Long-term trends in tobacco use assessed by wastewater-based epidemiology and its relationship with consumption of nicotine containing products.

Measurement of population tobacco use via wastewater-based epidemiology (WBE) provides objective data to evaluate the efficacy of tobacco control strategies. However, current WBE tobacco-use estimates based on nicotine metabolites (cotinine and hydroxycotinine) can be masked by use of non-tobacco nicotine-containing products. To better understand nicotine and tobacco use, we analysed tobacco-specific biomarkers, anabasine and anatabine, as well as nicotine metabolites, cotinine and hydroxycotinine, in wastewater samples collected for 6 weeks per year over 6 years (2012-2017) from an Australian wastewater treatment plant serving approximately 100,000 people. Population-normalised mass loads were used to estimate tobacco and nicotine use trends and were compared with surveys and taxation statistics. Significant annual declines were observed for anabasine, anatabine, cotinine and hydroxycotinine of -3.0%, -2.7%, -2.4%, and -2.1%, respectively. The results corresponded with the annual declining trends reported from surveys (-5%) and taxation statistics (-4%). Significant annual decreases in the ratios of anabasine to cotinine (-1.2%) and anatabine to cotinine (-1.0%) suggested a relative increase in the use of non-tobacco nicotine products at the same time that tobacco use was declining. Monitoring tobacco use with anabasine and anatabine removed influence from nicotine-containing products, showing larger reductions in this Australian city than via nicotine biomarkers, whilst also demonstrating their suitability for monitoring long-term trends.