key: cord-0858450-h5ni6tcx authors: Been, Frederic; Emke, Erik; Matias, João; Antonio Baz-Lomba, Jose; Boogaerts, Tim; Castiglioni, Sara; Campos-Mañas, Marina; Celma, Alberto; Covaci, Adrian; de Voogt, Pim; Hernández, Félix; Kasprzyk-Hordern, Barbara; ter Laak, Thomas; Reid, Malcolm; Salgueiro-González, Noelia; Steenbeek, Ruud; van Nuijs, Alexander L.N.; Zuccato, Ettore; Bijlsma, Lubertus title: Changes in drug use in European cities during early COVID-19 lockdowns - A snapshot from wastewater analysis date: 2021-03-26 journal: Environ Int DOI: 10.1016/j.envint.2021.106540 sha: c3d843a78f65b923bb6635b52cd4db55f11f5261 doc_id: 858450 cord_uid: h5ni6tcx The COVID-19 outbreak has forced countries to introduce severe restrictive measures to contain its spread. In particular, physical distancing and restriction of movement have important consequences on human behaviour and potentially also on illicit drug use and supply. These changes can be associated with additional risks for users due to reduced supplies, limited prevention and harm reduction activities. Furthermore, there have been limitations in the amount of data about drug use which can be collected due to restrictions. To goal of this study was to obtain information about potential changes in illicit drug use impacted by COVID-19 restrictions. Wastewater samples were collected in seven cities in the Netherlands, Belgium, Spain and Italy at the beginning of lockdowns (March-May 2020). Using previously established and validated methods, levels of amphetamine (AMP), methamphetamine (METH), MDMA, benzoylecgonine (BE, the main metabolite of cocaine) and 11-nor-9-carboxy-Δ(9)-tetrahydrocannabinol (THC-COOH, main metabolite of tetrahydrocannabinol (THC)) were measured and compared with findings from previous years. Important differences in levels of consumed drugs were observed across the considered countries. Whilst for some substances and locations, marked decreases in consumption could be observed (e.g., 50% decrease in MDMA levels compared to previous years). In some cases, similar or even higher levels compared to previous years could be found. Changes in weekly patterns were also observed, however these were not clearly defined for all locations and/or substances. Findings confirm that the current situation is highly heterogeneous and that it remains very difficult to explain and/or predict the effect that the present pandemic has on illicit drug use and availability. However, given the current difficulty in obtaining data due to restrictions, wastewater analysis can provide relevant information about the situation at the local level, which would be hard to obtain otherwise. The COVID-19 outbreak has forced countries to introduce severe restrictive measures to contain its spread. In particular, physical distancing and restriction of movement have important consequences on human behaviour and potentially also on illicit drug use and supply. These changes can be associated with additional risks for users due to reduced supplies, limited prevention and harm reduction activities. Furthermore, there have been limitations in the amount of data about drug use which can be collected due to restrictions. To goal of this study was to obtain information about potential changes in illicit drug use impacted by COVID-19 restrictions. Wastewater samples were collected in seven cities in the Netherlands, Belgium, Spain and Italy at the beginning of lockdowns (March-May 2020). Using previously established and validated methods, levels of amphetamine (AMP), methamphetamine (METH), MDMA, benzoylecgonine (BE, the main metabolite of cocaine) and 11-nor-9-carboxy-Δ⁹-tetrahydrocannabinol (THC-COOH, main metabolite of tetrahydrocannabinol (THC)) were measured and compared with findings from previous years. Important differences in levels of consumed drugs were observed across the considered countries. Whilst for some substances and locations, marked decreases in consumption could be observed (e.g., 50% decrease in MDMA levels compared to previous years). In some cases, similar or even higher levels compared to previous years could be found. Changes in weekly patterns were also observed, however these were not clearly defined for all locations and/or substances. Findings confirm that the current situation is highly heterogeneous and that it remains very difficult to explain and/or predict the effect that the present pandemic has on illicit drug use and availability. However, given the current difficulty in obtaining data due to restrictions, wastewater analysis can provide relevant information about the situation at the local level, which would be hard to obtain otherwise. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak forced countries to introduce various measures to constrain its spreading. Home confinement, physical distancing and closing of non-essential activities, have had profound effects on individuals and societies. These changes have also affected substance use and associated risks due to reduced supplies, limited prevention, treatment and harm reduction activities (EMCDDA 2020a; 2020b; Dietze and Peacock 2020) . These changes triggered the implementation of innovative methods to assess the impact lockdowns have on substance use and drug markets. In particular, mixed-method trend-spotter studies were introduced by the European Centre for Drugs and Drug Addiction (EMCDDA) to assess the situation (EMCDDA 2020b). These studies illustrate the complex and heterogeneous situation of drug use and markets in Europe during the early stages of the pandemic and highlighted the need for additional and localized surveys. In fact, due to differences in severity and timing of measures introduced in different countries, it remains difficult to explain and predict the effect that these will have on substance use and availability (Giommoni 2020; EMCDDA and Europol 2020) . While there appears to have been a decrease in stimulants consumption across Europe, in particular cocaine and MDMA, changes are not homogeneously distributed across the continent (EMCDDA 2020b). An increased use of amphetamine in Nordic cities has been reported, whereas changes in cannabis use were more variable (Finnish Institute for Health and Welfare 2020; EMCDDA 2020b). Specifically, a decrease or cessation was reported among occasional users in southern Europe, while the opposite was found among frequent users in northern Europe (EMCDDA 2020b). A complementary approach which could help gather additional information in these complex circumstances is wastewater-based epidemiology (WBE). While it cannot provide information about changes at the individual level (e.g., users switching to more harmful substances/administration routes), it can detect community-wide modifications in substance use, potentially symptomatic of changes in user behaviour or supply. Given the current circumstances, in which contacts with users are limited due to lockdown and social distancing measures (e.g., limited access to drug-related health services), WBE is a valuable source of complementary information. This has recently been shown also by Reinstadler et al. (2020) , who used WBE to assess changes in illicit drug, alcohol and pharmaceutical use in the city of Innsbruck (Austria) between March and April 2020. The authors showed that changes in mass loads of illicit drugs and alcohol could be observed when compared to data from previous years. The goal of this study was to apply WBE in order to evaluate spatial and temporal changes in illicit drug use in selected communities in the Netherlands, Belgium, Spain and Italy, during the early phase of the pandemic. Following standardized sampling protocols and validated analytical methods, levels of cocaine (COC) and its main metabolite benzoylecgonine (BE), amphetamine (AMP), 3,4-methylenedioxymethamphetamine (MDMA), methamphetamine (METH) and cannabis, through measurement of 11-nor-9-carboxy-Δ⁹-tetrahydrocannabinol (THC-COOH), were determined in wastewater samples collected from 7 locations between March and May 2020. Findings were then compared to results from previous monitoring campaigns, which have been carried out yearly since 2011. Influent wastewater samples were collected between March and May 2020 at the entrance of the wastewater treatment plants (WWTPs) of the 7 cities mentioned in Table 1 . All samples were 24-hour composite samples, collected following the protocols established in the yearly monitoring campaigns coordinated by the Sewage Analysis Core group Europe (SCORE) (González-Mariño et al. 2020; Castiglioni et al. 2013; SCORE 2020) . Historic data consists of results from yearly sampling campaigns, organized by SCORE between March and April since 2015 and published by the EMCDDA (EMCDDA, 2020c). Additional data such as population and wastewater flows were provided by the WWTPs personnel and were used to compute population normalized daily loads (expressed in milligrams of substance excreted per day per thousand inhabitants [mg/day.1000 inhabitants]). Analyses of COC, BE, AMP, MDMA, METH and THC-COOH were carried out using liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS). Analyses were performed in different laboratories and all methods used (van Nuijs et al. 2018; Bijlsma et al. 2014 ; Van Nuijs et al. 2009; Emke et al. 2018) , including those used for historic data (priori to 2020), have been fully validated and successfully passed yearly proficiency tests (van Nuijs et al. 2018) . Despite the uncertainties related to the measurement of THC-COOH, long-term monitoring within the same catchment provides useful insights in cannabis consumption (Burgard et al. 2019) . Hence, THC-COOH was determined in all locations, except in Milan, Brussels and Boom where no validated method was available for this metabolite. Historic data presented here can be retrieved from the EMCDDA (2020c) or can be found in the recent paper by González-Mariño et al. (2020) . Differences in population normalized loads of the measured illicit drugs were evaluated using the nonparametric Wilcoxon rank sum test, because in most cases data was not normally distributed. One-or two-sided tests were used to determine whether mass loads were higher, lower or equal compared to previous years, respectively. For comparison purposes, an overview of excreted mass loads of targeted drugs measured in the period from 2015 to 2020 is shown in Figure 1 . A thorough evaluation of trends in illicit drug use measured through WBE has been recently reported by (González-Mariño et al. 2020 ). It should be noted that for the cities of Eindhoven and Utrecht, not all data are reported as dumps of unused drugs or production waste were regularly detected. During such events, extremely high and erratic concentrations of AMP, MDMA and METH are detected, which cannot be explained by consumption only (Emke et al. 2018 ). In such cases, measured loads can thus not be used to evaluate trends. In the cities of Utrecht, Brussels and Boom, no decrease in AMP mass loads was observed in 2020 compared to 2019 (one-sided Wilcoxon rank sum test, p-value < 0.05). On the contrary, the increasing trend observed in the last three years appears to continue into 2020. This was not the case for Amsterdam, where a decrease of almost 50% in AMP was observed, comparable to mass loads measured back in 2017 (onesided Wilcoxon rank sum test, p-value > 0.05). In Castellon and Milan, AMP was either not detected or only sporadically, hence no comparison can be made. With respect to BE, loads measured in Utrecht and Brussels appear to still increase in 2020 despite the introduction of measures. However, statistical testing showed no significant difference between 2020 and 2019 for both cities (one-sided Wilcoxon rank sum test, p-value > 0.05). In Milan, however, a significant increase in BE loads was observed in 2020 compared to 2019 (one-sided Wilcoxon rank sum test, p-value = 0.001). For all other cities, an overall decrease in BE loads could be observed in 2020. In particular, Amsterdam (one-sided Wilcoxon rank sum test, p-value = 0.03) and Castellon (one-sided Wilcoxon rank sum test, p-value = 0.0003) showed a decrease in BE loads compared to previous years of more than 20% and 60%, respectively. In the case of MDMA, no significant change could be observed in Utrecht, Brussels, Boom and Castellon in 2020 compared to previous years (two-sided Wilcoxon rank sum test, p-value > 0.05). However, Amsterdam and Milan showed a significant decrease (one-sided Wilcoxon rank sum test, p-value < 0.05) of almost 50% on average. METH loads showed a decrease in 2020 compared to previous years, except for Brussels, where no significant change could be observed (two-sided Wilcoxon rank sum test, p-value > 0.05). In Amsterdam, measured loads varied greatly, however a significant decrease could be observed between 2020 and 2019 (one-sided Wilcoxon rank sum test, p-value=0.01). The other cities saw a decrease of 30% or more in METH loads in 2020 compared to 2019. Nevertheless, METH loads are in general low to very low in the selected cities, hence these changes should be interpreted with caution. THC-COOH was not monitored in all cities. In the case of Amsterdam and Castellon, a slight decrease in mass loads in 2020 compared to 2019 can be observed, however this was not statistically significant (one-sided Wilcoxon rank sum test, p-value > 0.05). Similarly, no significant difference was observed in Utrecht and Eindhoven in 2020 compared to 2019. A comparison between weekly patterns in drug mass loads in the period 2018-2020 was performed to determine if any change in consumption patterns could be observed. A selection of the most relevant patterns is shown in Figure 2 . Weekly patterns of all substances and cities are reported in the Supporting Information (Tables S1-S4). No obvious general trend was visible. For instance, in Amsterdam, the AMP weekly pattern was flatter (no weekend peak) compared to 2018 and 2019, while in Brussels and Boom no particular change could be observed. In Milan, AMP was detected only in 2019, hence no comparison could be made. Similarly, in Castellon, AMP was not detected between 2018-2020. For BE, the weekend peak in Amsterdam was less pronounced compared to previous years, while it remained the same in Utrecht, Brussels and Boom. In Castellon, a steep decrease in BE loads from Thursday to Wednesday can be observed, while in 2019, there was a clear peak on Sunday. However, also in 2018, the weekend BE peak was not clearly defined, in particular because of another important peak on Wednesday. In the case of MDMA, the weekend peak was less pronounced in Amsterdam compared to previous years. Little to no change in the weekend pattern of MDMA was observed in Utrecht, Brussels and Boom. In Milan, MDMA loads appear somewhat lower in 2020 compared to previous years. In particular, the weekend peak is almost a factor 10 lower compared to 2018 and 2019. In Castellon there is an important difference between 2020 and 2019, however when compared to 2018 this is less obvious, also because of the low detection frequency of MDMA. In fact, in samples from 2020, MDMA was not detected between Monday and Wednesday. In the case of METH, observed weekly patterns differ greatly between cities. For instance, in Milan, a peak was observed on Sundays, both in 2018 and 2019. Yet, in 2020 it appears to have decreased and shifted to Saturday. This also appears to be the case for Amsterdam. For THC-COOH, years differ greatly, and it is difficult to observe any overall change in 2020 compared to previous measurements. The obtained data clearly highlights the heterogenic effect that restrictions have caused on illicit drug use. More so, speculations about a generalized decrease in consumption does not seem to have materialized. While in Amsterdam and Castellon, consumption of stimulants, such as COC, AMP and MDMA, has seen a substantial decrease in 2020, findings are more nuanced for Milan and in particular Utrecht. Depending on the substance considered, constant or even higher drug loads were observed in 2020 compared to previous years. Some hypotheses can be formulated to explain these observations, yet they remain speculative due to the complexity and short period considered. The important decrease in tourism and/or closing of nightlife could explain observations made in Amsterdam and Milan, for instance. Yet, diminished availability or changes in consumption patterns (Finnish Institute for Health and Welfare 2020; EMCDDA 2020b) or differences in severity of measures could also have an effect. In this sense, it is interesting to notice that, according to the Government Response Stringency Index (GRSI) (University of Oxford 2020), Spain (GRSI 71.8 -85.2) had more stringent measures compared to the Netherlands (GRSI 62.0 -74.1) (Table 1) , the impact that these had on the situation of illicit drug use was not obvious. Both Amsterdam and Castellon saw a decrease in stimulant use, while Utrecht had a substantially different situation. Similarly, Belgium (GRSI 81.48) had comparable measures to Spain, yet the effect on illicit drug use appeared less pronounced, in particular for AMP and BE which remained stable or even increased in 2020 in Brussels. In the particular case of Milan, it should be noted that because of restrictions in place, sampling of wastewater took place only during the second half of May, which is later compared to the other cities considered here. Consequently, measures in place during this period were less stringent compared to the other cities considered here (i.e., GRSI of 64.9), which could explain why differences with previous years were less pronounced. These findings illustrate that it remains very difficult to attribute the observed changes to a single factor. With respect to synthesis of stimulants, results from Eindhoven, a renowned area for production, suggest that manufacturing of AMP, MDMA and METH in the area continued during the studied period. In fact, high loads of up to almost 350 mg/day 1000 inhabitants (or 1500 ng/L) of AMP were measured during weekdays. Similarly, loads of up to 92 mg/day 1000 inhabitants (or 340 ng/L) of METH were measured. Weekly patterns were also very heterogeneous. For AMP, MDMA and BE weekend peaks were less pronounced or even completely absent in 2020. In Milan, a factor 10 difference in the amount of MDMA consumed during the weekend peak could be observed between the considered years. Interestingly, in Castellon, BE mass loads showed a steep decrease during the sampled week, which corresponded to the introduction of a full lockdown halting all nonessential activities. Furthermore, in Utrecht, weekly patterns of MDMA and BE still showed a clear trend, whereas the weekend peak was less pronounced in Amsterdam and Eindhoven suggesting different dynamics in drug use. For cannabis (measured through THC-COOH), weekly patterns were very heterogeneous in all considered years, making comparisons particularly challenging. Both yearly and weekly trends observed in the considered cities support the heterogeneous situation hypothesized by Giommoni et al. and the EMCDDA (Giommoni 2020; EMCDDA 2020b) . This makes predictions about the effect that lockdown measures have on consumption very difficult, in particular due to national, regional and local characteristics, both in terms of drug use patterns as well as lockdown measures implemented and conformed by the public. Furthermore, given the continuously changing situation, with measures being introduced or relieved, findings highlight the need for and introduction of high resolution, localized and multi-indicator monitoring campaigns. The findings from this snapshot study show that the impact of the COVID-19 pandemic had heterogeneous effects on illicit drug use. In some of the studied locations, consumption remained stable or even increased compared to previous years, while clear reductions in drug uses were seen elsewhere. Patterns of drug use also seem to have changed slightly, but these are both substance-and location-dependent. Results obtained here further highlight the complexity of the situation, which might increase further due to cyclic introduction and relieving of measures, and the importance of continuous monitoring using all multiple indicators. Nevertheless, at the local level, WBE can help partly disentangle this complex situation, in particular given the limited information which can currently be obtained directly from users. This will be relevant also for substances not included in this study, such as opioids, antidepressants and alcohol, whose consumption might also change during the COVID-19 pandemic (EMCDDA 2020b, Reinstadler et al. 2020 ). Improvements in Analytical Methodology for the Determination of Frequently Consumed Illicit Drugs in Urban Wastewater Using Wastewater-based Analysis to Monitor the Effects of Legalized Retail Sales on Cannabis Consumption in Washington State, USA Evaluation of Uncertainties Associated with the Determination of Community Drug Use through the Measurement of Sewage Drug Biomarkers Illicit Drug Use and Harms in Australia in the Context of COVID-19 and Associated Restrictions: Anticipated Consequences and Initial Responses Impact of COVID-19 on Drug Services and Help-Seeking in Europe Impact of COVID-19 on Patterns of Drug Use and Drug-Related Harms in Europe Wastewater-Based Epidemiology and Drugs Topic Page Luxembourg: Office of the European Union Amphetamine Use Has Continued to Increase in the Helsinki Area during the Exceptional Circumstances -the Increase Also Visible in Traffic -Press Release -THL Spatio-Temporal Assessment of Illicit Drug Use at Large Scale: Evidence from 7 Years of International Wastewater Monitoring Multi-Year Inter-Laboratory Exercises for the Analysis of Illicit Drugs and Metabolites in Wastewater: Development of a Quality Control System Monitoring Drug Consumption in Innsbruck during Coronavirus Disease 2019 (COVID-19) Lockdown by Wastewater Analysis Coronavirus Government Response Tracker Analysis of Drugs of Abuse in Wastewater by Hydrophilic Interaction Liquid Chromatography-Tandem Mass Spectrometry closing of borders for all non-essential travel quarantine in areas of Lombardy and Veneto 9 March 2020: introduction of nation-wide measures *Government Response Stringency Index (GRSI) Changes in drug use in European cities during early COVID-19 lockdowns -A snapshot from wastewater analysis Istituto di Ricerche Farmacologiche Mario Negri IRCCS -Department of Environmental Health Sciences Writing -Original Draft, Writing -Review & Editing, Formal analysis, Visualization, Project administration Data Curation, Investigation, Writing -Original Draft Tim Boogaerts: Data Curation, Investigation, Writing -Original Draft, Writing -Review & Editing Sara Castiglioni: Conceptualization, Methodology, Investigation, Writing -Original Draft, Funding acquisition Marina Campos-Mañas: Data Curation, Investigation, Writing -Original Draft Alberto Celma: Data Curation, Investigation, Writing -Original Draft Writing -Review & Editing, Funding acquisition Pim de Voogt: Conceptualization, Methodology, Writing -Original Draft, Funding acquisition Félix Hernández: Conceptualization, Methodology, Writing -Original Draft Writing -Original Draft, Data Curation Noelia Salgueiro-González: Data Curation, Investigation, Writing -Original Draft Ruud Steenbeek: Data Curation, Investigation, Writing -Original Draft Writing -Original Draft, Writing -Review & Editing, Funding acquisition Ettore Zuccato: Conceptualization, Methodology, Writing -Original Draft, Funding acquisition Lubertus Bijlsma: Conceptualization, Methodology, Investigation, Writing -Original Draft, Writing -Review & Editing, Formal analysis, Visualization, Project administration Changes in drug use in European cities during early COVID-19 lockdowns -A snapshot from wastewater analysis Istituto di Ricerche Farmacologiche Mario Negri IRCCS -Department of Environmental Health Sciences Initiatives, EuSeME (project number 861602).