A nuanced picture of illicit drug use in 17 Italian cities through functional principal component analysis of temporal wastewater data

Background: Wastewater-based epidemiology is a novel approach in drug use epidemiology, which may provide more objective estimates of illicit drug use in a community. Functional data analysis (FDA) is a statistical framework specifically developed for analysing curves. We applied FDA to study weekly temporal patterns in wastewater curves for six different drugs in Italy. Methods: Wastewater samples were collected over seven consecutive days in November 2013, from the inlet of 17 wastewater treatment plants in 17 Italian cities. The weekly temporal features of the drug loads throughout the week were extracted using functional principal component analysis (FPCA), obtaining functional principal component (FPC) curves and corresponding FPC score variables. The FPC score variables were used as outcome variables in linear regression analyses. Results: The most important weekly features of the drug loads were captured by the first three FPCs. The first FPC represented the general level of drug in the wastewater, while the second and third FPCs represented the discrepancy between the weekend peak and midweek level, and the weekend peak timing respectively. Cannabis was the predominant drug in the Italian wastewater, while ecstasy (MDMA) was the drug with the highest discrepancy between the weekend peak and midweek level. The Italian cities showed different patterns of drug use depending on several characteristics of the cities. Conclusion: FPCA extracted detailed features of the weekly temporal patterns of the use of drugs derived from the wastewater analysis. This may help in understanding and monitoring the profile of drug use in a specific community.