Ecotoxicology and Environmental Safety 205 (2020) 111365

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Wastewater treatment plant workers’ exposure and methods for risk

evaluation of their exposure
Rui Lu a, b, Margit W. Frederiksen a, Katrine Uhrbrand a, Yanpeng Li b, Claus Østergaard c,
Anne Mette Madsen a, *
a
National Research Centre for the Working Environment, Lersø Parkallé 105, DK-2100, Copenhagen, Denmark
b
School of Water and Environment, Chang’an University, Xi’an, 710054, China
c
Department of Clinical Microbiology, Lillebaelt Hospital, Denmark

A R T I C L E I N F O A B S T R A C T

Keywords: Work in wastewater treatment plants (WWTPs) can be associated with respiratory symptoms and diarrhea. The
Airborne bacteria aim of this study was to obtain knowledge about WWTP workers’ exposure to airborne bacteria and endotoxin,
Endotoxin and the inflammatory potential (TIP) of their exposure, and to evaluate the risk posed by the exposure by 1)
Pathogens
calculating a hazard index and relating the exposure to suggested occupational exposure limits (OELs), 2) esti­
Inflammation
mating the potential deposition of bacteria in the airways, 3) relating it to the risk group classification of bacteria
by the European Union, and 4) estimating the TIP of the personal exposure. A cohort of 14 workers were followed
over one year. Bioaerosols were collected using personal and stationary samplers in a grid chamber house and an
aeration tank area. Airborne bacteria were identified using (MALDI-TOF MS), and TIP of exposure was measured
using HL-60 cells. A significant effect of season, work task, and person was found on the personal exposure. A
hazard index based on exposure levels indicates that the risk caused by inhalation is low. In relation to suggested
OELs, 14% and 34% of the personal exposure were exceeded for endotoxin (≥50 EU/m3) and bacteria (≥500
CFU/m3). At least 70% of the airborne bacteria in the grid chamber house and the aeration tank area could
potentially deposit in the lower respiratory tract. From the personal samples, three of 131 bacterial species,
Enterobacter cloacae, Staphylococcus aureus, and Yersinia enterocolitica are classified within Risk Group 2. Seven
additional bacteria from the stationary samples belong to Risk Group 2. The bacterial species composition was
affected significantly by season (p = 0.014) and by sampling type/area (p = 0.001). The TIP of WWTP workers’
exposure was higher than of a reference sample, and the highest TIP was measured in autumn. TIP of personal
exposure correlated with bacterial exposure. Based on the geometric average exposures to endotoxin (9.2 EU/
m3) and bacteria (299 CFU/m3) and based on the calculated hazard index, the risk associated with exposure is
low. However, since 43 of 106 exposure levels exceed suggested OELs, the TIP of exposure was elevated and
associated with bacterial exposure, and WWTP workers were exposed to pathogenic bacteria, a continued focus
on preventive measures is important. The identification of bacteria to species level in personal samples was
necessary in the risk assessment, and measurement of the microbial composition made the source tracking
possible.

1. Introduction 2017; Wang et al., 2019a), and they are considered a potential health
risk for the WWTP workers. Some of them are pathogenic bacteria, such
Work in wastewater treatments plants (WWTPs) can be associated as Enterococcus faecalis and E. faecium (Uhrbrand et al., 2017), Clos­
with respiratory symptoms and impaired pulmonary function tridium perfringens and Staphylococcus aureus (Fracchia et al., 2006).
(Cyprowski et al., 2015) and gastrointestinal problems such as diarrhea Airborne endotoxin has also been found in elevated levels in the air of
(Scarlett-Kranz et al., 1987). A large number of different airborne bac­ WWTPs (Oppliger et al., 2005). other studies also found it Airborne
teria has been found in WWTPs (Bruni et al., 2020; Uhrbrand et al., endotoxin has primarily been linked to lower respiratory and skin

* Corresponding author.
E-mail address: [email protected] (A.M. Madsen).

https://doi.org/10.1016/j.ecoenv.2020.111365
Received 13 June 2020; Received in revised form 14 August 2020; Accepted 13 September 2020
Available online 22 September 2020
0147-6513/© 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

symptoms (Smit et al., 2005). Exposure to endotoxin in WWTPs has also that a cohort of 14 WWTP workers were followed for a year to assess the
been associated with gastrointestinal problems (Jeggli et al., 2004). risk associated with exposure to microorganisms, as evaluated by
Whether endotoxin or specific gram-negative bacteria are the causal different risk assessment methods.
agent of the gastrointestinal symptoms is not known.
Using stationary samplers, the level of release and dispersal of bio­ 2. Materials and methods
aerosols in WWTPs was found to be associated with season (Fracchia
et al., 2006). The different steps in the treatment of wastewater consti­ 2.1. Wastewater treatment plant description
tute another important factor affecting the concentration of biaoerosol
components (Sánchez-Monedero et al., 2008; Uhrbrand et al., 2017). Sampling was conducted at a WWTP in Denmark, which serves
The grid is the first step in the wastewater treatment, and where the around 1 million people corresponding to 3 million liters of wastewater
highest levels of airborne bacteria may be present (Szyłak-Szydłowski every day. The WWTP is located on Zealand, close to the sea, and utilizes
et al., 2016). The aeration tank has also been identified as a source of an A2/O process, biological phosphorus removal and Bio-denitro
bacterial bioaerosols in WWTPs (Wang et al., 2019a). Exposure levels in process.
the WWTPs may also be affected by the work tasks performed (Heldal
et al., 2010). 2.2. Participants and sampling procedures
Several methods have been used to evaluate the risk associated with
exposure to bioaerosols in WWTPs. Some studies evaluated the risk of Sampling was performed by two occupational hygienists monthly
exposure by comparing the concentration of bacteria with local stan­ over a one-year period from June 2014 to June 2015.
dards. For example, Li et al. (2013a) reported that the concentration of To study personal exposure, a total of 14 wastewater workers
airborne actinomycetes and bacteria in general in a WWTP exceeded the employed at the WWTP were enrolled as participants and followed
threshold values recommended by American Conference of Govern­ through the one-year period. Of the 14 participants, 11 workers
mental Industrial Hygienists (ACGIH; 100 CFU/m3 for airborne actino­ belonged to the operations team that performed observations at the
mycetes and 1000 CFU/m3 for total airborne bacteria), and thus different wastewater processes, industrial cleaning, and maintenance,
potentially posed an increased health risk. A model based on daily and they are called wastewater (WW) workers in the results and dis­
exposure dose (US Environmental Protection Agency, 2011) has also cussion. The remaining three workers belonged to the service team,
been used to assess the risk of exposure of airborne bacteria (Li et al., which performed various maintenance and automation tasks at the
2013b), airborne endotoxin (Liu et al., 2018), and other potentially WWTP; they are called No-WW workers in the results and discussion
hazardous aerosol components (T. Yang et al., 2019) in WWTPs. The (Table s1). Of the participants, 11 were male and 3 female employees.
results of these WWTPs’ studies showed that compared with skin con­ Airborne inhalable microorganisms were sampled using personal
tact, inhalation is the main portal of entry into the human body for and stationary GSP samplers (Gesamtstaubprobenahme, CIS by BGI, INC
aerosol components. The size distribution of airborne bacteria is an Waltham, MA, USA) at a flow rate of 3.5 l/min. The samplers were
important factor in relation to inhalation and deposition of bacteria mounted with polycarbonate filters (pore size 1 μm, SUEZ – Water
(Wang et al., 2019b). However, the mentioned studies of exposure in Technologies & Solutions, Feasterville-Trevose, PA, USA) for quantifi­
WWTPs were based on data obtained from stationary rather than per­ cations of colony forming units (CFU) of bacteria or with Teflon filter
sonal sampling. Due to the differences between individuals and between (PTFE 3 μm, Millipore Sigma, Darmstadt, Germany) for sampling
work tasks, both personal exposure samples and stationary air samples airborne endotoxin. Polycarbonate filters with pore size 1 μm were used
are included in this study. as they have a high collection efficiency for particles larger than 0.2 μm
Another approach to evaluating the risk from occupational exposure (Burton et al., 2007), and in studies comparing filters the recovery rate
is to consider the presence and the concentration of the bacteria at of bacteria is highest for this filter type (Smith et al., 1993). Teflon filters
species level including the potential pathogens. Thus e.g. inhalation of were used for sampling endotoxin and 0.05% tween for extraction as a
bioaerosols with human pathogenic bacteria, such as Legionella spp., can comparative study revealed a high recovery using this method (Douwes
cause illnesses even in low concentrations (Azuma et al., 2013; Caicedo et al., 1995).
et al., 2016). It is important to identify at least to species level as the Airflows of the samplers were checked at least every 2 h. In addition,
same genus can contain species with very different pathogenicity, as one cassette with a filter was brought to the workplace each day. It was
seen for e.g. the genus Pantoea (Walterson and Stavrinides, 2015). not connected to a pump and was used as a blank filter. Each waste
Antibiotic resistant bacteria as e.g. Staphylococcus aureus have been worker carried a backpack with two pumps inside, connected to the
found in wastewater (Börjesson et al., 2009), and hence it is also rele­ samplers which were attached to the shoulder strap of the backpacks,
vant to include test of antibiotic resistance in the risk evaluation of close to the breathing zone. In addition, on every sampling day an
workers exposure. outdoor reference sample was taken using a stationary GSP sampler; this
Inflammatory responses act as a first line of defense for the immune sample was taken at the WWTP away from work activities, 1.5 m above
system, and production of reactive oxygen species (ROS) is considered a the ground. The average sampling period for the personal samples was
definitive marker of inflammation (Qu et al., 2017). The total inflam­ 283 min and for the outdoor reference 277 min. At the end of the
mation potential (TIP) assessment based on measurement of ROS pro­ workday, the pumps were switched off, and the samples were immedi­
duction by exposed human cells is another method for assessment of the ately transported to the laboratory. In total, 106 personal GSP samples
risk associated with exposure to inhalable aerosols. This method has the and 12 stationary GSP samples were taken over one year.
advantage that it measures the effect of the whole mixture of bioaerosols The stationary samples were taken in two working environments,
components on human cells, and it has previously been applied for e.g. one in the grid chamber house (called grid house), a room with coarse
waste collection workers’ exposure (Madsen et al., 2020). and fine grids, another one near the aeration tank. Stationary samples
The main objectives of this study were to obtain knowledge about were also taken with the Six-Stage Viable Andersen Cascade Impactor
WWTP workers’ 1) exposure to bacteria and endotoxin, and to apply 2) (ASCI, Thermo Fisher Scientific Inc. Waltham, MA, USA) in these envi­
different strategies for evaluation of the risk of exposure including ronments. The ASCI is an active size-selective sampler sampling directly
exposure levels combined with exposure limits and a hazard index, the onto six agar plates; particles of the following sizes were sampled on the
potential deposition of bacteria in the airways, the bacterial species following stages: Stage 1: >7 μm, stage 2: 4.7–7 μm, stage 3: 3.3–4.7 μm;
combined with Risk Group classification of bacteria, antibiotic resis­ stage 4: 2.1.-3.3 μm; stage 5: 1.1–2.1 μm, and stage 6: 0.65–1.1 μm.
tance as exemplified with a single selected species, and TIP of exposure, Stages 3 through 6 represent the respirable fraction of sampled aerosols
and about 3) potential sources of exposure. The novelty of this study is while stages 1 and 2 represent the aerosols which deposit in the upper

2
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

airways. resistance. Resistance to methicillin was established by susceptibility to

The ASCI was mounted with nutrient agar (NA; Thermo Fisher Sci­ cefoxitin in accordance with EUCAST recommendations, and confirmed
entific Oxoid, Basingstoke, UK) with actidione (cycloheximide; 50 mg/l; by presence or absence of PBP2a in a latex-agglutination test (SLIDEX®
Serva, Germany) for sampling and quantification of airborne bacteria MRSA detection BioMérieux).
and SSI agar (SSI; SSI Diagnostica, Hillerød, Denmark) for sampling of
henceforth called enteric bacteria. Sampling was performed with a flow 2.7. Risk assessment of inhalation of bacteria and endotoxin
rate of 28.3 L/min for 5 or 10 min on NA and for 30 min on SSI. The
geometric mean diameter (Dg) of the particles with culturable bacteria The inhalation risk assessment of bacteria was evaluated based on
was calculated. models developed by the EPA (US Environmental Protection Agency,
During air sampling, the air temperature and relative humidity were 2011). According to the EPA, the average daily dose (ADD) of inhaled
measured next to the outdoor reference using Tinytag Plus Data Loggers microbial aerosols can be given as:
(TGP-1500, United Kingdom; Table s2).
C × IR × EF × EDinhalation
ADDinhalation = (1)
2.3. Extraction of filters and plating BW × AT

where ADDinhalation represents the average daily dose of bacterial

Within 2 h after sampling, the bacteria collected on polycarbonate
exposure [CFU/(kg body weight⋅d)] or endotoxin exposure [EU/(kg⋅d)]
filters were extracted in 6.0 ml sterile solution (0.05% Tween 80 and
by inhalation; C is the airborne bacterial concentration (CFU/m3) or the
0.85% NaCl) by orbital shaking (500 rpm) for 15 min at room temper­
endotoxin concentration (EU/m3); IR is the inhalation rate (m3/d); EF is
ature. A serial dilution series was applied from 100 up to 10− 2 with 100
the exposure frequency (d/y); EDinhalation is the exposure duration (y);
μl of extracts were plated on NA and SSI plates for the quantification of
BW is the body weight (kg); and AT is the average lifetime (d). All the
bacteria.
parameters in Equation (1) are summarized in Table s3.
The endotoxin on the Teflon filters was extracted in 6.0 ml sterile
For non-carcinogenic risk evaluation, the daily dose was compared
solution (0.05% Tween 20) by orbital shaking (300 rpm) at room tem­
with the reference dose (RfD) for chronic exposure in order to estimate if
perature for 60 min, and subsequently the suspension was centrifuged
the contaminant presents a human health risk. Risk for non-carcinogenic
(1000×g) for 15 min to remove particulates.
pollutants was expressed by the hazard quotient (HQ) as follows:
2.4. Quantification of airborne bacteria, enteric bacteria, and endotoxin ADD
HQ = (2)
RfD
The number of bacteria on NA at 25 C and on SSI agar at 37 C were
◦ ◦
The RfD of bacteria has not yet been determined, and a limit of 500
counted after 3, 5, 7, and 14 days of incubation. The data are presented CFU/m3 was employed in this study because, according to the ACGIH,
as time-weighted average (TWA) exposures in CFU/m3 air. The dilution the exposure to culturable bacteria should not exceed 500 CFU/m3 on a
which gave optimal coverage and separation of individual CFU was used working day (Otten et al., 1986). For endotoxin exposure, 50 EU/m3 was
to represent that sample for identification. Between 1 and 88 bacterial used as the occupational exposure limit recommended by The Health
isolates from each sample were identified. In total, 2153 isolates from Council of the Netherlands (Heederik and Douwes, 1997).
personal exposure samples were identified. The hazard index (HI) represents the sum of hazard quotients for
NA agar plates and SSI plates from ASCI were incubated at 25 ◦ C and each target pollutant.
37 ◦ C, respectively, for up to 14 days. The plates were inspected every ∑
second day, and colonies were identified using matrix-assisted laser HI = HQi (3)
desorption/ionization time-of-flight mass spectrometry (MALDI-TOF
MS). When HQ ≤ 1 or HI ≤ 1, non-carcinogenic effects are not of concern
Endotoxin concentrations in samples were quantified using kinetic whereas if HQ > 1 or HI > 1, non-carcinogenic effects are of concern and
LAL test as described previously (Madsen et al., 2020). The data are the potential adverse health risks cannot be ignored. In this study, HI for
presented as EU/m3 air, and the limit of detection was 0.05 EU/ml personal exposure is the sum of the HQ of personal bacterial exposure
corresponding to 0.39 EU/m3. and the HQ of personal endotoxin exposure. HI based on stationary
samples from working environment only have bacterial concentration
2.5. Identification of bacteria by MALDI-TOF MS data.

Bacterial isolates were identified using the MALDI-TOF MS Biotyper 2.8. Total inflammatory potential
System (Bruker Daltonics, Bremen, Germany). A toothpick was used to
transfer a small amount of the bacterial colony onto the target plate Measurement of the TIP was conducted using an assay with
(MSP 96 target polished steel BC, Bruker Daltonics, Bremen, Germany). granulocyte-like cells originating from HL-60 cells. Upon exposure to
The sample was then overlaid with 1 μL 70% formic acid and allowed to inflammogens, the differentiated HL-60 cells produce reactive oxygen
dry before adding an HCCA matrix solution (α-cyano-4-hydroxycin­ species (ROS), which are quantifiable by a luminol-dependent chem­
namic acid, Bruker Daltonics). The MALDI-TOF MS analysis was per­ iluminometric assay. The cells were differentiated by adding all-trans
formed on a Microflex LT mass spectrometer (Bruker Daltonics) using retinoic acid and the growth medium (RPMI 1640, Biological In­
the Bruker Biotyper 3.1 software with the BDAL standard library. A dustries, USA) was not changed for 6–7 days. The cells were seeded at 3
bacterial test standard (Bruker Daltonics) was used to calibrate the in­ × 105 cells/ml and incubated at 5% CO2 at 37 ◦ C.
strument. Bacterial isolates that failed to give a positive identification To analyze the bioaerosol samples, amounts of 50 μl granulocyte
were subsequently extracted using the ethanol extraction method as cells and 100 μl of the twenty-fold diluted personal sample suspension
described elsewhere (Madsen et al., 2015). Bacterial pathogens were and the outdoor reference samples were inoculated in duplicate into
classified according to the European classification standard (European each well in a microtiter plate together with 30 μl Hanks’ balanced salt
Parliament, 2006). solution, 10 μl plasma, and 10 μl luminol. The chemiluminescence re­
action caused by sample activity was measured by a thermostated
2.6. Susceptibility testing of S. aureus (37 ◦ C) ORION II Microplate Luminometer (Berthold Detection Systems,
Germany), which measured relative-light units per second (RLU/s) for 1
Six S. aureus colonies were isolated and tested for antibiotic s every 120 s for 180 min. For every sample, accumulated RLU/s were

3
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

calculated by summing the RLU/s measurements throughout the 180

min period. To account for the reactivity level of the cells a blank (with
0.001% Tween) was included in every run. An outdoor air reference
from a residential area was taken in June 2019, and the microorganisms
were extracted as the other samples in this study, and cultivation on
DG18 and NA showed that it contained 45 CFU Cladosporium herbarum/
ml and 10 CFU Staphylococcus warneri/ml, and a subsamples of this
sample was included in every run.
The TIP is expressed as area under the curve (AUC), and the TIP of
each GSP sample was normalized by the TIP of blank using the following
formula:
TIPG × TIPavg(blank)
TIPNG =
TIPblank

where TIPNG represents the normalized TIP of GSP sample, TIPG and
TIPblank represent the TIP of the GSP sample and of the blank from the
same run, respectively, TIPavg(blank) is the average TIP of all blank values
in this study. The TIP data are presented as AUC/m3.
Fig. 1. Exposure to airborne bacteria, enteric bacteria, and endotoxin for
people working near (WW) and not working near (No-WW) wastewater as part
2.9. Statistical analysis of their workday during the four seasons.

The relative abundance (%) of bacterial species is the ratio of the

detected in 30 out of 106 personal samples, and thus present in 28.3% of
number of isolates of each species to the total number of isolates in each
the personal samples. The personal exposure was between 5 and 161
type of sample. The data analyses were performed in IBM SPSS Statistics
CFU/m3 (GM = 19 CFU/m3). No significant effect of season (p = 0.72),
25 for Windows and SAS 9.4. TIP and exposure values were log trans­
person (p = 0.94), and work task (p = 0.27) was found on the level of
formed to obtain normal distribution before analysis. The TIP and
exposure to enteric bacteria.
exposure values were examined for significant differences between the
The range of the endotoxin exposure for personal samples was from
seasons using a least significant difference one-way analysis of variance.
below the detection limit (0.39 EU/m3) to 707.3 EU/m3 (GM = 9.2 EU/
An independent sample t-test analysis was used to compare the differ­
m3). Significant effects of season (p = 0.029), work tasks (p = 0.0001),
ence, between two work tasks, between personal exposure and reference
and person (p = 0.0002) were found on the endotoxin exposure. The
samples, and between the TIP of personal exposure and outdoor refer­
highest endotoxin exposures were found in the winter (GM = 14.6 EU/
ence samples. Pearson’s correlation between the bacterial concentra­
m3), and it was significantly higher than in the summer (GM = 5.9 EU/
tion, endotoxin concentration and TIP were calculated. A p-value below
m3; p = 0.038). Exposure levels to endotoxin for WW workers (GM =
0.05 was considered statistically significant.
12.1 EU/m3) were higher than for No-WW workers (GM = 3.1 EU/m3; p
Statistical analyses for microbial community composition and
= 0.001). A WW worker, was exposed to the highest level of endotoxin
structure for the bacteria were performed and visualized using RStudio
(GM = 36.6 EU/m3, n = 9; Fig. s1). In total, 15 exposures were above 50
version 3.5.3 with the following R CRAN packages: ggplot 2, ampvis2,
EU/m3; they were sampled from 8 different workers on 8 different days.
ggpubr, and vegan. Canonical correspondence analysis (CCA) plots
A significant correlation was found between exposure to bacteria as
based on weighted Bray-Curtis distance measurements were used to
measured on NA and exposure to endotoxin (r = 0.24, p = 0.017, n =
explore the β-diversity. Comparisons of the microbial communities were
106).
done using analysis of similarity (ANOSIM) based on Bray-Curtis index
values for the different seasons and work tasks. Source tracker was used
to study the specific contribution of airborne bacteria in the grid house 3.2. Bacterial concentration and size distribution in two working
and aeration tank areas to bacteria in the personal samples. environments of the wastewater treatment plant (ASCI)

3. Results The concentration of airborne bacteria in the grid house was between
180 and 8085 CFU/m3 (GM = 1116 CFU/m3). Winter (GM = 2043 CFU/
3.1. Exposure levels to bacteria and endotoxin (GSP) m3) had the highest bacteria concentration and spring the lowest (GM =
219 CFU/m3). However, overall no significant of season was found (p =
The personal exposure to bacteria was between 11–5.89 × 103 CFU/ 0.059). In the aeration tank area, the airborne bacteria concentration
m3, with a geometric mean value (GM) of 299 CFU/m3, and it was was between 107 and 1838 CFU/m3 (GM = 368 CFU/m3). A significant
significantly higher (p < 0.001) than the bacterial concentration of effect of season (p = 0.015) was found. The bacterial concentration was
outdoor reference samples (GM = 15 CFU/m3). A significant effect of higher in summer than winter (p = 0.013) and spring (p = 0.003), and
season (p = 0.0236), work task (p = 0.0087), and person (p = 0.0002) higher in autumn than in spring (p = 0.048). In winter, the concentra­
was found on the exposure to bacteria as measured on NA. The workers tion of airborne bacteria in the grid house was significantly higher than
were exposed to the highest bacterial concentration in the autumn (GM in the aeration tank area (p = 0.036). No significant difference in bac­
= 499 CFU/m3) and it was significantly higher than in other seasons (p terial concentration was found between the grid house and the aeration
< 0.05). The lowest exposure was measured in the winter (GM = 223 tank area during other seasons (p > 0.05). The concentration of airborne
CFU/m3). WW workers were exposed to higher levels of bacteria (GM = bacteria in the two working environments (GM = 546 CFU/m3) tended
352 CFU/m3) than No-WW workers (GM = 158 CFU/m3; p = 0.01; to be higher than the personal exposures (GM = 299 CFU/m3, p =
Fig. 1). The person with the highest exposure was exposed to 2.52 × 103 0.055). The average respirable fraction of bacteria was 89% in the grid
CFU/m3 (GM; WW worker; Fig. s1). In total, 36 exposures were above house and 70% in the aeration tank area (Fig. 2).
500 CFU/m3; these exposures were sampled from 13 different workers Enteric bacteria were found in 5 of 12 samples from the grid house in
on 12 different days. concentrations from 5 to 18 CFU/m3 (GM = 9.4 CFU/m3). Eight out of
Enteric bacteria, here defined as bacteria growing on SSI agar, were 12 samples from the aeration tank area contained enteric bacteria, and

4
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

Fig. 2. Concentration and size distribution of airborne bacteria in the grid house and aeration tank area during the four seasons as measured on NA-agar. Dg = the
geometric mean diameter.

the concentration was between 2 and 43 CFU/m3 (GM = 5.3 CFU/m3; no 14.4% and 23.3%, and summer: 9.4% and 23.6%). In winter, Micro­
further data shown). coccus (14.6%), Staphylococcus (12.7%), and Cellulosimicrobium (10.6%)
were found to be the dominant genera. Thirty out of 131 species from
3.3. The hazard index of bacteria and endotoxin personal samples were gram-negative bacteria and the highest relative
abundance was found in spring (10.1%; Table s4).
The hazard index (HI) of inhalable airborne bacteria for both adult In total, 45 genera and 91 species of different bacteria were identi­
genders was less than 1 (GMmale = 0.046, GMfemale = 0.044; no further fied from the samples taken in the grid house; 31 out of these species
data shown). The HI of the combined exposure to airborne bacteria and were gram-negative, with a relative abundance of 8.7%. From the
endotoxin was also less than 1 (GMmale = 0.074, GMfemale = 0.071; aeration tank area, 39 genera and 94 species of bacteria were identified;
Table 1) and was significant affected by the season (p = 0.009) with the 42 of these species were gram-negative with a relative abundance of
highest HI in autumn. 10.9% (Table s5). No significant effect of season was found on the
The HI of inhalable airborne bacteria in the grid house was signifi­ bacterial composition in the grid house (p = 0.61, Fig. 3B) or in the
cantly higher than that of the aeration tank area and than that of the aeration tank area (p = 0.17, Fig. 3C). Significant differences in both
personal samples (p < 0.05), and no significant difference was found bacterial composition (p = 0.001, Fig. 3D) and gram-negative bacterial
between the aeration tank area and personal samples (p = 0.863). For
samples from the aeration tank area, the HI of airborne bacteria in
summer was higher than that of other seasons (p < 0.05).

3.4. Airborne bacterial species on SSI and NA agar plates

In total, 41 different genera and 131 different species of bacteria

were identified from personal samples. A seasonality was found in
species composition (p = 0.014), but no significant difference in species
composition was found between WW and No-WW workers (p = 0.17;
Fig. 3A). During all seasons Micrococcus and Staphylococcus constituted
around 10% of all identified isolates (autumn: 24.5% and 17.7%, spring:

Table 1
The hazard index (HI) of inhalable airborne bacteria and endotoxin from per­
sonal samples, and HI of inhalable airborne bacteria from stationary samples in
the grid house and the aeration tank area.
Work site Personal samples Grid house Aeration tank

Season Male Female Male Female Male Female Fig. 3. Canonical correspondence analysis (CCA) of airborne bacteria during
Autumn 0.107 0.102 0.186 0.177 0.077 0.073 the four seasons; (A) personal exposure samples constrained by season; (B) grid
Spring 0.067 0.064 0.041 0.038 0.024 0.022 house samples constrained by season; (C) aeration tank area samples con­
Summer 0.060 0.050 0.179 0.169 0.235 0.223
strained by season, and (D) all samples constrained by sample type (personal,
Winter 0.080 0.070 0.317 0.300 0.044 0.041
grid house, aeration tank area, and outdoor references).

5
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

composition (p = 0.001, Fig. s2) were found between the samples from
the grid house and the aeration tank area and the personal exposure
samples.
Three species, Enterobacter cloacae, Staphylococcus aureus, and Yer­
sinia enterocolitica, which belong to Risk Group 2 were found in the
personal exposure samples. Further antibiotic susceptibility analysis
found that the six S. aureus isolates were methicillin-susceptible. The
S. aureus isolates were from the sampling day in April and were from a
WW and a No-WW worker.
Enterobacter cloacae was found only in the autumn and was the only
pathogen found both in personal exposure and working environment
samples. Risk Group 2 bacteria: Escherichia coli, Klebsiella oxytoca,
Klebsiella pneumoniae, and Morganella morganii were found both in the
grid house and in the aeration tank, but not in any personal samples.
Pseudomonas aeruginosa was only found in the aeration tank area, while
Neisseria meningitidis and Providencia rettgeri were only found in the grid
house (Fig. 4).

3.5. Total inflammatory potential

There was no significant difference (p = 0.709) between the TIP of

personal samples (GM = 4.59 × 107 AUC/m3) and WWTP stationary
reference samples (GM = 4.92 × 107 AUC/m3), but the TIP of exposure
was significantly higher than the TIP of the residential outdoor reference
sample (GM = 1.42 × 107 AUC/m3, p = 0.001). Significantly different
TIPs of exposure were found in the different seasons (p < 0.000, Fig. 5A)
and for the different persons (p = 0.037, Fig. 5B), while no significant
difference was found between the WW and No-WW workers (p = 0.076,
Fig. s3). Exposures measured in autumn were higher (7.45 × 107 AUC/
m3) than in spring (4.33 × 107 AUC/m3, p = 0.02) and summer (2.78 ×
107 AUC/m3, p < 0.0001). The TIP in winter (5.54 × 107 AUC/m3) was
higher than in summer (p = 0.003; Fig. 5A).
A significant correlation was found between exposure to bacteria and
TIP of exposure (r = 0.548, p < 0.0001), while endotoxin did not
correlate significantly with TIP (r = 0.108, p = 0.288).
The TIP of personal exposure was divided into quartile values (Q1 =

Fig. 5. Variations between seasons (A) and persons (B) in the total inflamma­
tory potential (TIP, AUC/m3) of personal exposures. Box frames represent the
upper and lower quartiles, the horizontal line represents the median, whiskers
denote range, the black dots represent outliers.

2.57 × 107, Q2 = 4.85 × 107, Q3 = 8.10 × 107 AUC/m3), and the

bacteria composition was constrained by these quartiles (p = 0.19;
Fig. 6). Sixteen gram-negative species were found in the highest TIP
quartile (TIP > 8.10 × 107 AUC/m3), and the number of gram-negative
species tended to be higher than that in the other quartiles (14, 9, 8
species in TIP quartile decreasing order).

4. Discussion

In this study, we apply different approaches to evaluating the risks

associated with exposure during work on a WWTP; these approaches
included: assessing workers’ bacterial and endotoxin exposure levels
and comparing them to suggested exposure limits, a hazard index, the
potential deposition of bacteria in the airways, specific bacteria species
in workers’ exposure, and the inflammatory potential of their exposure.
A cohort of WWTP 14 workers were followed throughout a year.

Fig. 4. Airborne bacterial species which belong to Risk Group 2, found in the
personal exposure samples, grid house, and the aeration tank area. Overlap
indicates species identified in the different sample types.

6
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

Yang et al., 2019). However, in the present study we have investigated

personal exposure and have sampled repeatedly on whole workdays
while the study by Yang used stationary samplers. In this study, the HI
was higher in the grid house than the HI of the workers’ exposure. This is
probably because the workers only performed work tasks in areas with
exposure from the wastewater for short periods. The HI of the concen­
tration in the aeration tank area was not different from the HI of the
workers. This highlights the importance of personal sampling in the risk
assessment.

4.3. Risk assessment based on potential deposition in the airways and

swallowing

In this study, airborne bacteria were found in all studied size frac­
tions, and thus they may be deposited in different areas of the airways.
The Dg of particles with airborne bacteria were between 2.0 μm and 4.3
μm. A fraction of 89% of the bacteria in the grid house and 70% of the
bacteria in the aeration tank area were found in the respirable fraction
(aerodynamic diameter < 4.7 μm). This indicates that most bacterial
Fig. 6. Canonical correspondence analysis (CCA) of airborne bacteria con­
aerosols in the working environments could potentially deposit in the
strained by the 4 quartiles of TIP levels (Q1 = 2.57 × 107, Q2 = 4.85 × 107, Q3
lower respiratory tract in humans. Infectious bacteria, such as Klebsiella
= 8.10 × 107 AUC/m3).
pneumoniae, may if deposited in respiratory system, cause acute in­
fections in susceptible hosts (Egbe et al., 2011). The airborne bacterial
4.1. Exposure levels related to previous findings
particle size in the grid house (Dg = 2.2 μm) was smaller than that at the
aeration tank (Dg = 3.5 μm) indicating working in the grid house poses a
The personal exposure to bacteria varied between seasons with
higher risk of effects on the deeper airways.
similar results reported in Switzerland (Oppliger et al., 2005). Compared
Some of the airborne bacteria are species known to cause gastroin­
to personal exposure from May 2010 (GM = 837 CFU/m3; Uhrbrand
testinal problems e.g. E. coli, Bacillus cereus, Yersinia enterocolitica, and
et al., 2011), the exposure in this study in the same month was relatively
S. aureus. Gastrointestinal problems have been reported repeatedly
low (GM = 299 CFU/m3). The airborne bacteria concentration in the
among WWTP workers e.g. (Scarlett-Kranz et al., 1987), and because a
grid house was significantly higher than that from aeration tank area.
part of the larger sized, inhaled particles will be swallowed it is relevant
Most airborne bacteria in the WWTP were gram-positive bacteria,
also to consider this. In this study, airborne E. coli and B. cereus were
similar results were found in Switzerland with less than 10% being
found on particles with aerodynamic diameters of 7–12 μm, and
gram-negative (Oppliger et al., 2005). The personal exposure to endo­
consequently they may be swallowed.
toxin (GM = 7.9 EU/m3) was slightly higher than that found in different
Since the risk assessment results based on OEL and HI as well as the
work stations in a Danish hospital WWTP (GM, 1.3–6.5 EU/m3; Uhr­
species composition were different for the personal exposure and
brand et al., 2017), but lower than the workers’ exposure in a waste­
working environment samples, a risk assessment based on particle size
water sludge thermal drying room in France (GM = 53 EU/m3; Schlosser
distribution from samples from the personal inhalation zone still needs
et al., 2011) and in WWTPs in lowa city, USA (GM = 91 EU/m3; Lee
to be explored.
et al., 2007).
In this study, waste workers who worked for part of a workday in
4.4. Risk assessment based on bacterial species
areas near wastewater were exposed to higher levels of bacteria and
endotoxin than workers without these tasks. A WWTP study from
Ten different bacterial species found in personal samples and sta­
Switzerland showed that work with cleaning of tanks caused a higher
tionary samples belonged to Risk Group 2. All of them, except S. aureus,
exposure risk to endotoxin than routine work (Oppliger et al., 2005).
were gram-negative. Yersinia enterocolitica, is a well known agent of
Workers handling dry sludge had more airway and systemic symptoms
foodborne, zoonotic bacterial gastroenteritis (Helms et al., 2003).
than office workers in a WWTP study from Norway (Heldal et al., 2010).
Enterobacter cloacae, was the only pathogen found both in personal and
Hence, WW workers seem to have higher occupational exposure risk
stationary samples in the same season, and it can cause a variety of in­
than No-WW workers.
fections, such as urinary tract infections and intraabdominal infections
(Jones et al., 1999). All E. cloacae (Davin-Regli and Pagès, 2015) and
4.2. Exposure levels related to suggested occupational exposure limits some Y. enterocolitica (Baumgartner et al., 2007) are also reported to be
(OELs) and hazard index (HI) intrinsically resistant to some antibiotics, like ampicillin and amox­
icillin/clavulanic acid.
The OEL for bacteria of 500 CFU/m3 suggested by the ACGIH (Otten Escherichia coli (Levine, 1987), Klebsiella oxytoca (Kashiwagi et al.,
et al., 1986), was exceeded in 34% personal exposure samples. There­ 2007), and K. pneumoniae (Song et al., 2008) are also classified as Risk
fore, the exposure may cause a potential health risk. Personal exposure Group 2 bacteria and they can be found in the gastrointestinal tract.
to endotoxin exceeded the OEL of 50 EU/m3 for 14% measurements. In a These bacteria, and a rare human pathogen, Morganella morganii
previous study, this exposure level was associated with elevated risk of (Abdalla et al., 2006) were found both in the air in the grid house and in
diarrhea (Smit et al., 2005). Five workers were exposed to more than the air in the aeration tank area. Aeromonas veronii, which may cause
200 EU/m3, and this exposure level has previously been associated with gastroenteritis (Schuetz, 2019), was found in the aeration tank area, but
flu-like and upper respiratory symptoms in WWTP workers (Smit et al., not in the personal samples. This highlights the importance of limiting
2005). occupancy in these areas.
The risk assessment based on the personal exposure of inhalable According to the WHO, antibiotic resistance is one of the biggest
airborne bacteria and endotoxin by ADDinhalation showed a hazard index threats to global health (Shrivastava et al., 2018), therefore it is also
(HI) smaller than 1, thus the risk associated with inhalation can be relevant to include this in the risk evaluation of WWTP workers’ expo­
negligible. These HI results are in accordance with a previous study (K. sure. As an example, we tested the six S. aureus isolates for antibiotic

7
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

resistance, and found that they were all methicillin-susceptible. Even 26% of the airborne intestinal bacteria seemed to be from the BRT
though they are methicillin-susceptible they are not necessarily less (Wang et al., 2019a). Some pathogenic bacterial species (5.7%–36.4%)
virulent than resistant versions (MRSA; Rozgonyi et al., 2007), but the in the air at a WWTP seemed also to originate from wastewater (K. Yang
treatment is faster. et al., 2019). In this study, sourcetracker analysis showed that 22.36% of
Overall the species identification showed that the workers were the bacteria potentially inhaled by WWTP workers’ seem to be from the
exposed to airborne bacteria which can potentially cause gastrointes­ air around the aeration tank and 22.40% from the grid house (Fig. s4).
tinal problems and are classified in Risk Group 2. The risk classification Unidentified sources could be the sludge dewatering house (Han et al.,
specifically deals with bacterial infections and the severity of these, and 2018; K. Yang et al., 2019), the fine screen room or from the ambient air
thus the bacteria causing the airway symptoms often reported by WWTP (Xu et al., 2018). Some differences were observed in the composition of
workers are not risk classified on this basis, and we cannot yet evaluate gram-negative bacterial species between personal exposure and sta­
the species in regard to this. tionary samples. Hence, further source analysis for bacteria, especially
for pathogenic bacteria, in WWTPs is important to be able to implement
4.5. Risk assessment based on inflammatory potential the right preventive measures.

The total inflammation potential (TIP) of the WWTP workers expo­ 5. Conclusion
sure was found to be positively correlated with bacterial exposure, and a
similar relationship was also found in a study on waste collection Based on the different measures to evaluate the risk of exposure via
workers’ exposure (Madsen et al., 2020). The TIP of exposure in this the air, it is concluded that the GM exposure to endotoxin (9.2 EU/m3)
study (GM = 4.59 × 107 AUC/m3) seems higher than that in a study of and bacteria (299 CFU/m3) are low and below suggested OELs. This is
waste collection workers (max = 3.8 × 107 AUC/m3; Madsen et al., also confirmed by the HI, and by the TIP which is at the same level as a
2020), even though the bacterial exposure of WWTP workers is lower reference measurement at the WWTP. Thus, an accordance between
than that of waste collection workers. This might be due to different these different measures of risk associated with exposure via air is found.
bacterial compositions. In addition, non-culturable bacteria could also However, a continued focus on preventive measures is important, as this
have contributed to the TIP of exposure. study found that 1) 15 of 106 personal measurements of exposure to
When the bacterial composition was constrained by the four quar­ endotoxin and 36 of 106 measurements of exposures to bacteria exceed
tiles of TIP levels the difference was not significant (p = 0.17), but more suggested OELs, 2) some of these bacteria are present on particles of a
gram-negative species were found in the highest TIP level quartile. size which may enter the lower respiratory tract and some are of a size
Furthermore, autumn was associated with the highest TIP of exposure, which may be swallowed, 3) the TIP of exposure was elevated compared
and the highest exposure to bacteria, which was also reflected in a to a reference from a residential area, and TIP was associated with
higher HI of these exposures, and in addition, the bacterial composition bacterial exposure level, and finally 4) the WWTP workers’ were
was affected by season. Thus, some accordance is found between the exposed to bacterial species known to cause gastrointestinal problems
different methods used for risk evaluation. and these bacteria were present in the fraction which may be swallowed.
In this study we measured the ROS production during 3 h of exposure The identification of microorganisms to species level was useful
of human granulocytes to the workers personal exposure. It has the especially in relation to risks of gastrointestinal problems, while TIP and
strength that it is a human cell line, it is ‘real’ exposures from different endotoxin are relevant mainly in relation to risks of symptoms of the
tasks and seasons. It should however be noted that long term exposure to airways, while the knowledge of size distribution of airborne bacterial
airborne inflammogens may have an inflammatory response that is not species is relevant in relation to both gastrointestinal and airway prob­
significant in an in vitro study. A study with mice showed the dose of lems. Together this shows that different approaches are important in the
endotoxin causing tissue or cell damage in mice’ lung was 5 times lower evaluation of occupational exposure risks.
than the dose causing acute inflammation (Xue et al., 2018). Therefore, Measurement of the microbial composition made the source tracking
the long-term sub-chronic exposure might cause health problems for the possible, and showed that 45% of the bacterial species in the workers’
WWTP workers later in life, and the in vivo study could be suggested to exposure seems to be from the air in the grid house or the aeration tank
use for the a risk assessment study in the future. area. On the other hand constrained redundancy analysis showed a
significantly different species composition in personal and working
4.6. Seasonal variation and source analysis of airborne bacteria environment samples, and this highlights the importance of analyzing
personal samples in the evaluation of risk of exposure.
Seasonal variation of bacterial species were found both in personal
exposure and in working environment samples. We have found no other Credit author statement
studies with personal exposure assessment through a year combined
with species identification. However, a study with stationary measure­ Rui Lu: Data curation, Investigation, Visualization，Writing - orig­
ments also found seasonality in the concentrations of coliforms and inal draft.Margit W. Frederiksen ：Investigation,Katrine Uhrbrand ：
staphylococci (Szyłak-Szydłowski et al., 2016). In this study, the per­ Investigation,Yanpeng Li ：Writing - Review & Editing,Claus
sonal samples were dominated by normal skin bacteria as Micrococcus Østergaard：; Investigation,Anne Mette Madsen ：;Conceptualization，
and Staphylococcus. In the aeration tank area, Staphylococcus (13.2%) Supervision， Project administration，Funding acquisition， Formal
and Aeromonas (11.7%) were the dominant genera, while Acinetobacter analysis，Writing - Review & Editing.
(25.6%) was the dominant genus in grid house. Therefore, the Micro­
coccus and Staphylococcus may partly derive from the workers them­ Declaration of competing interest
selves or from their indoor work activities. Micrococcus and
Staphylococcus are common in indoor air (Madsen et al., 2018). Nine The authors declare that they have no known competing financial
species of Aeromonas were found in the winter and autumn, and some of interests or personal relationships that could have appeared to influence
them are known to cause gastroenteritis (Pal et al., 2016). Eleven species the work reported in this paper.
of Acinetobacter were found with the highest richness in the autumn.
In addition to the identification of potentially inhaled bacteria spe­ Acknowledgments
cies, it is also necessary for risk assessment to investigate their sources. A
previous study found a biochemical reaction tank (BRT) as a main source The study was supported by The Danish Research Council (DFF –
in a WWTP (Wang et al., 2018), and upon further study they found that 1335–00183) and the Danish Working Authority Foundation (32-2019-

8
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

09 20195100806). We also want to thank the workers at the WWTP who Levine, M.M., 1987. Escherichia coli that cause diarrhea: enterotoxigenic,
enteropathogenic, enteroinvasive, enterohemorrhagic, and enteroadherent. J. Infect.
participated in the study.
Dis. 155, 377–389. https://doi.org/10.1093/infdis/155.3.377.
Li, Y., Yang, L., Meng, Q., Qiu, X., Feng, Y., 2013a. Emission characteristics of microbial
Appendix A. Supplementary data aerosols in a municipal sewage treatment plant in Xi’an, China. Aerosol Air Qual.
Res. 13, 343–349. https://doi.org/10.4209/aaqr.2012.05.0123.
Li, Y., Zhang, H., Qiu, X., Zhang, Y., Wang, H., 2013b. Dispersion and risk assessment of
Supplementary data to this article can be found online at https://doi. bacterial aerosols emitted from rotating-brush aerator during summer in a
org/10.1016/j.ecoenv.2020.111365. wastewater treatment plant of xi’an, China. Aerosol Air Qual. Res. 13, 1807–1814.
https://doi.org/10.4209/aaqr.2012.09.0245.
Liu, H., Zhang, Z., Wen, N., Wang, C., 2018. Determination and risk assessment of
References airborne endotoxin concentrations in a university campus. J. Aerosol Sci. 115,
146–157. https://doi.org/10.1016/j.jaerosci.2017.09.002.
Abdalla, J., Saad, M., Samnani, I., Lee, P., Moorman, J., 2006. Central nervous system Madsen, A.M., Frederiksen, M.W., Bjerregaard, M., Tendal, K., 2020. Measures to reduce
infection caused by Morganella morganii. Am. J. Med. Sci. 331, 44–47. https://doi. the exposure of waste collection workers to handborne and airborne microorganisms
org/10.1097/00000441-200601000-00013. and inflammogenic dust. Waste Manag. 101, 241–249. https://doi.org/10.1016/j.
Azuma, K., Uchiyama, I., Okumura, J., 2013. Assessing the risk of Legionnaires’ disease: wasman.2019.10.023.
the inhalation exposure model and the estimated risk in residential bathrooms. Madsen, A.M., Moslehi-Jenabian, S., Islam, M.Z., Frankel, M., Spilak, M., Frederiksen, M.
Regul. Toxicol. Pharmacol. 65, 1–6. https://doi.org/10.1016/j.yrtph.2012.11.003. W., 2018. Concentrations of Staphylococcus species in indoor air as associated with
Baumgartner, A., Küffer, M., Suter, D., Jemmi, T., Rohner, P., 2007. Antimicrobial other bacteria, season, relative humidity, air change rate, and S. aureus-positive
resistance of Yersinia enterocolitica strains from human patients, pigs and retail pork occupants. Environ. Res. 160, 282–291. https://doi.org/10.1016/j.
in Switzerland. Int. J. Food Microbiol. 115, 110–114. https://doi.org/10.1016/j. envres.2017.10.001.
ijfoodmicro.2006.10.008. Madsen, A.M., Zervas, A., Tendal, K., Nielsen, J.L., 2015. Microbial diversity in
Börjesson, S., Melin, S., Matussek, A., Lindgren, P.E., 2009. A seasonal study of the mecA bioaerosol samples causing ODTS compared to reference bioaerosol samples as
gene and Staphylococcus aureus including methicillin-resistant S. aureus in a measured using Illumina sequencing and MALDI-TOF. Environ. Res. 140, 255–267.
municipal wastewater treatment plant. Water Res. 43, 925–932. https://doi.org/ https://doi.org/10.1016/j.envres.2015.03.027.
10.1016/j.watres.2008.11.036. Oppliger, A., Hilfiker, S., Duc, T.V., 2005. Influence of seasons and sampling strategy on
Bruni, E., Simonetti, G., Bovone, B., Casagrande, C., Castellani, F., Riccardi, C., assessment of bioaerosols in sewage treatment plants in Switzerland. Ann. Occup.
Pomata, D., Di Filippo, P., Federici, E., Buiarelli, F., Uccelletti, D., 2020. Evaluation Hyg. 49, 393–400. https://doi.org/10.1093/annhyg/meh108.
of bioaerosol bacterial components of a wastewater treatment plant through an Otten, J., Feeley, J., Burge, H., 1986. ACGIH committee activities and reports. Appl. Ind.
integrate approach and in vivo assessment. Int. J. Environ. Res. Publ. Health 17, 273. Hyg. 1 https://doi.org/10.1080/08828032.1986.10390461. R-19-R-23.
https://doi.org/10.3390/ijerph17010273. Pal, B.B., Pattnaik, S.K., Mohanty, A., Samal, S.K., Khuntia, H.K., Nayak, S.K., 2016.
Burton, N.C., Grinshpun, S.A., Reponen, T., 2007. Physical collection efficiency of filter Incidence of Aeromonas species isolated from Diarrhoea patients and water samples
materials for bacteria and viruses. Ann. Occup. Hyg. 51, 143–151. https://doi.org/ from coastal districts of Odisha, India. Int. J. Curr. Microbiol. Appl. Sci. 5, 990–999.
10.1093/annhyg/mel073. https://doi.org/10.20546/ijcmas.2016.507.111.
Caicedo, C., Beutel, S., Scheper, T., Rosenwinkel, K.H., Nogueira, R., 2016. Occurrence of Qu, J., Li, Y., Zhong, W., Gao, P., Hu, C., 2017. Recent developments in the role of
Legionella in wastewater treatment plants linked to wastewater characteristics. reactive oxygen species in allergic asthma. J. Thorac. Dis. https://doi.org/10.21037/
Environ. Sci. Pollut. Res. 23, 16873–16881. https://doi.org/10.1007/s11356-016- jtd.2017.01.05.
7090-6. Rozgonyi, F., Kocsis, E., Kristóf, K., Nagy, K., 2007. Is MRSA more virulent than MSSA?
Cyprowski, M., Sobala, W., Buczyńska, A., Szadkowska-Stańczyk, I., 2015. Endotoxin Clin. Microbiol. Infect. https://doi.org/10.1111/j.1469-0691.2007.01780.x.
exposure and changes in short-term pulmonary function among sewage workers. Int. Sánchez-Monedero, M.A., Aguilar, M.I., Fenoll, R., Roig, A., 2008. Effect of the aeration
J. Occup. Med. Environ. Health 28, 803–811. https://doi.org/10.13075/ system on the levels of airborne microorganisms generated at wastewater treatment
ijomeh.1896.00460. plants. Water Res. 42, 3739–3744. https://doi.org/10.1016/j.watres.2008.06.028.
Davin-Regli, A., Pagès, J.M., 2015. Enterobacter aerogenes and Enterobacter cloacae; Scarlett-Kranz, J.M., Babish, J.G., Strickland, D., Lisk, D.J., 1987. Health among
Versatile bacterial pathogens confronting antibiotic treatment. Front. Microbiol. municipal sewage and water treatment workers. Toxicol. Ind. Health 3, 311–319.
https://doi.org/10.3389/fmicb.2015.00392. https://doi.org/10.1177/074823378700300303.
Douwes, J., Versloot, P., Hollander, A., Heederik, D., Doekes, G., 1995. Influence of Schlosser, O., Huyard, A., Catalán, V., 2011. Exposure to dust, endotoxin and airborne
various dust sampling and extraction methods on the measurement of airborne microorganisms in wastewater sludge thermal drying units. Water Sci. Technol. 64,
endotoxin. Appl. Environ. Microbiol. 61, 1763–1769. https://doi.org/10.1128/ 1073–1080. https://doi.org/10.2166/wst.2011.514.
aem.61.5.1763-1769.1995. Schuetz, A.N., 2019. Emerging agents of gastroenteritis: Aeromonas, Plesiomonas, and
Egbe, C.A., Ndiokwere, C., Omoregie, R., 2011. Microbiology of lower respiratory tract the diarrheagenic pathotypes of Escherichia coli. Semin. Diagn. Pathol. 36, 187–192.
infections in Benin city, Nigeria. Malays. J. Med. Sci. 18, 27–31. https://doi.org/10.1053/j.semdp.2019.04.012.
European Parliament, 2006. Directive 2000/54/EC of the European Parliament and of Shrivastava, S.R., Shrivastava, P.S., Ramasamy, J., 2018. Responding to the challenge of
the Council of 18 September 2000 on the protection of workers from risks related to antibiotic resistance: world health organization. J. Res. Med. Sci. 23, 3–4. https://
exposure to biological agents at work. Off. J. Eur. Communities 262, 21–45. doi.org/10.4103/1735-1995.228593.
Fracchia, L., Pietronave, S., Rinaldi, M., Giovanna Martinotti, M., 2006. Site-related Smit, L.A.M., Spaan, S., Heederik, D., 2005. Endotoxin exposure and symptoms in
airborne biological hazard and seasonal variations in two wastewater treatment wastewater treatment workers. Am. J. Ind. Med. 48, 30–39. https://doi.org/
plants. Water Res. 40, 1985–1994. https://doi.org/10.1016/j.watres.2006.03.016. 10.1002/ajim.20176.
Han, Y., Wang, Y., Li, L., Xu, G., Liu, J., Yang, K., 2018. Bacterial population and Smith, L., Carroll, K., Mottice, S., 1993. Comparison of membrane filters for recovery of
chemicals in bioaerosols from indoor environment: sludge dewatering houses in nine legionellae from water samples. Appl. Environ. Microbiol. 59, 344–346. https://doi.
municipal wastewater treatment plants. Sci. Total Environ. 618, 469–478. https:// org/10.1128/aem.59.1.344-346.1993.
doi.org/10.1016/j.scitotenv.2017.11.071. Song, H.J., Shim, K.N., Jung, S.A., Choi, H.J., Lee, M.A., Ryu, K.H., Kim, S.E., Yoo, K.,
Heederik, D., Douwes, J., 1997. Towards an occupational exposure limit for endotoxins ? 2008. Antibiotic-associated diarrhea: candidate organisms other than Clostridium
Ann. Agric. Environ. Med. difficile. Korean J. Intern. Med. 23, 9–15. https://doi.org/10.3904/
Heldal, K.K., Madsø, L., Huser, P.O., Eduard, W., 2010. Exposure, symptoms and airway kjim.2008.23.1.9.
inflammation among sewage workers. Ann. Agric. Environ. Med. 17, 263–268. Szyłak-Szydłowski, M., Kulig, A., Miaśkiewicz-Pęska, E., 2016. Seasonal changes in the
Helms, M., Vastrup, P., Gerner-Smidt, P., Mølbak, K., 2003. Short and long term concentrations of airborne bacteria emitted from a large wastewater treatment plant.
mortality associated with foodborne bacterial gastrointestinal infections: registry Int. Biodeterior. Biodegrad. 115, 11–16. https://doi.org/10.1016/j.
based study. Br. Med. J. 326, 357. ibiod.2016.07.008.
Jeggli, S., Steiner, D., Joller, H., Tschopp, A., Steffen, R., Hotz, P., 2004. Hepatitis E, Uhrbrand, K., Schultz, A.C., Koivisto, A.J., Nielsen, U., Madsen, A.M., 2017. Assessment
Helicobacter pylori, and gastrointestinal symptoms in workers exposed to waste of airborne bacteria and noroviruses in air emission from a new highly-advanced
water. Occup. Environ. Med. 61, 622–627. https://doi.org/10.1136/ hospital wastewater treatment plant. Water Res. 112, 110–119. https://doi.org/
oem.2003.011411. 10.1016/j.watres.2017.01.046.
Jones, R.N., Kugler, K.C., Pfaller, M.A., Winokur, P.L., 1999. Characteristics of pathogens Uhrbrand, K., Schultz, A.C., Madsen, A.M., 2011. Exposure to airborne noroviruses and
causing urinary tract infections in hospitals in North America: results from the other bioaerosol components at a wastewater treatment plant in Denmark. Food
SENTRY Antimicrobial Surveillance Program, 1997. Diagn. Microbiol. Infect. Dis. Environ. Virol. 3, 130–137. https://doi.org/10.1007/s12560-011-9068-3.
35, 55–63. https://doi.org/10.1016/S0732-8893(98)00158-8. US Environmental Protection Agency, 2011. Exposure Factors Handbook: 2011 Edition.
Kashiwagi, Y., Sato, S., Nakamura, M., Kuboshima, S., Numabe, H., Kawashima, H., U.S. Environ. Prot. Agency EPA/600/R-, 1–1466.EPA/600/R-090/052F.
Takekuma, K., Hoshika, A., Matsumoto, T., 2007. Klebsiella oxytoca septicemia Walterson, A.M., Stavrinides, J., 2015. Pantoea: insights into a highly versatile and
complicating rotavirus-associated acute diarrhea [2]. Pediatr. Infect. Dis. J. https:// diverse genus within the Enterobacteriaceae. FEMS Microbiol. Rev. https://doi.org/
doi.org/10.1097/01.inf.0000253041.67603.f5. 10.1093/femsre/fuv027.
Lee, J.A., Thorne, P.S., Reynolds, S.J., O’Shaughnessy, P.T., 2007. Monitoring risks in Wang, Y., Lan, H., Li, L., Yang, K., Qu, J., Liu, J., 2018. Chemicals and microbes in
association with exposure levels among wastewater treatment plant workers. bioaerosols from reaction tanks of six wastewater treatment plants: survival factors,
J. Occup. Environ. Med. 49, 1235–1248. https://doi.org/10.1097/ generation sources, and mechanisms. Sci. Rep. 8, 1–12. https://doi.org/10.1038/
JOM.0b013e3181568b40. s41598-018-27652-2.

9
R. Lu et al. Ecotoxicology and Environmental Safety 205 (2020) 111365

Wang, Y., Li, L., Xiong, R., Guo, X., Liu, J., 2019a. Effects of aeration on microbes and Xue, J., Zhang, J., Wu, Q.Y., Lu, Y., 2018. Sub-chronic inhalation of reclaimed water-
intestinal bacteria in bioaerosols from the BRT of an indoor wastewater treatment induced fibrotic lesion in a mouse model. Water Res. 139, 240–251. https://doi.org/
facility. Sci. Total Environ. 648, 1453–1461. https://doi.org/10.1016/j. 10.1016/j.watres.2018.04.008.
scitotenv.2018.08.244. Yang, K., Li, L., Wang, Y., Xue, S., Han, Y., Liu, J., 2019. Airborne bacteria in a
Wang, Y., Li, L., Xue, S., Han, Y., Yang, K., 2019b. Characteristics and formation wastewater treatment plant: emission characterization, source analysis and health
mechanism of intestinal bacteria particles emitted from aerated wastewater risk assessment. Water Res. 149, 596–606. https://doi.org/10.1016/j.
treatment tanks. Water Res. 163, 114862. https://doi.org/10.1016/j. watres.2018.11.027.
watres.2019.114862. Yang, T., Han, Y., Zhang, M., Xue, S., Li, L., Liu, J., Qiu, Z., 2019. Characteristics and
Xu, G., Han, Y., Li, L., Liu, J., 2018. Characterization and source analysis of indoor/ exposure risks of potential pathogens and toxic metal(loid)s in aerosols from
outdoor culturable airborne bacteria in a municipal wastewater treatment plant. wastewater treatment plants. Ecotoxicol. Environ. Saf. 183, 109543. https://doi.org/
J. Environ. Sci. (China) 74, 71–78. https://doi.org/10.1016/j.jes.2018.02.007. 10.1016/j.ecoenv.2019.109543.

10