations requires re-evaluation.

Objective: We evaluated concentrations of urinary nicotine biomarkers, cotinine and trans-3'-hydroxycotinine (3HC), and questionnaire responses. We used machine learning and prediction modeling to understand sources of tobacco smoke exposure for infants from the CHILD Cohort Study.

Methods: Multivariable linear regression models, chosen through a combination of conceptual and data-driven strategies including random forest regression, assessed the ability of questionnaires to predict variation in urinary cotinine and 3HC concentrations of 2017 3-month-old infants.

Results: Although only 2% of mothers reported smoking prior to and throughout their pregnancy, cotinine and 3HC were detected in 76 and 89% of the infants' urine (n = 2017). Questionnaire-based models explained 31 and 41% of the variance in cotinine and 3HC levels, respectively. Observed concentrations suggest 0.25 and 0.50 ng/mL as cut-points in cotinine and 3HC to characterize SHS exposure. This cut-point suggests that 23.5% of infants had moderate or regular smoke exposure.

Significance: Though most people make efforts to reduce exposure to their infants, parents do not appear to consider the pervasiveness and persistence of secondhand and thirdhand smoke. More than half of the variation in urinary cotinine and 3HC in infants could not be predicted with modeling. The pervasiveness of thirdhand smoke, the potential for dermal and oral routes of nicotine exposure, along with changes in public perceptions of smoking exposure and risk warrant further exploration.

Keywords: Biomarker; Childhood asthma; Cotinine; Secondhand smoke; Thirdhand smoke; Variable importance.

Fig. 1. Questionnaire-determined density plots of cotinine and trans-3’-hydroxycotinine concentrations.

The distribution of the log-transformed urinary cotinine (left) and 3HC (right) concentrations by two questions: “How many cigarettes (on average) are smoked at the home daily in the child’s early life”, and “How many smokers lived at the home during pregnancy?”. Vertical lines reflect cut-points of assumed exposure to very little to no SHS or THS (left), moderate SHS (middle), and regular SHS exposure (right). The dashed lines reflect 0.25 and 30 ng/mL, while dotted lines indicate 0.50 and 60 ng/mL. Density curves were not created for categories with two or fewer participants.

Fig. 2. Cotinine multivariable linear regression model.

Standard regression coefficients (point) and their 95% confidence intervals (line) are displayed for each variable in a model predicting log-transformed urinary cotinine concentration. Variables related to secondhand smoke are shown in red, variables not related to smoking in blue, and variables related to household characteristics in gray. Note that the ingestion of breast milk is an indirect source of nicotine from secondhand or thirdhand smoke, or dietary nicotine sources. Intervals with a point estimate of the regression coefficients that are displayed as a circle are based on bivariate analysis between each predictor and urinary cotinine, while estimates displayed with a triangle reflect estimates from a multivariable model.

Fig. 3. Trans-3’-hydroxycotinine multivariable linear regression model.

Standard regression coefficients (point) and their 95% confidence intervals (line) are displayed for each variable in a model predicting log-transformed urinary trans-3’-hydroxycotinine concentration. Variables related to secondhand smoke are shown in red, variables not related to smoking in blue, and variables related to household characteristics in gray. Note that the ingestion of breast milk is an indirect source of nicotine from secondhand or thirdhand smoke, or dietary nicotine sources. Intervals with a point estimate of the regression coefficients that are displayed as a circle are based on bivariate analysis between each predictor and urinary trans-3’-hydroxycotinine concentration while estimates displayed with a triangle reflect estimates from a multivariable model.

Fig. 4. Measured vs. predicted log-transformed urinary cotinine concentration.

Scatterplots show the relationship between predicted and measured log-transformed urinary concentrations of cotinine (top) and trans-3’-hydroxycotinine (bottom) based on their MLR models (red, dashed line).