Tracing methamphetamine and amphetamine sources in wastewater and receiving waters via concentration and enantiomeric profiling

Science of the Total Environment 601–602 (2017) 159–166

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Tracing methamphetamine and amphetamine sources in wastewater and receiving waters via concentration and enantiomeric profiling XuZeqiong, DuPeng, LiKaiyang, GaoTingting, WangZhenglu, FuXiaofang, LiXiqing ⁎ Laboratory of Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, 100871 Beijing, PR China

a r t i c l e

i n f o

Article history: Received 1 March 2017 Received in revised form 4 May 2017 Accepted 4 May 2017 Available online xxxx Editor: Kevin V. Thomas Keywords: Amphetamine Methamphetamine Sources Concentration and enantiomeric profiling Wastewater and receiving waters

a b s t r a c t Wastewater analysis is a promising approach to monitor illicit drug abuse of a community. However, drug use estimation via wastewater analysis may be biased by sources other than abuse. This is especially true for methamphetamine and amphetamine as their presence in wastewater may come from many sources, such as direct disposal or excretion following administration of prescription drugs. Here we traced methamphetamine and amphetamine sources via concentration and enantiomeric profiling of the two compounds from black market to receiving waters. Methamphetamine in wastewater was found to predominantly arise from abuse, proving the feasibility of using wastewater analysis for estimating its consumption in China. Amphetamine abuse was previously considered negligible in East and Southeast Asia. However, we found that amphetamine was abused considerably (up to 90.7 mg/1000 inh/day) in a significant number (N20%) of major cities in China. Combined concentration and enantiomeric profiling also revealed direct disposal into receiving waters of methamphetamine manufactured by different processes. These findings have important implications for monitoring of and law enforcement against methamphetamine/amphetamine abuse and related crimes in China and abroad. © 2017 Published by Elsevier B.V. H I G H L I G H T S • • • •

Methamphetamine and amphetamine sources were traced via concentration and enantiomeric profiling. Methamphetamine in Chinese wastewater was found to predominantly arise from abuse. Amphetamine was abused considerably in a significant number of major cities in China. Combined concentration and enantiomeric profiling revealed direct methamphetamine disposal into receiving waters.

1. Introduction Amphetamine-type stimulants (ATS), including mainly amphetamine, methamphetamine, and ecstasy-group substances (e.g., 3,4methylenedioxymethamphetamine), are the second most widely used class of drugs worldwide (after cannabis) (UNODC, 2015). East and Southeast Asia, with about one third of the global population, has some of the most established ATS markets in the world (Global SMART Programme, 2011, 2013, 2015). Methamphetamine (METH) is the primary ATS and its use continues to increase across the region. For example, METH seizure in China has roughly quadrupled from 6.15 t in 2008 to 25.9 t in 2014, far exceeding seizures of other drugs (Office of China National Narcotic Control Commission, 2009, 2015). Some countries also seized significant amounts of ecstasy pills (e.g., over 4 million in Indonesia in 2012) (Global SMART Programme, ⁎ Corresponding author. E-mail address: [email protected] (X. Li).

http://dx.doi.org/10.1016/j.scitotenv.2017.05.045 0048-9697/© 2017 Published by Elsevier B.V.

2013). In contrast, no amphetamine (AMP)-related seizure, arrest, and manufacturing facility was recorded in the past 2–3 years (Global SMART Programme, 2011, 2013), indicating AMP abuse in the region was minor. Traditional drug monitoring methods, based largely on population surveys, are time consuming and may be inaccurate (Zuccato et al., 2008). In the past decade, a new approach has emerged that estimates drug consumption by measuring drug concentrations in wastewater and taking into account population serviced, stability of drug residues, excretion rates, and wastewater volumes (Zuccato et al., 2005). This approach, being more objective and much less time-consuming (Zuccato et al., 2008), has been applied in many countries (e.g., Castiglioni et al., 2015; Du et al., 2015; Kim et al., 2015; Lai et al., 2013; Metcalfe et al., 2010; Nuijs et al., 2009; Thomas et al., 2012). While this approach represents a significant improvement, it also has uncertainties and biases. For example, drug release into wastewater may arise from sources other than its abuse (e.g., direct disposal; metabolism of other drugs), which may lead to significant biases. This is especially true for AMP

160

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

and METH, as both drugs have legal medical uses and can be excreted following administration of a number of medicines (Cody, 2002). Most illicit drugs are chiral and exist in the form of two or more enantiomers (Kasprzyk-Hordern et al., 2010). Different manufacturing processes can yield illicit drugs of completely different enantiomeric compositions. For example, the Leuckart or reductive amination process that uses phenyl-2-propanone as a precursor yields racemic METH products, whereas the Emde or Nagai processes using ephedrine or pseudoephedrine as precursors yield solely S(+)-METH (Remberg and Stead, 1999). Furthermore, METH, AMP, and their precursor drugs are metabolized by the human body with characteristic enantioselectivity (Cody, 2002). Thus, comparing enantiomeric compositions has the potential to shed light on sources of chiral drugs. Chiral analysis linked the excessively high mass loads of MDMA in wastewater during a sampling campaign in Utrecht to direct MDMA disposal, demonstrating the potential of enantiomeric profiling for source tracing of illicit drugs (Emke et al., 2014). Using concentration and enantiomeric distribution of fluoxetine in raw wastewater and other information (e.g., prescription data), Petrie et al. (2016) demonstrated direct disposal of fluoxetine. In addition, the authors proposed a framework to differentiate consumed and nonconsumed loads in wastewater. However, the scheme could only apply to simple cases when the enantiomeric form of the drug in question is known, and there are no other sources of the parent drug and its metabolites. Such a framework does not apply to METH and AMP, as METH is produced in different enantiomeric forms by different routes. Furthermore, METH and AMP are also metabolites of many other pharmaceuticals (e.g., deprenyl, benzphetamine) (Cody, 2002), which complicates the source apportionment of METH and AMP in wastewater dramatically. In this work, enantiomeric profiling of METH and AMP was expanded to METH drugs seized from suppliers, the urines of abusers, wastewater, and rivers across China. In addition to enantiomeric profiling, AMP and METH concentrations, as well as concentration ratios between the two drugs were also compared. Factors that might affect the concentrations and the enantiomeric compositions of METH and AMP in wastewater and rivers were fully discussed. The combined profiling approach yielded unique insights and unequivocally revealed different METH and AMP sources in wastewater and receiving waters in China. To our knowledge, this is the first report of simultaneous concentration and enantiomeric profiling of two closely related drugs throughout its cycle from the supply market to receiving waters. 2. Materials and methods 2.1. Sample collection In total, 67 crystalline and 54 tablet samples were randomly picked for analysis from METH drugs seized in Heilongjiang, Beijing, Ningxia, Sichuan, Zhejiang, Hubei, Shandong, and Guangdong provinces. These provinces cover all the geographic regions of China. No sample of “Shenxian Shui”, a mixed liquid drug that also contains METH, was collected. According to the Bureau of Narcotics Control (personal communication), seizure of this drug was negligible compared to those of crystalline and tablet METH. Furthermore, METH concentrations in this drug were very low (0.04–1.3%) (Zhu et al., 2014). Urine samples of METH abusers were collected in Shandong (21 samples) and Guangdong (31 samples) provinces in the first half of 2015 with the assistance of local rehabilitation centers, in accordance with a protocol approved by the ethics committee of Peking University and with the informed consent of the addicts (The ethics approval number is IRB0000105216029). Time proportional composite influent wastewater samples were collected for two days (i.e., 2 samples) using autosamplers or manually from 19 STPs at 14 major cities across China in the summer of 2014 and 2015 (Fig. S1). Time proportional composite wastewater effluents (10 samples) were also collected from 5 STPs in Beijing, Guangzhou, and Shenzhen. Twelve and eight grab samples were collected in the

summer of 2015 along Liangshui (LSR) and Shenzhen (SZR) rivers that flow through Beijing and Shenzhen, respectively (Fig. S2). The Liangshui River receives effluent from STPs BJ-2 and BJ-3, whereas the Shenzhen River receives effluent from STP SZ-1. Details of sample collections are available in Supplementary Content (Table S1, S2). 2.2. Sample preparation and analysis Seized METH samples were dissolved in methanol (MeOH) to roughly 1 mg mL−1, filtered using 0.22 μm centrifuge filters, and diluted 5000 and 1000 times for LC-MS/MS analysis of METH and AMP, respectively. Urine samples were first diluted by MeOH by a factor of 6, vortexed for 20s, and centrifuged for 1 min at 13000g. Aliquots of supernatants were spiked with deuterated internal standards and then further diluted 2 to 40 times by MeOH. Wastewater and river waters samples were pretreated using solid phase extraction. An Oasis HLB cartridge (60 mg, 3 mL, Waters, UK) was conditioned in sequence with 2 mL MeOH and 2 mL deionized water at pH = 7.5 (adjusted using ammonium hydroxide). Wastewater (50 mL) or river water (200 mL), filtered using a 0.45 mm glass fiber membrane and spiked with internal standards, was loaded to the cartridge at a flow rate of 1–2 mL min−1. The cartridge was then rinsed using 5% MeOH solution, dried under vacuum, and eluted using 4 mL MeOH. The eluate was evaporated to dryness, redissolved in 200 μL MeOH, and further cleaned using a centrifugal filter. Pretreated samples were analyzed using a UFLCXR-LC system (Shimadzu, Japan) coupled with a Chirobiotic V2 column (250 mm × 2.1 mm, 5 μm) (Sigma-Aldrich, UK) at an injection volume of 5 μL. Enantiomer separation was undertaken at 20 °C and a flow rate of 0.25 mL min−1, under isocratic conditions with a mobile phase composed of MeOH, glacial acetic acid, and ammonium hydroxide (100:0.1:0.025, v/v). Baseline separation of both METH and AMP enantiomers was achieved (Fig. S3). Concentrations of individual enantiomers were determined using an API 4000 triple quadrupole mass spectrometer (AB SCIEX, USA) equipped with an electrospray interface operating in positive ionization mode (Table S3). Enantiomeric fractions (EFs) were derived by dividing the S(+)-enantiomer concentrations by the summed concentrations of the two enantiomers. EFs greater and lower than 0.5 indicate enrichment of S(+)- and R(−)-enantiomers, respectively. Method quantification limits (MQLs) of METH and AMP were 2 ng L−1 for wastewater and 0.5 ng L−1 for river water, respectively. Recoveries and matrix effects of the enantiomers were N 92% and 84%, respectively. Details of sample analysis and method validation are provided in Supplementary content (Tables S3–5). 2.3. Load and consumption estimation The total daily mass loads of AMP and METH at a specific STP were estimated using the following equation:

Load





mg ¼ 1000 inh
d

Conc:

ng

 influent flow

L Population served 1000

  L d



  mg ng 10 1

6

where conc. represents summed concentrations of the two enantiomers. The contribution of AMP abuse to the total load of AMP was derived by subtracting the contribution of METH metabolism: LoadAMP abuse ¼ LoadAMP total −LoadMETH  0:07 where 0.07 represents the upper bound of the concentration ratio of AMP and METH in wastewater following METH abuse (details provided in below sections). Consumption of AMP and METH was back-calculated from loads of abuse by multiplying correction factors (2.77 and 4.4,

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

161

respectively) that account for excretion following its abuse (Gracia-Lor et al., 2016). 3. Results 3.1. Enantiomeric fractions (EFs) of METH and AMP contents in seized METH drugs S(+)-METH was the only enantiomer detected (EF = 1) in all but two samples of seized crystalline METH (Table S6). In the two samples in which R(−)-METH was detected, EF values were 0.54 and 0.97, respectively. In 46 out of 54 tablet METH samples, S(+)-METH was also the only enantiomer detected (Table S7). Considerable fractions of R(−)-METH (EF b 0.8) were detected in only four tablet samples. These results indicate that in the Chinese market, S(+)-METH is the predominant enantiomer. AMP/METH ratio was between 0.01 and 0.02 in 2 crystalline samples and between 0.001 and 0.01 in 7 crystalline and 3 tablet samples. In the other seized samples, AMP was below the detection limit. These results indicate that AMP impurity in seized METH drugs was negligible. In addition, in the 12 samples with AMP/ METH ratios N 0.001, only S(+)-AMP was detectable. 3.2. Concentrations and EFs of METH and AMP in urine samples METH and AMP concentrations ranged from 24.1 to 191,520 and from 2.1 to 24,888 ng mL−1, respectively (Table S8). The average ratios of AMP and METH concentrations were 0.142 (± 0.073) and 0.123 (±0.108) for the samples collected in Shandong and Guangdong provinces, respectively (overall average ratio = 0.135 ± 0.100). METH and AMP in urines collected in this work were predominantly the S(+)-enantiomer (Table S8). R(−)-METH was only detected in 5 samples, with EFs ranging from 0.92 to 0.99. The predominance of the S(+)-enantiomer in urine samples is consistent with the observation that S(+)-enantiomer was the dominant form in seized METH. R(−)-AMP was only detected in 2 samples, in which R(−)-METH was also detected. This is understandable, as during metabolism of METH into AMP, there is no chiral inversion taking place, i.e., S(+)-METH is exclusively metabolized into S(+)-AMP, whereas R(−)-METH is exclusively metabolized into R(−)-AMP. (Cody, 2002). In the two samples with both AMP and METH detection, EFs of AMP were slightly higher than those of METH (0.98 vs. 0.93 and 0.94 vs. 0.92, respectively). Higher EFs of AMP in these samples can be explained by the fact that in the human body more S(+)-METH is metabolized (into S(+)-AMP) than R(−)METH (into R(−)-AMP) (Cody, 2002), which leads to further enrichment of the S(+)-enantiomer of AMP relative to METH in urines.

Fig. 1. Concentrations and enantiomeric fractions of METH and AMP in influents at STPs in major cities of China: Haerbin (HRB), Dalian (DL), Beijing (BJ), Taiyuan (TY), Qingdao (QD), Suzhou (SU), Hangzhou (HZ), Wuhan (WH), Yinchuan (YC), Chengdu (CD), Kunming (KM), Guangzhou (GZ), Shenzhen (SZ), and Nanning (NN).

In contrast, the enantiomeric compositions of AMP in wastewater influents were quite different. Although the quantification frequency of R(−)-AMP (6 out of 14 cities and 8 out of 19 STPs) (Fig. 1, Table S9) was only slightly higher than that of R(−)-METH, AMP EFs were considerably lower than 1.0 at the majority of the 8 plants where R(−)-AMP was detected. Except at CD-1 (EF = 0.89), EFs were below 0.75 at all the other seven STPs. At BJ-3, GZ-1, and SZ-1, AMP EFs were even b0.65. These results indicate that in wastewater influents at about 40% of STPs in major Chinese cities, significant fractions of AMP (up to over 35%) existed in the form of R(−)-enantiomer. This observation has important implications, which will be discussed in detail in the Discussion section. 3.4. Concentrations and EFs of METH and AMP in effluent wastewater

3.3. Concentrations and EFs of METH and AMP in influent wastewater Concentrations of METH in wastewater influent ranged from 81.4 (± 53.4) ng L− 1 at WH-1 to 1153.3 (± 85.8) ng L− 1 at CD-1 (Fig. 1, Table S9). High METH concentrations were also observed at HRB-1 (713.0 ± 47.8 ng L−1), QD-1 (978.0 ± 53.7 ng L−1), and QD-2 (811.8 ± 130.4 ng L−1). METH concentrations at the overwhelming majority of the STPs were between 100 and 700 ng L− 1. AMP concentrations ranged from 4.6 ± 3.0 ng L−1at WH-1 to 132.6 (±26.9) ng L−1 at QD2. The ratios of AMP to METH concentrations varied significantly, ranging from 0.033 (±0.023) at NN-1 to 0.33 (±0.068) at GZ-1. METH present in the wastewater was predominantly the S(+)-enantiomer. The R(−)-enantiomer was quantifiable only in 5 out of 14 cities and at 6 out of 19 STPs (Fig. 1, Table S9). At the 6 STPs where R(−)METH was quantifiable, EF values were N 0.96 at four STPs (WH-1, KM1, KM-2, SZ-1) and N0.93 at one STP (CD-1). The only STP where EF was below 0.9 was GZ-1 (EF = 0.89 ± 0.07). That METH existed predominantly in the form of S(+)-enantiomer in wastewater influents is consistent with the observation that METH in seized drugs and urines of abusers was almost exclusively the S(+)-enantiomer.

Concentrations and EFs of METH and AMP were determined in effluents at five STPs (BJ-1,2,3, GZ-1, and SZ-1). METH concentrations ranged from 20.7 (±1.1) ng L−1 at BJ-2 to 153.2 (±27.6) ng L−1 at SZ-1 (Table S9). Apparent removal of METH ranged from 61.2 (±8.1) % at GZ-1 to 92.8 (±2.1) % at BJ-2, consistent with removal rates reported in previous studies in China (Du et al., 2015; Jing et al., 2014). R(−)-METH was below method quantification limit (MQL) in effluents of BJ-1, BJ2, as well as in influents. R(−)-METH was quantifiable in effluents at BJ-3, GZ-1 SZ-1, due to presence of R(−)-METH in influents at the three plants. EF values of METH in effluents either did not change or slightly decreased relative to values in influent. These results indicating that fractions of R(−)-METH increased following wastewater treatment, consistent with the finding by Kasprzyk-Hordern and Baker (2012). AMP in effluents was below method detection limit (MDL) at the majority of STPs examined, consistent with the finding that AMP removal was near complete at most STPs in China (Du et al., 2015; Jing et al., 2014). The only exception was effluent of BJ-3 on June 10, 2015, in which AMP was detected at a concentration of 4.9 ng L−1 and with an EF value of 0.51.

162

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

3.5. Concentrations and EFs of METH and AMP in rivers In the Liangshui River that flows through Beijing, METH, in the sole form of S(+)-enantiomer, decreased progressively from upstream (15.8 ng L−1 at LSR1) to middle stream (3.1 ng L−1 at LSR6), whereas neither enantiomer of AMP was detected upstream of LSR8 (Fig. 2, Table S10). From LSR8 downward, METH and AMP concentrations increased, to 37.6 and 7.6 ng L−1, respectively. More importantly, at these sampling points, both R(−)-METH and R(−)-AMP were quantifiable, with EFs ranging from 0.82 to 0.98 and from 0.41 to 0.66, respectively. In Shenzhen River, both enantiomers of METH and AMP were observed at all sampling points (Fig. 2, Table S10). METH and AMP concentrations ranged from 61.6 to 292.1 ng L−1 and 1.0 to 10.9 ng L−1, respectively, much higher than the concentration levels in the Liangshui River. Both METH and AMP concentrations showed a sudden increase from SZR3 to SZR4. This coincided with a sudden increase in EFs for both drugs from SZR3 to SZR4. The three upstream points (SZR1,2,3) were located in a tributary of the Shenzhen River. At these points, AMP was nearly racemic (0.51 b EFs b 0.58), where METH was racemic or slightly enriched with R(−)-enantiomer. At sampling points in the main river, both METH and AMP were enriched with the S(+)-enantiomers, with EFs ranging from 0.86 to 0.99 and from 0.67 to 0.91, respectively. 4. Discussion 4.1. Sources of METH in wastewater Although METH itself has no medical use in China, METH may be released into wastewater as metabolite of other medicines. Drugs that are

metabolized into METH include deprenyl, benzphetamine, dimethylamphetamine, famprofazone, fencamine, and furfenorex (Cody, 2002). However, prescription of these drugs except deprenyl is prohibited in China. Deprenyl, also known as selegiline, is used mainly to treat Parkinson's disease (Cody, 2002). Contributions of deprenyl metabolism to METH loads in Chinese wastewater are thought to be negligible at most (CMEIN, 2014; Khan et al., 2014). Lack of R(−)METH in wastewater at the majority of STPs examined in this study confirms, unambiguously, that METH from deprenyl metabolism was negligible at these STPs, as deprenyl is exclusively metabolized into R(−)METH (Cody, 2002) which should be detected if deprenyl use was noticeable. Since significant regional variation of deprenyl use in the same country is unlikely, deprenyl contributions in wastewater at STPs with R(−)-METH detection are also negligible. In fact, according to the medical statistical yearbook, there was no record of deprenyl production in China in 2014 (CMEIN, 2014), whereas production of deprenyl HCl in 2013 was 0.2 mg/1000 inh/d (CMEIN, 2013). Using the prescription information of 2013 and taking excretion rate (Khan and Nicell, 2012) into account, deprenyl contribution to METH loads in wastewater would be 0.03 mg/1000 inh/d at most. Thus, it can be concluded that METH detected in Chinese wastewater is exclusively of illegal origin. Small fractions of R(−)-METH in wastewater at some STPs (e.g., GZ-1) is consistent with the observation that a small fraction of seized METH drugs on the Chinese market is near racemic. METH of illegal origin enters wastewater via excretion following METH consumption. It may also enter wastewater via direct disposal by drug manufacturers/dealers who may dump the drug whenever they feel threatened (e.g., by visits of strangers). Direct dumping can potentially lead to significant bias on back-calculation of

Fig. 2. Concentrations and enantiomeric fractions of METH and AMP in Liangshui (a) and Shenzhen (b) rivers. The two rivers flow through Beijing and Shenzhen, respectively. AMP EFs were not plotted for LS1 to LS8 is because AMP at LS1 to LS8 was not detected.

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

methamphetamine consumption based on wastewater analysis. However, dumping would significantly reduce the concentration ratio between AMP and METH in wastewater, as dumped METH does not go through human metabolism. Thus, comparing AMP/METH ratios in wastewater to those in urines could provide insights into the occurrence and magnitude of METH dumping. In this work, among the 36 influent samples collected, AMP was detected in 15 samples at 8 STPs with considerable fractions of the R(−)enantiomer (EFs b 0.9) (Fig. 1, Table S9), indicating AMP input other than METH metabolism (which only yields S(+)-AMP) at these STPs. For the other 21 samples, both METH and AMP are predominantly (at KM-2) or exclusively (at other 10 STPs) in the form of S(+)-enantiomers. Since S(+)-AMP content in seized METH drugs is negligible, S(+)-AMP in these samples was almost exclusively from metabolism of S(+)-METH. Among these samples, AMP/METH ratios fell within (in 18 samples) or slightly beyond (in 2 samples, with ratios of 0.049 and 0.076) a narrow range of 0.055 to 0.070 (Fig. 3, Table S9). There was only one sample (NN-1 on July 18, 2015) in which AMP/METH ratio (0.017) was significantly below this range. AMP/METH in wastewater may be affected by the stability of the two substances in sewers and during sample collection. METH was found to be quite stable in wastewater, whereas reports on AMP stability were variable in the literature (e.g., Senta et al., 2014; Thai et al., 2014; van Nuijs et al., 2012). In this study, experiments were performed to examine the stability of METH and AMP using fresh wastewater collected from the Xiaojiahe plant in Beijing and composite wastewater freshly collected from eight other STPs in Beijing. For each wastewater, AMP and METH enantiomers were added at concentrations around 600 ng L−1 to two sets of samples. One set of the samples were kept for 6 h at room temperature to examine potential degradation of the two substances in sewer. The other set of the samples were kept for 24 h at 4 °C to examine stability during sampling. After 6 h at room temperature, the concentrations of AMP and METH enantiomers were 99.5– 101.6% of its original concentrations (Table S11), indicating no in-sewer degradation of the two compounds. After 24 h at 4 °C, concentrations of and AMP and METH ranged from 92.4 to 94.0% and from 96.2 to 99.8%, respectively, of its original concentrations. Thus, our results indicate that degradation of AMP and METH during sampling would not cause the ratio of the two compounds in wastewater to differ significantly from the average ratio in urines (i.e., the ratio of METH and AMP excretion rates).

163

Since AMP/METH ratios at the STPs without R(−)-AMP detection were all much lower than the average ratio in urine samples (0.135 ± 0.100), one would conclude, at first glance, that direct dumping occurred in the catchment of these plants. However, it is worth noting that METH and AMP concentrations in urine samples varied by over three orders of magnitude, presumably because some of the urine samples (with low concentrations) were collected at much later post-dose time (The time points when the urine samples of this study were collected following administration were unknown). The AMP/METH ratios in urine samples increased significantly with decreasing METH concentrations (corresponding to later post-dose time), from 0.067 (±0.042) to 0.208 (± 0.118) for samples with METH concentrations over 100,000 ng L−1 (n = 7) and b10,000 ng mL−1 (n = 14), respectively (Table S8). Hence, the average AMP/METH ratio in urines may significantly overestimate the excretion ratio of AMP/METH, as the majority of METH and AMP excretion occurred within the first 12 h after administration and excretion at later post-dose times should figure much less in the estimation of the excretion ratio. Valentine et al. (1995) also reported increasing AMP/METH ratio with post-dose time, from about 0.025 at 2 h to around 0.15 at 12 h post-dose. The AMP/METH excretion ratio was found to be 0.1 following oral administration of 30 mg S(+)-METH. In addition, the authors reported a saturable process of METH metabolism into AMP, i.e., the excretion ratio would be lower than 0.1 if METH was administrated at greater doses. Given that METH abusers in China consume METH with daily doses up to 0.4–0.7 g (Wang et al., 2013), excretion ratio between AMP and METH is expected to be much b 0.1 in China. Hence, low AMP/ METH ratios (b 0.07) observed in this study at STPs would not necessarily indicate significant dumping in the catchment of these plants. Furthermore, METH concentrations on two consecutive sampling days differed by a factor of 3.0 and 2.5 at SU-1 and KM-1, respectively (Table S9). If the much higher concentrations on one sampling day were caused by dumping, the AMP/METH ratios would be expected to vary significantly between the two sampling days. However, AMP/ METH ratios almost did not change at the two plants (0.076 vs. 0.068 at KM-1 and 0.058 vs. 0.055 at SU-1). Hence, AMP/METH ratios between 0.055 and 0.070 indicate that dumping either did not occur or was not captured by sampling. The observation that the overwhelming majority of AMP/METH ratios fell within this narrow range indicates that likelihood of dumping was low in most Chinese cities and therefore back-calculation using METH concentrations in wastewater would not significantly overestimate METH consumption. In addition, this range represents the excretion ratio between METH and AMP following METH consumption and can be used to determine if significant dumping occurred at a particular plant. For example, AMP/METH ratio of 0.017 at NN-1 on July 18 of 2015 indicates high likelihood of dumping, as confirmed by the observation that AMP concentration at the plant increased slightly from 5.1 ng L− 1 on July 17 to 6.0 ng L− 1 on July 18, whereas METH concentrations increased dramatically from 104.8 to 359.2 ng L−1 (Table S9). In this case, METH consumption can be estimated based on AMP concentrations instead, by taking into account the above range of AMP/METH ratios in absence of dumping.

4.2. Sources of AMP in wastewater

Fig. 3. Enantiomeric fractions of AMP vs. concentration ratios of AMP to METH in influent wastewater. Minor fractions of R(−)-AMP detected at KM-1, 2 with AMP/METH ratios b 0.08 were likely from metabolism of R(−)-METH observed at the two plants (Table S9). The vertical line at 0.07 represents the upper bound of AMP/METH in the wastewater in the absence of AMP consumption (i.e., AMP in wastewater was solely from METH metabolism).

R(−)-AMP was detected with significant fractions (EFs b 0.9) in 15 samples at 8 STPs. Plotting AMP EFs against AMP/METH concentration ratios revealed an interesting pattern: samples detected with significant fraction of R(−)-AMP had AMP/METH ratios no b 0.081, much higher than the AMP/METH ratios observed in most samples without significant R(−)-AMP detection (Fig. 3, Table S9). As stated earlier, lack of R(−)-AMP means that AMP in influent wastewater is predominantly from METH metabolism. Hence, Fig. 3 clearly demonstrates that other than METH metabolism, there are additional inputs of AMP in the wastewater at the 8 STPs. In fact, since METH in seized products, urines

164

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

and wastewater is almost exclusively the S(+)-enantiomer (which is metabolized into S(+)-AMP), significant fractions of R(−)-AMP (EFs b 0.9) must not come from METH metabolism. As demonstrated above, the likelihood that direct dumping significantly changes measured METH concentrations in wastewater is very low (i.e., measured METH was solely from METH consumption). In addition, the range of AMP/METH ratio following METH metabolism was known (0.055–0.07). Thus, AMP contribution from METH metabolism can be deducted from total AMP occurrence in wastewater using METH concentrations and the ratio of AMP/METH excretion rates following METH consumption. To derive the lower bound of AMP contributions from other sources, the upper bound (0.07) of AMP/METH excretion ratio following METH consumption was used. AMP contribution from other sources calculated this way accounted for 13.1% (BJ-2, August 25, 2015) to 81.7% (GZ-1, July 7, 2015) of total AMP in wastewater (Fig. 4, Table S12). The average contributions at BJ-1, BJ-3, QD-2, HZ2, and GZ-1 were N50%, indicating that at some STPs, AMP of other sources was the major contributor to its presence in wastewater. AMP in wastewater has more potential sources, including medical use of itself, metabolism of other legal drugs, metabolism of METH, AMP impurity in METH products, and its abuse. Like METH, AMP is not used for clinical purposes in China. Metabolism of METH in China yields predominantly S(+)-AMP, thus is ruled out as the source for significant fractions of R(−)-AMP. As demonstrated earlier, AMP impurity in METH drugs seized across China is exclusively the S(+)-enantiomer (which arises from reduction of norephedrine impurity present in ephedrine that is used as a precursor to produce S(+)-METH). As such, AMP impurity in METH products is also precluded as the source of R(−)-AMP in wastewater.

Drugs that have legal usage in China that can be metabolized in AMP include deprenyl and prenylamine. Prenylamine is a drug used to treat angina pectoris. Like deprenyl, this drug can yield both R(−)- and S(+)-AMP upon metabolism (Cody, 2002). However, the absence of R(−)-AMP in wastewater at the majority of STPs indicates that the contribution of the two drugs to AMP in wastewater is negligible. Indeed, there was no record of prenylamine production in the medical statistical yearbook of 2013 and 2014 (CMEIN, 2013, 2014), which strongly suggests that prenylamine use in China was negligible. Based on the production data of deprenyl HCl in 2013 (0.2 mg/1000 inh/d) (CMEIN, 2013) and the excretion rate (Khan and Nicell, 2012), an average AMP load of 0.01 mg/1000 inh/d in wastewater was expected. This contribution from deprenyl was negligible compared to AMP loads observed in wastewater. Hence, AMP abuse (likely in form of racemate) is the only possible source that can account for the significant fractions of R(−)AMP in wastewater. To evaluate the magnitude of AMP abuse, AMP loads and consumption were back-calculated based on AMP concentrations (contribution from METH metabolism subtracted), excretion rate, wastewater flow, and the populations that the 8 STPs served. Potential AMP degradation in the sewer and during sampling was not taken into account, leading to underestimation of AMP consumption. However, even the underestimated AMP loads at the STPs, ranging from 0.9 (BJ-2, June 10, 2015) to 39.7 (QD-2, August 24, 2015) mg per 1000 inhabitants per day (mg/1000 inh/d) (Fig. 4, Table S12), were significant. Loads at QD-2 (34.1 ± 8.0 mg/1000 inh/d) and GZ-1 (32.7 ± 3.1 mg/1000 inh/d) were higher than those observed in many countries in Europe and Canada (e.g., Metcalfe et al., 2010; Thomas et al., 2012). In addition, estimated AMP consumption was N 5% of estimated METH consumption at five

Fig. 4. Estimated AMP loads (unit: mg/1000 inh/d) in wastewater arising from abuse (a), contribution of AMP loads by abuse to total loads (b), and comparison of AMP and METH consumption (c). The two bars for each STP represents two sampling days. Only at the 8 STPs in the figure were there considerable contributions of AMP consumption to its occurrence in wastewater.

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

STPs (BJ-1, BJ-3, HZ-2, QD-2, GZ-1). At GZ-1 in particular, AMP consumption (90.7 ± 8.6 mg/1000 inh/d, translating to about 75 kg per year by the population it serves) was even N 15% of METH consumption (Fig. 4, Table S12). These results indicate that AMP abuse is significant within these cities and cannot be overlooked. The finding that there is considerable AMP consumption in a significant number of Chinese cities is intriguing, as reports of recent years indicate that AMP abuse was negligible or did not exist at all in China and in East and Southeast Asia at large (Office of China National Narcotic Control Commission, 2016; Global SMART Programme, 2011, 2013). According to the Bureau of Narcotics Control of China, no AMP manufacturing facility was identified in China in the past few years, whereas 486 METH manufacturing facilities were dismantled in 2015 alone (Office of China National Narcotic Control Commission, 2016). Significant AMP consumption and lack of AMP manufacturing facilities strongly suggest that there is a possibility of AMP trafficking to China and perhaps to other countries in East and Southeast Asia as well. Future research is needed to identify the origin(s) and route(s) of AMP trafficking to the region and to provide direction for international drug control efforts. 4.3. Sources of METH and AMP in receiving waters METH, in the sole form of S(+)-enantiomer, at LSR1 to LSR7 of the Liangshui River, may come from effluent discharge at an STP upstream of the river (Fig. 2, Fig. S2, Table S10). According to the Beijing Urban Drainage Monitoring Center, a small portion (about 5%) of wastewater in the area may be discharged without treatment, which also contributes to METH occurrence at those points. For LSR8 and downward, both S(+)- and R(−)-enantiomers of METH and AMP were observed. Observation of R(−)-enantiomers at these points is consistent with presence of R(−)-METH and R(−)-AMP in the effluent of BJ-3, which is located just upstream of LSR8. According to a local monitoring station, the flow of the river at a cross section just upstream of LSR8 was 1.96 m3 s−1 (daily average) on the sampling day. This means that the effluent of BJ-3 (650,000 m3 per day, translating to a daily average of 7.5 m3 s−1) was diluted by a mere factor of 1.26. Thus occurrence of METH and AMP at these points could be accounted for by effluent discharge, as concentrations of the two drugs in the effluent of BJ-3 and river were roughly at the same level. In contrast, METH and AMP concentrations and enantiomeric compositions in Shenzhen River differed dramatically from those in the effluents of STPs in Shenzhen. METH concentrations were quite high (N60 ng L− 1) at the three upstream sampling sites in the tributary (SZR1-SZR3) (Fig. 2, Table S10). In addition, METH and AMP at these points were nearly racemic or slightly enriched with R(−)-enantiomer. Given that METH in wastewater in China was predominantly the S(+)enantiomer, METH in the tributary could not be from wastewater. Rather, METH here was most likely from direct disposal by manufacturing using Leuckart or reductive amination process which yields racemic products. The Leuckart or reductive amination process can generate racemic AMP impurity (Kunalan et al., 2009), which explains the occurrence of near racemic AMP of low concentrations at these points. It can also explain the occurrence of low concentrations of R(−)-enantiomers of METH and AMP in the main stream. In the main stream (SZR4-SZR8), METH concentrations were close to or even significantly higher than the highest effluent concentrations observed in Shenzhen (Fig. 2, Table S10). There is another STP of Shenzhen at the far upstream of the river (Fig. S2) and possibly also STPs influencing the river on the Hong Kong side, it cannot be precluded that METH in the main stream came from effluents of the STPs. On the other hand, both the Hong Kong side of the river and the Shenzhen side at the far upstream are suburban areas. Previous studies indicate that in Chinese cities much less METH was consumed in suburban areas than in urban centers (Du et al., 2015; Jing et al., 2014). METH concentrations in influents at the STPs and in untreated wastewater along the river were

165

expected to be lower than in the influent of the SZ-1, as the later plant treats wastewater from the central urban area of Shenzhen where night clubs and other entertainment venues concentrate. Thus, this wastewater, even discharged without treatment, would only account for a small fraction of the high AMP and METH concentrations observed in the main stream. Without direct METH dumping, influent METH concentrations at a STP rarely varied by a factor 2 (Table S9). According to a local monitoring authority, the annual average flow of the river is about 17 m3 s−1. Since July is the rainy season in Shenzhen, flow during sampling is expected to be higher than the annual average, which means effluent from STP SZ-1 (about 290,000 m3 per day, 3.4 m3 s−1) would be diluted by a factor of at least 5. Thus, with direct dumping into influent SZ-1, contribution of effluent ofSZ-1 to METH occurrence in the main stream would be minor. In fact, with this dilution factor, METH concentrations in the river would be less than half of the highest concentration observed, even if influent of SZ-1 (with concentrations of July 4 and 6, 2015) was directly discharged (without treatment). Contribution of effluent to AMP occurrence would be even less as AMP in effluent was below detection limit, whereas AMP concentrations in the main stream were over 5 ng L−1 at some sampling points. Hence, it is unlikely that the majority of METH and AMP in the main stream came from human excretion following METH abuse. Rather, it is more likely that clandestine manufacturing was the major contributor. Clandestine manufacturing may lead to high river concentrations by direct disposal into the Shenzhen River, or via discharge into the influent of SZ-1, which lead to exceedingly high influent and effluent concentrations. Since METH exists mainly in the form of S(+)-enantiomer, it can be concluded that the ephedrine process was adopted in the manufacturing (International Narcotics Control Board, 2014). In this process, S(+)AMP was generated as an impurity (Lee et al., 2006), which is consistent with the presence of the enantiomer at these points. 5. Conclusions Concentration and enantiomeric profiling of METH and AMP from market to receiving waters revealed: 1) METH in wastewater was predominantly from abuse and wastewater analysis can be used to estimate its consumption in China; 2) there was considerable AMP abuse (up to 90.7 mg/1000 inh/d) in a significant number of major cities in China, indicating a likelihood of AMP trafficking into the country; 3) there may be at least two clandestine manufacturing facilities using different processes that released METH and AMP into the Shenzhen River. These results are valuable for monitoring and controlling the abuse of the two drugs in China. We also demonstrated that combined concentration and enantiomeric profiling provided unique evidence for source identification and quantification that were otherwise impossible if concentrations or enantiomeric compositions were profiled separately. For example, without enantiomeric composition data, abuse of AMP could not be concluded, whereas without concentration data, the magnitude of AMP abuse could not be estimated. There are several pairs of chiral illicit drugs and precursors that are closely related and have multiple sources (e.g., ephedrine/ methcathinone). The combined profiling approach demonstrated here may also be applicable to trace the sources of these drugs. Acknowledgements This work was funded by the National Natural Science Foundation of China (Grant No. 41371442 and 41401566). The authors wish to thank all the personnel involved in sample collection. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2017.05.045.

166

Z. Xu et al. / Science of the Total Environment 601–602 (2017) 159–166

References Castiglioni, S., Borsotti, A., Senta, I., Zuccato, E., 2015. Wastewater analysis to monitor spatial and temporal patterns of use of two synthetic recreational drugs, ketamine and Mephedrone, in Italy. Environ. Sci. Technol. 49 (9):5563–5570. http://dx.doi.org/10. 1021/es5060429. China Medicine Economic Information Net (CMEIN), 2013. Chinese Medical Statistical Yearbook. China Medicine Economic Information Net (CMEIN), Beijing, China. China Medicine Economic Information Net (CMEIN), 2014. Chinese Medical Statistical Yearbook. China Medicine Economic Information Net (CMEIN), Beijing, China. Cody, J.T., 2002. Precursor medications as a source of methamphetamine and/or amphetamine positive drug testing results. J. Occup. Environ. Med. 44 (5):435–450. http://dx. doi.org/10.1097/00043764-200205000-00012. Du, P., Li, K., Li, J., Xu, Z., Fu, X., Yang, J., Zhang, H., Li, X., 2015. Methamphetamine and ketamine use in major Chinese cities, a nationwide reconnaissance through sewagebased epidemiology. Water Res. 84:76–84. http://dx.doi.org/10.1016/j.watres.2015. 07.025. Emke, E., Evans, S., Kasprzyk-Hordern, B., de Voogt, P., 2014. Enantiomer profiling of high loads of amphetamine and MDMA in communal sewage: a Dutch perspective. Sci. Total Environ. 487:666–672. http://dx.doi.org/10.1016/j.scitotenv.2013.11.043. Global SMART Programme, 2011. Patterns and Trends of Amphetamine-type Stimulants and other drugs: Asia and the Pacific 2011. United Nations Office on Drugs and Crime, Vienna, Austria http://www.unodc.org/documents/southeastasiaandpacific// 2011/11/ats-2011/2011_Patterns_and_Trends_of_ATS_and_Other_Drugs.pdf. Global SMART Programme, 2013. Patterns and Trends of Amphetamine—Type Stimulants and Other Drugs: Challenges for Asia and the Pacific. United Nations Office on Drugs and Crime, Vienna, Austria http://www.unodc.org/documents/ southeastasiaandpacific//Publications/2013/ats-2013/2013_Regional_ATS_Report_ web.pdf. Global SMART Programme, 2015. The Challenge of Synthetic Drugs in East and South-East Asia and Oceania: Trends and Patterns of Amphetamine-type Stimulants and New Psychoactive Substances. United Nations Office on Drugs and Crime, Vienna, Austria https://www.unodc.org/documents/southeastasiaandpacific/Publications/2015/ drugs/ATS_2015_Report_web.pdf. Gracia-Lor, E., Zuccato, E., Castiglioni, S., 2016. Refining correction factors for back-calculation of illicit drug use. Sci. Total Environ. 573:1648–1659. http://dx.doi.org/10.1016/ j.scitotenv.2016.09.179. International Narcotics Control Board, 2014. Precursors and Chemicals Frequently Used in the Illicit Manufacture of Narcotic Drugs and Psychotropic Substances. United Nations, Vienna, Austria. Jing, L., Hou, L., Peng, D., Yang, J., Li, K., Xu, Z., Wang, C., Zhang, H., Li, X., 2014. Estimation of amphetamine and methamphetamine uses in Beijing through sewage-based analysis. Sci. Total Environ. 490:724–732. http://dx.doi.org/10.1016/j.scitotenv.2014.05. 042. Kasprzyk-Hordern, B., Baker, D.R., 2012. Enantiomeric profiling of chiral drugs in wastewater and receiving waters. Environ. Sci. Technol. 46 (3):1681–1691. http://dx.doi. org/10.1021/es203113y. Kasprzyk-Hordern, B., Kondakal, V.V., Baker, D.R., 2010. Enantiomeric analysis of drugs of abuse in wastewater by chiral liquid chromatography coupled with tandem mass spectrometry. J. Chromatogr. A 1217 (27):4575–4586. http://dx.doi.org/10.1016/j. chroma.2010.04.073. Khan, U., Nicell, J.A., 2012. Sewer epidemiology mass balances for assessing the illicit use of methamphetamine, amphetamine and tetrahydrocannabinol. Sci. Total Environ. 421:144–162. http://dx.doi.org/10.1016/j.scitotenv.2012.01.020. Khan, U., Nuijs, A.L.N.V., Jing, L., Maho, W., Peng, D., Li, K., Hou, L., Zhang, J., Meng, X., Li, X., 2014. Application of a sewage-based approach to assess the use of ten illicit drugs in four Chinese megacities. Sci. Total Environ. 487 (14):710–721. http://dx.doi.org/10. 1016/j.scitotenv.2014.01.043. Kim, K.Y., Lai, F.Y., Kim, H.Y., Thai, P.K., Mueller, J.F., Oh, J.E., 2015. The first application of wastewater-based drug epidemiology in five South Korean cities. Sci. Total Environ. 524-525:440–446. http://dx.doi.org/10.1016/j.scitotenv.2015.04.065. Kunalan, V., Nic Daéid, N., Kerr, W.J., Buchanan, H.A.S., McPherson, A.R., 2009. Characterization of route specific impurities found in methamphetamine synthesized by the Leuckart and reductive amination methods. Anal. Chem. 81 (17):7342–7348. http:// dx.doi.org/10.1021/ac9005588. Lai, F.Y., Bruno, R., Leung, H.W., Thai, P.K., Ort, C., Carter, S., Thompson, K., Lam, P.K.S., Mueller, J.F., 2013. Estimating daily and diurnal variations of illicit drug use in Hong

Kong: a pilot study of using wastewater analysis in an Asian metropolitan city. Forensic Sci. Int. 233 (1–3):126–132. http://dx.doi.org/10.1016/j.forsciint.2013.09.003. Lee, J.S., Han, E.Y., Lee, S.Y., Kim, E.M., Park, Y.H., Lim, M.A., Chung, H.S., Park, J.H., 2006. Analysis of the impurities in the methamphetamine synthesized by three different methods from ephedrine and pseudoephedrine. Forensic Sci. Int. 161 (2–3): 209–215. http://dx.doi.org/10.1016/j.forsciint.2006.02.054. Metcalfe, C., Tindale, K., Li, H., Rodayan, A., Yargeau, V., 2010. Illicit drugs in Canadian municipal wastewater and estimates of community drug use. Environ. Pollut. 158 (10): 3179–3185. http://dx.doi.org/10.1016/j.envpol.2010.07.002. Nuijs, A.L.N.V., Pecceu, B., Theunis, L., Dubois, N., Charlier, C., Jorens, P.G., Bervoets, L., Blust, R., Neels, H., Covaci, A., 2009. Cocaine and metabolites in waste and surface water across Belgium. Environ. Pollut. 157 (1):123–129. http://dx.doi.org/10.1016/j. envpol.2008.07.020. Office of China National Narcotic Control Commission, 2009. Annual Report on Drug Control in China. Office of China National Narcotic Control Commission, Beijing, China http://www.nncc626.com/index/ndbg.htm. Office of China National Narcotic Control Commission, 2015. Annual Report on Drug Control in China. Office of China National Narcotic Control Commission, Beijing, China http://www.nncc626.com/index/ndbg.htm. Office of China National Narcotic Control Commission, 2016. Drug Situation Report of 2015. Office of China National Narcotic Control Commission, Beijing, China http:// news.xinhuanet.com/live/2016-02/18/c_128730815.htm. Petrie, B., Youdan, J., Barden, R., Kasprzyk-Hordern, B., 2016. New framework to diagnose the direct disposal of prescribed drugs in wastewater—a case study of the antidepressant fluoxetine. Environ. Sci. Technol. 50 (7):3781–3789. http://dx.doi.org/10.1021/ acs.est.6b00291. Remberg, B., Stead, A., 1999. Drug characterization/impurity profiling, with special focus on methamphetamine: recent work of the United Nations International Drug Control Programme. Bull. Narc. 51 (1), 2. Senta, I., Krizman, I., Ahel, M., Terzic, S., 2014. Assessment of stability of drug biomarkers in municipal wastewater as a factor influencing the estimation of drug consumption using sewage epidemiology. Sci. Total Environ. 487:659–665. http://dx.doi.org/10. 1016/j.scitotenv.2013.12.054. Thai, P.K., Jiang, G., Gernjak, W., Yuan, Z., Lai, F.Y., Mueller, J.F., 2014. Effects of sewer conditions on the degradation of selected illicit drug residues in wastewater. Water Res. 48:538–547. http://dx.doi.org/10.1016/j.watres.2013.10.019. Thomas, K.V., Bijlsma, L., Castiglioni, S., Covaci, A., Emke, E., Grabic, R., Hernández, F., Karolak, S., Kasprzyk-Hordern, B., Lindberg, R.H., 2012. Comparing illicit drug use in 19 European cities through sewage analysis. Sci. Total Environ. 432 (16):432–439. http://dx.doi.org/10.1016/j.scitotenv.2012.06.069. UNODC, 2015. World drug report 2015, United Nations Office of drugs and Crime. Austria, Vienna http://www.unodc.org/wdr2015/. Valentine, J.L., Kearns, G.L., Sparks, C., Letzig, L.G., Valentine, C.R., Shappell, S.A., Neri, D.F., Dejohn, C.A., 1995. GC-MS determination of amphetamine and methamphetamine in human urine for 12 hours following oral administration of dextro-methamphetamine: lack of evidence supporting the established forensic guidelines for methamphetamine confirmation. J. Anal. Toxicol. 19 (7), 581–590. van Nuijs, A.L., Abdellati, K., Bervoets, L., Blust, R., Jorens, P.G., Neels, H., Covaci, A., 2012. The stability of illicit drugs and metabolites in wastewater, an important issue for sewage epidemiology? J. Hazard. Mater. 239-240:19–23. http://dx.doi.org/10.1016/ j.jhazmat.2012.04.030. Wang, G., Shi, J., Chen, N., Xu, L., Li, J., Li, P., Sun, Y., Lu, L., 2013. Effects of length of abstinence on decision-making and craving in methamphetamine abusers. PLoS One 8 (7), e68791. http://dx.doi.org/10.1371/journal.pone.0068791. Zhu, B.L., Meng, L., Zheng, K.F., 2014. Inspection and analysis of mixed drugs recently seized in China. Forensic Sci. Int. 242:E44–E47. http://dx.doi.org/10.1016/j.forsciint. 2014.07.013. Zuccato, E., Chiabrando, C., Castiglioni, S., Calamari, D., Bagnati, R., Schiarea, S., Fanelli, R., 2005. Cocaine in surface waters: a new evidence-based tool to monitor community drug abuse. Environ. Health 4 (1):1–7. http://dx.doi.org/10.1186/1476-069X-4-14. Zuccato, E., Chiabrando, C., Castiglioni, S., Bagnati, R., Fanelli, R., 2008. Estimating community drug abuse by wastewater analysis. Environ. Health Perspect. 116 (8): 1027–1032. http://dx.doi.org/10.1289/ehp.11022.