Alcohol and cocaine co-consumption in two European cities assessed by wastewater analysis

The quantitative determination of urinary biomarkers in raw wastewater has emerged in recent years as a promising tool for estimating the consumption of illicit drugs, tobacco and alcohol in a population and for comparing local and temporal trends. In this study, a three-year monitoring campaign (2012-2014) was conducted to compare alcohol and cocaine use in two European cities (Santiago de Compostela, Spain, and Milan, Italy) by wastewater analysis. Ethyl sulphate and benzoylecgonine were used, respectively, as biomarkers of ethanol and cocaine consumption and cocaethylene as an indicator of co-consumption of both substances. Biomarkers were measured using liquid chromatography-tandem mass spectrometry and concentrations were converted to rates of consumption using specific correction factors. Results were statistically compared in terms of geographic and temporal tendencies. Alcohol intake was significantly higher in Santiago than in Milan (13.6L versus 5.1L ethanol/1000 people day, averages). Cocaine use was higher in Milan than in Santiago de Compostela (800 versus 632 mg/1000 people day, averages). A significant higher consumption of both alcohol and cocaine was observed during the weekends (~23-75% more than on weekdays) in both cities. In terms of years, slight changes were observed, but no clear trends as representative of the whole year could be identified because of the limited number of days sampled. Co-consumption was evaluated using the cocaethylene/benzoylecgonine ratio, which was higher during the weekend in both cities (58% in Santiago and 47% in Milan over the non-weekend day means), indicating a greater co-consumption when cocaine is used as a recreational drug. Wastewater-based epidemiology gave estimates of alcohol and cocaine use in agreement with previous wastewater studies and with recent European surveillance and prevalence data, and weekly profiles of use and preferential patterns of consumption could be plot.