Wastewater Surveillance for Benzodiazepines in Wuhu, China: Occurrence, Removal, and Consumption Patterns

Abstract

Benzodiazepines (BZDs), potent sedative and hypnotic drugs widely prescribed in psychiatry, pose a high risk of dependence and are globally abused. This study used wastewater-based epidemiology to investigate the consumption patterns of BZDs across four wastewater treatment plants (WWTPs) in Wuhu, China. A total of 16 BZDs and three metabolites were detected in influents and effluents, with concentrations reaching up to 90 ng/L (quetiapine fumarate) and 18.4 ng/L (diazepam). Most BZDs had a poor removal efficiency except quetiapine fumarate (>98% removal). The consumptions of BZDs in WWTPs ranged from <0.02 (lormetazepam) to 2700 mg/day/1000 people (quetiapine fumarate). Seasonal variation was found in BZD usage, where the consumptions in winter and spring were significantly higher than those in summer and autumn. It was worth noting that nimetazepam may be abused during the sampling campaign. Urban areas with higher housing prices match higher BZD consumption, correlating with greater stress and insomnia rates. This study reveals the relationship between socioeconomic factors and BZD consumption patterns, provide a new path to addressing community public health.

1. Introduction

Benzodiazepines (BZDs), including alprazolam and lorazepam, are widely prescribed in Chinese hospital outpatient clinics for insomnia due to their direct central nervous system effects, low toxicity, and minimal side effects, making them a first-line treatment for improving sleep quality [1,2,3]. However, prolonged use of BZDs can lead to dependence, and misuse beyond medical purposes can result in addiction. BZDs are frequently combined with alcohol, opioids, and other addictive substances, prompting many countries and regions to classify them as key monitored drugs [4,5,6]. The United States Food and Drug Administration (FDA) has issued a black-box warning highlighting their risks of abuse and dependence [7]. Additionally, the FDA monitors BZD use through Prescription Drug Monitoring Programs (PDMPs), which help healthcare providers identify patients with excessive or inappropriate co-use of these drugs [5]. In China, 38 types of BZDs are listed as controlled psychotropic substances. However, illegal activities such as drug diversion and dark web sales have complicated regulatory efforts [8,9]. Therefore, monitoring BZD use is critical for preventing drug abuse.

Studies indicate that BZDs enter wastewater treatment systems through human excretion and improper disposal, but conventional treatment processes often fail to fully remove them, leading to their discharge into surface and groundwater [10,11]. It has been reported that alprazolam, oxazepam, temazepam, and diazepam were the most commonly detected BZDs, with concentration levels of 1–1766 ng/L, in wastewater treatment facilities from the UK [12], Slovakia [13], India [14], and Europe [15]. Studies have shown that BZDs are widely present in the aquatic environment. For example, BZDs (including alprazolam, diazepam, clonazepam, lorazepam, and oxazepam) are present in various aquatic matrices in Brazil, such as drinking water, surface water, groundwater, seawater, estuaries, and influents/effluents, with the concentrations ranging from 0.14 to 89,900 ng/L [16]. At present, there are no standards and safe concentrations for BZDs in the pollutant emission standards. The residual BZDs affect the ecological environment and public health, which should be paid attention to.

Wastewater-based wastewater epidemiology (WBE) is a method that can provide information on drug consumption level, spatial distribution, and temporal changes in a short time by using parameters such as wastewater detection, human excretion rate, and population data [17]. It has been widely used to analyze the abuse of cocaine, cannabis, MDMA, methamphetamine, and other illicit drugs [18,19,20,21], but also to estimate the consumption of legally addictive substances (e.g., tobacco and alcohol) [22,23,24] and commonly used drugs (e.g., antibiotics and metformin) [25,26,27,28]. It has also been employed to evaluate lifestyles and address public health concerns. After people take BZDs, the drugs enter the urban wastewater system in their original form or as metabolites, with concentrations ranging from ng/L to μg/L [29,30,31,32]. Thus, WBE can achieve real-time monitoring of BZDs. Previous studies have reported per capita consumption, removal rate, and emission of BZDs into surface water though wastewater surveillance. Baker et al. estimated diazepam (28 mg/day/1000 people) and temazepam (75 mg/day/1000 people) consumptions in the UK, consistent with 37 and 54 mg/day/1000 people based on National Health Service prescription data [33]. Lei et al. investigated the occurrence of 17 BZDs and their three transformation products in Guangdong, China [29]. The results showed that most of the BZDs were found in influents, effluents, and sludge. Gracia-Lo et al. and Rousis et al. estimated BZD usage in different periods of the COVID-19 pandemic [34,35]. In addition, diazepam was detected in drinking water, with the concentration reaching 1.22 ng/L [36]. However, seasonal variation of BZD use is rarely involved, due to the limited number of samples in those studies mentioned previously.

The objective of this study was to examine BZD occurrence and the spatial–temporal variations in BZD consumption by analyzing wastewater samples collected quarterly from four urban WWTPs (covering ~89% of the resident population) in Wuhu, China.

2. Material and Methods

2.1. Standards and Reagents

The 25 sedative and hypnotic drugs included 23 benzodiazepines (alprazolam, α-hydroxy alprazolam, chlorodiazepam, clonazepam, 7-aminoclonazepam, delorazepam, diazepam, nordiazepam, oxazepam, temazepam, estazolam, etizolam, flubromazepam, flunitrazepam, lormetazepam, lorazepam, meclonazepam, midazolam, nimetazepam, 7-aminonimetazepam, nitrazepam, pyrazolam, and triazolam) and 2 non-benzodiazepines (zolpidem and quetiapine fumarate). These standards (100 μg/mL, purity > 99%) and their internal standards (10 μg/mL, purity > 99%) were dissolved in methanol. All reagents were of high purity (see Supporting Information).

2.2. Collection of Wastewater Samples

Wuhu is one of the core cities of the Yangtze River Delta urban agglomeration. Due to its unique geographical advantages, Wuhu is developing rapidly and has been ranked among the third-tier cities in China. Influent and effluent wastewater samples were collected from four WWTPs in Wuhu, China (Figure S1). These WWTPs (WWTP-1, WWTP-2, WWTP-3, and WWTP-4), located in different main urban areas, capture more than 89% of the resident population in Wuhu (Table S1). Each WWTP received primarily domestic wastewater (>90%) and operated normally during the sampling period. All wastewater influent samples were collected as 24 h flow-integrated composite samples using an automatic sampler. The sampling campaign was conducted from 19 to 25 September 2020 (autumn), 18 to 24 January 2021 (winter), 10 to 16 May 2021 (spring), and 24 to 30 July 2021 (summer). A total of 54 samples were collected each quarter (in months 1, 5, 7, and 9, representing the typical four seasons in China) over 4 quarters (total of 212 samples). Sampling dates were selected to avoid public holidays and festivals to ensure representative wastewater samples. After, collection samples were immediately acidified to pH = 2 using HCl and stored at −20 °C until analysis.

2.3. Sample Preparation and Analysis

The sample was pretreated using solid-phase extraction (SPE) with a PCX column to reduce matrix effects and enrich the target substances [37]. The specific method is described in Text S1. In this study, the samples were analyzed using high-performance liquid chromatography–tandem mass spectrometry (HPLC-MS/MS; Exion LC-6500+ Triple Quad, AB Sciex, Framingham, MA, USA). The 25 target substances were separated using an Agilent Infinity Lab Poroshell HPH-C18 liquid chromatography column (2.1 × 100 mm, 2.7 μm, Santa Clara, CA, USA). The mobile phase consisted of a 10 mM ammonium format with 0.1% formic acid (phase A) and acetonitrile (phase B). The gradient elution procedure was as follows: 0–0.5 min, 10% phase B; 0.5–5 min, 10–90% phase B; 5–5.4 min, 90% phase B; 5.4–5.8 min, 90–10% phase B; 5.8–8 min, 10% phase B. The total run time was 8 min. The mass spectrometer was operated in scheduled multiple reaction monitoring (MRM) mode using an electrospray ionization (ESI) source. The ion source voltage was set to 4500 V, the ion source temperature to 450 °C, the gas curtain pressure to 25 psi, the drying gas pressure to 65 psi, the auxiliary gas pressure to 55 psi, and the collision cell pressure to 6 psi. Other mass spectrometry parameters are listed in Table S2.

2.4. Quality Assurance and Quality Control

Pure water samples and wastewater samples with different gradients (1, 5, and 10 ng/mL) were prepared to determine recovery rates and matrix effects. Recovery rates ranged from 80% to 120%, and matrix effects were within ±20%. A series of gradient standard samples (0.01–10 ng/mL) in a 1:1 methanol–water solution was used. The correlation coefficients (r²) of the standard curves for all substances were above 0.99, accuracy ranged from 85% to 115%, and RSD values were less than 15%. See Text S2 and Tables S5 and S6 for details on the experimental setup and data from the method validation.

2.5. WBE Estimation and Statistical Analysis

The apparent removal rate can be used to measure the removal effect of various drugs by wastewater treatment facilities. Its calculation formula is as follows:

$$Apparent removal rate \left(\%\right)={\displaystyle \frac{C_{in}-C_{eff}}{C_{in}}}\times 100\%$$

(1)

In Formula (1), $C_{in}$ represents the concentration of the target substance in influents and $C_{eff}$ represents the concentration of the target substance in effluents. Based on the principles of wastewater epidemiology, the load and consumption of sedative and hypnotic drugs in this study were calculated using the drug concentration, excretion rate, influent flow of the WWTP, and service population. The calculation method is shown in the following formula.

$$Load \left(\mathrm{m}\mathrm{g}/\mathrm{d}\mathrm{a}\mathrm{y}/1000 \mathrm{p}\mathrm{e}\mathrm{o}\mathrm{p}\mathrm{l}\mathrm{e}\right)={\displaystyle \frac{C_{in} \left(\mathrm{n}\mathrm{g}/\mathrm{L}\right)\times F_{in} \left({\mathrm{m}}^3\right)}{\frac{POP}{1000}}}\times {\displaystyle \frac{1}{{10}^6}}\left({\displaystyle \frac{\mathrm{m}\mathrm{g}}{\mathrm{n}\mathrm{g}}}\right)$$

(2)

In Formula (2), ${\mathrm{C}}_{\mathrm{i}\mathrm{n}}$ is the concentration of the target in influents, $F_{in}$ stands for the daily influent flow of the WWTP, and $POP$ stands for the population served by the WWTP. The concentrations that were below the LOD or LOQ were substituted with LOD (LOQ)/2 values for statistical analysis (this treatment is also used for the concentration in Formula (1)).

$$Consumption \left(\mathrm{m}\mathrm{g}/\mathrm{d}\mathrm{a}\mathrm{y}/1000 \mathrm{p}\mathrm{e}\mathrm{o}\mathrm{p}\mathrm{l}\mathrm{e}\right)=Load \left(\mathrm{m}\mathrm{g}/1000 \mathrm{p}\mathrm{e}\mathrm{o}\mathrm{p}\mathrm{l}\mathrm{e}/\mathrm{d}\mathrm{a}\mathrm{y}\right)\times {\displaystyle \frac{1}{R_e}}\times {MW}_{ratio}$$

(3)

In Formula (3), $\mathrm{L}\mathrm{o}\mathrm{a}\mathrm{d}$ is the drug loading level, $R_e$ represents the excretion rate of biomarkers of various substances in human urine; and ${MW}_{ratio}$ represents the molar mass ratio of the target substance matrix to its biomarker.

Consumptions of oxazepam and lorazepam were estimated by Formulas (4) and (5), while others used Formula (3). The urine excretion rate of target substances can be found in Table S4.

$${Consumption}_{OZP}=\left[{Load}_{OZP}-\left({Consumption}_{DZP}\times R_{e1}\times {MW}_1\right)\right]\times {\displaystyle \frac{1}{R_{e2}}}\times {MW}_2$$

(4)

In Formula (4), ${Load}_{OZP}$ is the oxazepam loading level, ${Consumption}_{DZP}$ is the diazepam consumption level, $R_{e1}$ represents the urinary excretion rate of oxazepam as metabolites of diazepam, ${MW}_1$ represents the molar mass ratio of oxazepam to diazepam, $R_{e2}$ represents the urinary excretion rate of oxazepam as maternal drug, and ${MW}_2$ represents the molar mass ratio of the metabolites of oxazepam to oxazepam.

$${Consumption}_{LZP}=\left[{Load}_{LZP}-\left({Consumption}_{LMP}\times R_{e3}\times {MW}_3\right)\right]\times {\displaystyle \frac{1}{R_{e4}}}\times {MW}_4$$

(5)

In Formula (5), ${\mathrm{L}\mathrm{o}\mathrm{a}\mathrm{d}}_{\mathrm{L}\mathrm{Z}\mathrm{P}}$ is the lorazepam loading level, ${\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{L}\mathrm{M}\mathrm{P}}$ is the lormetazepam consumption level, $R_{e3}$ represents the urinary excretion rate of lorazepam as metabolites of lormetazepam, ${MW}_3$ represents the molar mass ratio of lorazepam to lormetazepam, $R_{e4}$ represents the urinary excretion rate of lorazepam as a maternal drug, and ${MW}_4$ represents the molar mass ratio of the metabolites of lorazepam to lorazepam.

Different sedative and hypnotic drugs have varying therapeutic effects, leading to differences in their frequency of use and single doses. To standardize the dosage, the adult defined daily dose ($DDD$) is used, allowing for the equivalent conversion of different drugs. The calculation formula is as follows:

$${DDD}_s \left(\mathrm{d}\mathrm{o}\mathrm{s}\mathrm{e}\mathrm{s}/\mathrm{d}\mathrm{a}\mathrm{y}/1000 \mathrm{p}\mathrm{e}\mathrm{o}\mathrm{p}\mathrm{l}\mathrm{e}\right)={\displaystyle \frac{Consumption}{DDD}}\left({\displaystyle \frac{\mathrm{m}\mathrm{g}/1000 \mathrm{p}\mathrm{e}\mathrm{o}\mathrm{p}\mathrm{l}\mathrm{e}/\mathrm{d}\mathrm{a}\mathrm{y}}{\mathrm{m}\mathrm{g}}}\right)$$

(6)

In Formula (6), ${DDD}_s$ represents the average daily dose of each drug per thousand people, and $\mathrm{D}\mathrm{D}\mathrm{D}$ represents the average daily dose (mg) of the drug used by adults for main therapeutic purposes. These data are from the WHO.

Data analysis was performed using SPSS 25. One-way ANOVA was used to compare spatiotemporal data, with differences considered statistically significant at p < 0.05.

3. Results and Discussion

3.1. Occurrence of Benzodiazepines and Metabolites in Wastewater

Table 1. The detection frequency (%), concentration (ng/L), and load (mg/day/1000 people) of 25 substances in WWTPs.

Compounds	Influent Concentration				Effluent Concentration				Load a
	Freq	Range	Mean ± SD	Median	Freq	Range	Mean ± SD	Median	Mean ± SD
Alprazolam	93.9	<LOD~2.5	0.9 ± 0.5	0.8	99.0	<LOD~1.9	0.8 ± 0.4	0.7	0.2 ± 0.2
α-Hydroxy alprazolam	30.1	<LOD~16.8	2.1 ± 3.7	<LOD	26.4	<LOD~18.8	1.9 ± 3.7	<LOD	0.8 ± 0.9
Clonazepam	19.8	<LOD~14.5	0.8 ± 2.2	<LOD	14.1	<LOD~11.0	0.4 ± 1.3	<LOD	0.2 ± 0.6
7-Aminoclonazepam	16.9	<LOD~3.4	0.2 ± 0.5	<LOD	4.7	<LOD~0.7	0.01 ± 0.02	<LOD	0.03 ± 0.1
Diazepam	100	0.7~52.8	6.7 ± 7.9	4.2	100	0.7~18.4	4.0 ± 3.0	3.0	1.7 ± 2.3
Nordiazepam	19.8	<LOD~1.7	0.1 ± 0.2	<LOD	21.6	<LOD~0.5	0.1 ± 0.1	<LOD	0.04 ± 0.1
Temazepam	2.83	<LOD~0.8	0.01 ± 0.1	<LOD	25.4	<LOD~0.6	0.1 ± 0.1	<LOD	0.002 ± 0.01
Oxazepam	12.2	<LOD~4.6	0.2 ± 0.7	<LOD	23.5	<LOD~6.2	0.4 ± 0.9	<LOD	0.02 ± 0.1
Estazolam	35.8	<LOD~0.8	0.1 ± 0.2	<LOD	39.6	<LOD~0.8	0.1 ± 0.1	<LOD	0.03 ± 0.1
Flunitrazepam	3.5	<LOD	<LOD	<LOD	3.5	<LOD~2.3	0.07 ± 0.4	<LOD	-
Lormetazepam	13.2	<LOD~0.8	0.04 ± 0.1	<LOD	2.8	<LOD~0.3	<LOD	<LOD	0.01 ± 0.03
Lorazepam	6.6	<LOD~5.1	0.2 ± 0.8	<LOD	28.3	<LOD~10.9	1.3 ± 2.4	<LOD	0.1 ± 0.2
Midazolam	17.9	<LOD~0.2	0.01 ± 0.03	<LOD	10.3	<LOD~0.2	<LOD	<LOD	0.004 ± 0.01
Nitrazepam	3.7	<LOD~0.2	0.01 ± 0.03	<LOD	1.8	<LOD~4.9	0.1 ± 0.6	<LOD	0.04 ± 0.3
Nimetazepam	39.6	<LOD~3.7	0.4 ± 0.7	<LOD	50.0	<LOD~2.3	0.3 ± 0.4	0.052	0.1 ± 0.2
7-Aminonimetazepam	4.7	<LOD~0.9	0.03 ± 0.1	<LOD	5.6	<LOD~0.5	0.01 ± 0.03	<LOD	0.01 ± 0.04
Triazolam	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Chlorodiazepam	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Delorazepam	0.9	<LOD~0.1	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Etizolam	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Flubromazepam	0.9	<LOD~0.4	<LOD	<LOD	2.8	<LOD~0.4	0.01 ± 0.01	<LOD	0.001 ± 0.01
Meclonazepam	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Pyrazolam	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Zolpidem	0	<LOD	<LOD	<LOD	0	<LOD	<LOD	<LOD	-
Quetiapine fumarate	97.1	<LOD~90.0	4.4 ± 12.4	1.2	28.3	<LOD~0.6	0.05 ± 0.1	<LOD	1.3 ± 3.7

Note: ^a Calculation based on influent concentration.

details the detection frequency, concentration range, and load levels of 25 BZDs in influents and effluents of four WWTPs. Among 25 BZDs, 19 were found in the influent and 18 in the effluent, with concentrations from <LOD to 90 ng/L (clozapine), peaking in WWTP-3. Diazepam ranked second, with concentrations at 52.8 ng/L (WWTP-1) and the highest average load (1.7 ± 2.3 mg/day/1000 people), followed by quetiapine fumarate (1.3 ± 3.4 mg/day/1000 people). Diazepam had a 100% detection rate in influent, with quetiapine fumarate and alprazolam at 97.1% and 93.3%, respectively. Third-generation drugs like triazolam and etizolam, and the new sedative zolpidem, were absent in Wuhu’s wastewater. These results suggest that BZDs are widely used in the Wuhu area. Notably, clonazepam and nimetazepam were detected at rates of 13.2% and 39.6%, respectively. Lormetazepam, an illegal drug globally [6], and lorazepam, a common prescription drug and its metabolite [38], were also identified. With 75% of lorazepam excreted in urine [39], the origin of lormetazepam—whether from human metabolism or improper disposal—remains unclear. Nimetazepam, a third-generation drug, and its metabolite 7-aminonimetazepam, were found in all three WWTPs. Effluent samples contained most of the influent-detected substances, except lorazepam, with concentrations up to 18.4 ng/L (diazepam). These results align with Kosjek et al.’s findings in Slovenia [10] and studies by Lei, Wang, and others [29,30,40] in China.

At the same time, the apparent removal rates of benzodiazepines in four WWTPs in Wuhu were compared (Figure 1). The results demonstrated varying removal efficiencies of BZDs across different wastewater treatment processes. The four WWTPs in this study primarily employed two distinct processes: anaerobic/anoxic/oxic (A2O) processes (WWTP-1, -2, and -4) and the oxidation ditch activated sludge process (WWTP-3). With the exception of quetiapine fumarate (showing >98% removal efficiency), α-hydroxy alprazolam, temazepam, and nimetazepam exhibited moderate removal rates of 39.5%, 45.6%, and 38.8%, respectively, through A2O processes. However, nine BZDs, including alprazolam, clonazepam, oxazepam, estazolam, lorazepam, and midazolam, demonstrated poor removal efficiency, with some even showing negative removal rates—indicating increased effluent concentrations after A2O treatment. Similarly, the oxidation ditch process showed limited effectiveness for most BZDs. Apart from quetiapine fumarate (>99% removal) and α-hydroxy alprazolam (60.2% removal), only four compounds (clonazepam, diazepam, midazolam, and nimetazepam) achieved >10% removal rates. Notably, alprazolam and lorazepam concentrations actually increased after oxidation ditch treatment. According to the result, for different BZDs we can choose different treatment processes; for example, for alprazolam, delorazepam, and nimetazepam we can choose A2O processes, and for clonazepam, oxazepam, and midazolam we can choose the oxidation ditch activated sludge process. In conclusion, current wastewater treatment processes in the Wuhu region demonstrate generally poor removal efficiency for BZDs, with some compounds showing concentration increases through treatment processes. The results showed that the removal rate of psychotropic drugs in urban WWTPs has been very low [41,42]. In Shanghai [43], the removal rates of alprazolam, nordiazepam, and oxazepam were all below 40%, while the removal rates of diazepam were negative, and the removal rates of lorazepam and temazepam were better. The removal efficiency of various BZDs in wastewater and sludge in 11 WWTPs in Guangdong Province was studied by Lei et al. [29]. The results showed that the apparent removal rate of alprazolam and temazepam reached more than 50%, but the removal rate of oxazepam was low. Similarly, the results of the whole process section of two different WWTPs in in Guangzhou [29] showed that the oxidation ditch process and A2O process do not have an obvious removal effect on BZDs. This is consistent with the results of the present study. The biodegradation and sludge adsorption processes in wastewater treatment systems have little effect on the degradation and adsorption of BZDs and can even increase the concentration of drugs in effluent. Currently, BZDs cannot be completely removed by urban WWTPs, and these drugs can be discharged into surface water, posing a threat to the aquatic environment [16,43,44,45,46]. Therefore, it is necessary to develop more effective technologies for their removal, such as advanced oxidation technologies (UV/H₂O₂ [47], Fenton oxidation [48], and UV/hydrogen peroxide treatment [48]).

3.2. Temporal and Spatial Variations of BZD Consumption in Wuhu

Statistically significant correlations (p < 0.05) were found between parent BZDs (alprazolam, diazepam, nimetazepam, and clonazepam) and their metabolites (α-hydroxy alprazolam, nordiazepam, temazepam, 7-aminometazepam, and 7-anminoclonazepam). These results indicated the BZDs’ residues in wastewater samples highly likely originated from human consumption. Thus, metabolites can be used to back-calculate the consumption of parent substances. The results show that among the four water treatment plants in Wuhu, the consumption range of BZDs is <0.02–2700 mg/day/1000 people. Quetiapine fumarate is the most consumed, at 127.7 ± 373.8 mg/day/1000 people, followed by midazolam (21.1 ± 55.9 mg/day/1000 people). Alprazolam, estazolam, nimetazepam, and nitrazepam have similar consumption levels. Among the 13 drugs, alprazolam (4.7 ± 4.9 doses/day/1000 people) has the highest per capita dosage, followed by midazolam (1.4 ± 3.7 doses/day/1000 people) and estazolam (1.2 ± 1.9 doses/day/1000 people). Quetiapine fumarate has the lowest per capita dosage, at 0.3 ± 0.9 doses/day/1000 people. More detailed drug data can be found in Tables S7 and S8.

There are significant differences in per capita drug dosages across different WWTPs, running up to 8.8 doses/day/1000 people (WWTP-1, alprazolam). More detailed drug data can be found in Table S7 and Figure S3. Alprazolam consumption is higher in WWTP-1 (8.8 doses/day/1000 people) and WWTP-2 (5.3 doses/day/1000 people) than in WWTP-3 (2.8 doses/day/1000 people) and WWTP-4 (1.8 doses/day/1000 people). Similarly, midazolam’s per capita drug dose is higher in WWTP-1 and WWTP-2 than in WWTP-3 and WWTP-4 (Figure S3). In WWTP-1 and WWTP-2, the per capita drug dose of nimetazepam is the highest, at 1.1 doses/day/1000 people, while in WWTP-3, the per capita drug dose of lormetazepam is the highest, at 0.4 doses/day/1000 people (Figure S3). One-way ANOVA results showed a significant spatial difference in per capita total dose (p < 0.05). The consumption level in WWTP-1 and WWTP-2, located in the city center, is significantly higher than that in WWTP-3 and WWTP-4, located in the suburban economic development zone. The spatial variation trend of per capita BZD consumption in Wuhu is higher in city centers and lower in towns and suburbs, similar to the findings of Wang et al. [30] on diazepam, oxazepam, and temazepam in wastewater from the Beijing WWTP. This spatial variation pattern may be related to the economic development level of the service areas of the WWTPs.

In addition, seasonal influent samples from Wuhu revealed significant seasonal differences in the consumption of all sedative–hypnotic medications (Figure 2; p < 0.05). The consumption of commonly used BZDs shows seasonal variations, with higher levels in spring and winter and lower levels in summer and autumn. Further comparisons of drug usage in spring and winter show that diazepam (0.07 doses/day/1000 people in spring vs. 0.15 doses/day/1000 people in winter) and alprazolam (5.0 doses/day/1000 people in spring vs. 6.4 doses/day/1000 people in winter) have higher consumptions in winter. Notably, there is a sharp increase in the consumption of nitrazepam and nimetazepam in autumn in the Wuhu region (Figure 2), reaching 2.7 doses/day/1000 people and 3 doses/day/1000 people, respectively. Nitrazepam and 7-aminonimetazepam are metabolites of the illegal drug nimetazepam, and these three substances were detected in influents of all four WWTPs in autumn. Lormetazepam was primarily detected in summer and autumn, with doses of 0.04 doses/day/1000 people and 0.02 doses/day/1000 people, respectively. Additionally, a study on the weekly variation in the total dose of drugs in the influent (Figure 3) showed no significant differences in BZD consumption over seven consecutive days (p > 0.05). This suggests that fluctuations in sedative doses in the influent samples from urban WWTPs are random and not substantial over the course of one week. This indicates that residents in the Wuhu area use sedatives at a consistent dose every day.

Generally speaking, the housing price level of a region can represent the economic development level of that region. In this study, house prices were selected as an evaluation index of the economic level of each region, and the relationship between the economic level and the consumption level of sedative and hypnotic drugs in these regions was further explored. The relationship between house prices and the corresponding consumption levels in the service areas of the WWTPs was tested using a t-test. The significant p-value of 0.027, which is less than 0.05, indicates that the difference is significant. As shown in Figure 4, the higher the house prices in a region, the higher the consumption level of sedatives. Monitoring sedative and hypnotic drugs in wastewater can inversely evaluate the sleep health of people in the area, as drugs are the first choice to indicate population health and related diseases. The higher the house prices in a region, the higher the insomnia rate among the population. Many domestic psychological studies have proven this trend [49,50,51], showing that in more developed areas, people face greater life pressures, making them more prone to anxiety and insomnia. Stress can affect the regulation of the nervous system and hormones, leading to anxiety and depressive behaviors [52,53,54,55,56]. Employees often face greater work pressure in winter, making them more prone to mental problems such as anxiety and insomnia [57,58,59]. Domestic studies have found that the peak of anxiety disorder visits mostly occurs during the winter solstice, while insomnia in winter shows a significant upward trend [57,58]. These results are similar to the seasonal variation study of 58 kinds of drugs and metabolites in the Tagus River in Spain conducted by Valcárcel et al. [60], which found that antidepressants and anxiolytics had the highest detection concentrations in winter. Similarly, a study by Mendoza et al. [61] on the seasonal variation of commonly used drugs and metabolites in surface water in Madrid, Spain, showed that the average concentrations of alprazolam, diazepam, and lorazepam in winter were higher than in summer (with a detection rate of 100%). These research results indicate that the consumption levels of drugs obtained through wastewater epidemiology and the corresponding disease patterns have good consistency, proving that the residual drug information in domestic wastewater can objectively reflect the disease and health status of the population in different time periods. This can provide a scientific basis for disease prevention and control and further analyze other factors in this area in the future, such as climate conditions, entertainment facilities, and their relationship with the quality of sleep of the population.

3.3. Comparison of BZD Consumption with Other WBE Studies and Data Source

There is a lack of research specifically reporting the consumption of BZDs. In this study, we selected several studies with population data and compared the consumption and dosage of different drugs across various regions using Formula (3). These regions include the United States, Britain, Australia, and several cities and regions in China. We used relevant prescription data from Wuhu and data from the “China Pharmaceutical Statistics Yearbook 2018” to “China Pharmaceutical Statistics Yearbook 2020” (referred to as the “Yearbook”), which includes three years of statistical data. Based on the data from the seventh national population census, we applied Formula (6) to estimate and compare drug dosages. According to the national production data of sedatives (including alprazolam, clonazepam, diazepam, estazolam, lorazepam, midazolam, nitrazepam, quetiapine fumarate, and zolpidem) in the “Yearbook”, the average dose of these nine drugs over the past three years was 20.8 doses/day/1000 people. The average dose of lorazepam was the highest, at 16.4 doses/day/1000 people, followed by alprazolam, with an average dose of 2.4 doses/day/1000 people. The average dose of BZDs calculated by wastewater-based epidemiology (WBE) in Wuhu is higher than that estimated by prescription data in Wuhu. Additionally, except for alprazolam and diazepam, most BZDs exceeded the estimated values in the “Yearbook”. The “Yearbook” and Wuhu prescription materials did not provide relevant data for nimetazepam, but nimetazepam and its metabolites were detected in the influent of four WWTPs in Wuhu. The dosage of nimetazepam and nitrazepam suddenly increased in autumn, and the dosage of nitrazepam was particularly high on weekends (2.6 doses/day/1000 people), while the dosage of nimetazepam was higher on Tuesdays (2.6 doses/day/1000 people) and weekends (1.1 doses/day/1000 people) (Figure S2). Nimetazepam may be abused in the Wuhu area, and according to the spatial distribution of nimetazepam and nitrazepam, it is very likely that benzodiazepines are being used illegally in the service area of WWTP-2.

Table 2. Load and dose (in brackets) of BZDs in different areas.

Location	Year	No. of WWTPs	Population (Million)	DDDS (doses/day/1000 people)
				Alprazolam	Diazepam	Oxazepam	Temazepam	Estazolam	Lorazepam
UK—London [11]	2011	1	3.4	-	ND (7.3 c)	9.3 (0) e	58.5 (3.1)	-	-
Australia—South [67]	2019	4	1.2	0.2 (1.7)	1.2 (23.9)	153.9 (2)	74.9 (3.4)	-	2.3 (1.2)
US—eastern Kentucky [68]	2018	2	0.19	66	3	3	16	-	8
			0.15	68	<LOQ	4	37	-	12
US—ten states [69]	2020	35	5.6	ND	ND	-	21.1 (3.1)	-	44.3 (68)
US—New York [70]	2013	2	0.12	1.7	0.92	4.84	-	-	7.74
UK—south-west [12]	2014~18	- a	0.89	-	-	54.7	44.4	-	-
Slovakia [13]	2013	8	1.11	-	-	30.1	-	-	-
Norway—Oslo [71]	2014	1	0.6	ND	-	140.7	-	-	-
Norway—Trondheim [71]	2014	1	0.18	630.2	-	106.5	-	-	-
India—Saidpur [15]	2013	1	0.35	0.5	0.4	-	-	-	-
India—Beur [15]	2013	1	0.26	0.3	0.3	-	-	-	-
India—Coimbatore [15]	2013	1	0.35	-	-	1.6	-	-	-
India—Udupi [15]	2013	1	0.01	-	1.3	-	-	-	4.7
India—Manipal [15]	2013	1	0.01	1.2	32.6	2.3	-	-	3.3
China—Beijing [30]	2013	13	11.6		0.5 (0.3 d)	1.9 (0)	0.6 (0)	-	-
China—Shanghai [43]	2014	5	1.0	4(19.8)	2.8 (0.7 c)	5 (0.1)	6.9 (0.3)	1.6 (53.1)	18.7 (9.5)
China—Guangdong [29]	2018~19	11	3.9	ND	0.3 (0.3 c)	1 (0)	-	ND	-
China—Taipei [72]	2022~24	4	6.5	15.12	22.22	-	-	2.98	-
China—Wuhu (this study)	2020~21	4	1.8	0.2 (1.3 b)	1.3 (0.1 c)	0 (0)	0 (0)	ND (1.2)	ND
Medical statistics data	2018~20	-	(2.37)	(0.08)	(0.0007)	-	(1.2)	(0.02)	-

Notes: ^a means data are not available; ^b using α-hydroxy alprazolam as DTR; ^c using nordiazepam as DTR; ^d using temazepam as DTR; ^e 0 in brackets means only found as diazepam metabolites; unlabeled data mean using parent compounds as DTR; ND = none detected; Medical statistics data means data from the China Medical Statistics Annual Report, which were provided by government departments: the Industry and Information Sectors.

shows that there are significant differences in drug consumption preferences across different regions. Among several commonly used benzodiazepines, the consumption of diazepam is higher in the UK and Australia, reaching up to 7.3 doses/day/1000 people and 23.9 doses/day/1000 people, respectively. The USA may lean towards consuming lorazepam, with a maximum of 68 doses/day/1000 people. In China, according to the analysis of prescription data from psychiatric clinics in recent years, the use of alprazolam, estazolam, and zolpidem is consistent [1,2,3,62,63,64], which is consistent with the results of this study. This indicates that traditional benzodiazepines (BZDs) are still the main sedative and hypnotic drugs used in Wuhu. Considering that sedative–hypnotic drugs are prescription drugs in most areas, their use may be related to their effectiveness, price, and doctors’ prescription preferences. Alprazolam, estazolam, nitrazepam, and clonazepam are relatively low in price, have a higher number of users, and are more widely accepted, which may lead to irrational drug use, such as the combined use of multiple sedative–hypnotics [1,2,65,66]. Due to different sampling times in each region and the global experience of the COVID-19 pandemic since 2020, the lives of residents in each country have been greatly affected. Therefore, the total consumption of drugs in different regions is not directly comparable. In the future, more research on benzodiazepines can be carried out, and the consumption levels in different regions can be compared horizontally to realize the analysis and evaluation of the sleep health of the global population.

3.4. Uncertainty Analysis

Wastewater-based epidemiology (WBE) involves several uncertainties, including sampling, analysis, the stability of target substances in wastewater, human metabolism data, and population data used in back-calculation. The relationship among these variables and the fingerprint of the chemical analyses lead to the obtained result. Preanalytical measurement error may occur during in-pipe and holding times. The samples’ travel time from Wuhu to Beijing took 3 days maximum. All wastewater samples in this study were stored at −20 °C and pH = 2 to ensure the stability of BZDs in actual samples. Deconjugation was not included in the sample preparation and may have caused underestimation of the consumptions of several BZDs (alprazolam, etizolam, lormetazepam, etc.) [73]. This study did not account for adsorption of these compounds to sewer biofilms, which may impact the per capita loadings reported in this study. In addition, while most drugs were analyzed using drug target residues (DTRs) as the matrix, limited research on maternal excretion rates compared to metabolites has resulted in lower accuracy in back-calculation. Therefore, future studies will prioritize metabolite-based calculations. In this study, population estimates were obtained from the survey data for calculating the per capita mass loading in Wuhu, which could have resulted in an over- or under-estimation of drug consumption rates, depending on the quality of the data. Additionally, this study was limited by the number of WWTPs, low sampling frequency, and limited sample size, necessitating long-term monitoring in future research.

4. Conclusions

This study comprehensively analyzed BZD consumption patterns across four WWTPs in Wuhu, China, integrating temporal, spatial, and socioeconomic factors. The data showed significant seasonal variation in BZD usage, where BZD consumption in winter and spring was significantly higher than in the other seasons. It was worth noting that nimetazepam may be abused in the catchment of WWTP-2. BZD consumption was higher in urban areas and lower in suburban areas, consistent with the trend of regional housing prices. The level of housing prices is related to residents’ life pressures, which in turn affect sleep health. In summary, this work provides important public health information regarding the use of BZDs, and offers a corresponding basis for future regulation.