TY - JOUR
AU - Nordby,, Karl-Christian
AB - Abstract Background The purpose of this study is to investigate whether airborne exposure to endotoxins, hydrogen sulphide (H2S), and inhalable particles negatively impacts the respiratory system and inflammatory blood proteins in sewage plant and sewer net system workers and, further, to determine dose-response associations between exposure and health outcomes. Methods In total, 148 waste water workers (WWWs) from urban and rural sewage plants and the sewer net system participated. One hundred and twenty-one workers were exposed to sewage, 46 from sewage plants and 75 from the sewer net system. Twenty-seven workers were characterized as little or not exposed and served as an internal reference group. Personal inhalable samples were analysed for endotoxins (Limulus assay), particle dust (gravimetrically) and Salmonella and Yersinia spp. (polymerase chain reaction method, PCR). Levels of H2S were measured using personal electro chemical sensors. Intercellular adhesion molecule 1 (ICAM-1), interleukin 8 (IL-8), surfactant protein D (SP-D), club cell protein 16 (CC16), and macrophage inflammatory protein (MIP) were determined by enzyme-linked immunosorbent assay and C-reactive protein (CRP) by an HS-MicroCRP assay in blood samples. Results Workers in sewage plants were exposed to significantly higher levels of endotoxins compared to workers in the sewer net system [median 55 EU m−3 (4–262 EU m−3) and median 27 EU m−3 (1–304 EU m−3), respectively]. The estimated H2S index showed higher values when working in the sewer net system [median 3.1 (0.5–78.1)] compared to workers at the sewage plants [median 1.3 (0.5–9.3)], and the most excessive exposure was collecting sewage from cesspools (273 p.p.m.). No viable airborne Salmonella and Yersinia spp. were detected. The exposed workers had significantly higher CRP compared to the referents [1.2 µg ml−1 (0.1–19.0 µg ml−1) and 0.8 µg ml−1 (0.1–5.0 µg ml−1), respectively] and lower forced expiratory volume in 1 s (FEV1)% [92.6%, standard deviation (SD) 14.6 and 102.0%, SD 10.1, respectively], with numbers given as mean and SD. The serum concentration of CRP was significantly and negatively associated with FEV1% (β = −7.7, R2 = 0.05) and forced vital capacity % (β = −8.5, R2 = 0.08), and the serum concentration of ICAM-1 with the estimated exposure to H2S (β = −19.9, R2 = 0.07). Conclusion Despite moderate levels of endotoxin and H2S exposure, the results indicate an impact of these agents on lung function and the adhesion molecule ICAM-1, and a low-grade systemic inflammation was indicated in increased levels of CRP. endotoxins, exposure, hydrogen sulphide, inflammatory markers, lung function, sewer net workers, waste water workers Introduction Waste water workers (WWWs) at sewage treatments plants are exposed to a complex mixture of non-infectious and infectious microorganisms, microbial components, and toxic gases. Several studies have shown that the range of symptoms are also broad, extending from central nervous system (CNS), respiratory, and gastrointestinal symptoms to infection and flu-like symptoms (Lundholm and Rylander, 1983; Melbostad et al., 1994; Douwes et al, 2001; Thorn et al., 2002; Heldal et al., 2010). Of all the airborne exposure associated with handling waste water at sewage treatment plants, hydrogen sulphide (H2S) from degradation of organic material in the sewage and endotoxins from Gram-negative bacteria are of particular interest (Adelson and Sunchine, 1966; Beauchamp et al., 1984; Latinen et al., 1994; Farahat and Kishk, 2010). High-dose acute exposure to H2S can be extremely hazardous, being able to cause pulmonary oedema (>300 p.p.m.) and sudden unconsciousness (>500 p.p.m.) or death by even a single exposure at levels above 1000 p.p.m. (NEG, 2001). However, repeated exposure to low levels of H2S (1–10 p.p.m.) may possibly result in chronic symptoms from the CNS, such as fatigue, headache, poor memory, and concentration difficulties (Tvedt et al., 1991; Richardson, 1995; Watt et al., 1997; Lee et al., 2007; Farahat and Kishk, 2010). Whether effects result from short-duration high exposure peaks or from slightly elevated average H2S exposure is of interest in the analysis of associations with health outcomes, and levels and variation of exposure in the work environment may be important to study in relation to CNS dysfunction. Results across industry-based studies warrant caution, are partly conflicting and suffer from methodological shortcomings regarding confounding, selection bias and other bias-related problems (Lewis and Copley 2015; Lim et al., 2016). When handling sewage, endotoxins can be present in levels exceeding those that may give rise to symptoms and illness (Rylander, 1999). Symptoms from the airways and impaired lung function among sewage plant workers compared to controls have been related to endotoxin exposure (Douwes et al., 2001; Thorn and Beijer 2004, Heldal et al., 2010). Recent studies have reported associations between exposure when handling sewage-related dust and both specific inflammatory pneumoproteins and systemic inflammatory biomarkers (Heldal et al., 2013, 2016). Both the inflammatory effects of endotoxins and irritant effects of H2S may have the respiratory system as a target organ. The potential effect from exposure to a mixture of these contaminants in low levels on the airways is less studied (Lee et al., 2007). Most of the published studies regarding exposure and health effects when handling sewage were carried out in waste water treatment facilities; thus, published reports regarding work in the sewer net including sewer underground pipes, chambers, and pumps are scarce (Neumann et al, 2002; Duquenne et al., 2014). The aim of this study is to investigate whether airborne exposure to endotoxin, H2S, and inhalable particles negatively impact the respiratory system and central nervous symptoms in WWWs, both at sewage plants and in the sewer net system. Further, we aimed to detect exposure to Salmonella and Yersinia spp. by personal sampling. A further aim was to investigate dose-response associations between exposure to endotoxin and health outcomes and to quantitate possible synergic effects of exposure on the respiratory system. Methods Description of work tasks The sewer net system is a complex network collecting domestic waste water, rain water, and industrial effluent for treatment at sewage plants. It comprises underground pipe galleries, channels, and pump stations leading waste water to the plants. Raw waste water carries a large amount of suspended matter that may deposit in the pipes and disturb the waste water flow and even obstruct the pipes. The sewage treatment process removes physical, chemical, and biological contaminants from the waste water by coarse filtering of small particles flocculated after addition of chemicals [FeCl3, Al2(SO4)3] in sedimentation tanks. The sediment sludge is then dewatered by pressing or centrifugation. The main tasks for workers at the urban sewage plants were daily inspection and cleaning activities. The work at the rural plants was more complex and additional tasks were inspection and cleaning of the network system such as pipes, pits, and pump stations connected to the plant. Study population Workers from four sewage treatment plants, being representatives of municipal plants operated to treat waste water from large urban cities (Oslo, Trondheim) and from four sewage treatment plants in small rural communities (Steinkjer, Støren, Klæbu, Selbu) were invited to participate in the study. Workers handling waste water in the sewer net system consisting of sewage pipes and pump stations were recruited from the same areas. In total, 148 workers (99%) of the invited workers accepted to participate (Table 1). One hundred and twenty-one workers were characterized as exposed to sewage, 46 from sewage treatment plants, and 75 from the sewer net system including underground pipe galleries, channels, and pump stations. Based on a questionnaire of job operations and our own inspections, 27 workers from the same working sites were characterized as low or not exposed to toxicants from the sewage. These data include workers from the TV-inspection group which have minimal contact with waste water, workers dealing with the fresh water net system in ditches, and workers with administration tasks with lower than 10% exposed working time. Workers relocated to administrative work because of health problems were not included in the reference group. This combined low-exposed group had a similar socioeconomic status as the exposed group, and they served as an internal reference group. At the beginning of the study, all participants were invited to the National Institute of Occupational Health in Oslo or St. Olav’s Hospital in Trondheim for a health examination consisting of spirometry, and a blood sample was collected for determination of biomarkers. Participants completed a self-administrated questionnaire of work-related symptoms, including questions regarding airway symptoms (irritations and cough), gastro-intestinal symptoms (nausea and diaré), and neuropsychological symptoms from the CNS (tiredness, concentration difficulties, and forgetfulness). Smoking and respiratory health status was recorded by a general questionnaire (Melbostad et al., 2001). The data were collected from April to June 2015. The study was approved by the Regional Medical Ethics Board. All participants gave their written informed consent. Table 1. Characteristics of the study population. Characteristics Referents n = 27 Workers All n = 121 Sewage plant n = 46 Sewage net system n = 75 Male, % 96 94 89 97 Age, yeara 40.5 (7.7) 44.6 (9.5) 42.9 (11.0) 45.7 (8.3) Atopy positive, % 29,6 22 27,3 18,9 Smoking, % 11,1 20,7 17,4 22,7 Body mass index, BMIa 28.4 (5.2) 28.6 (5.1) 27.7 (4.8) 29.2 (5.2) Characteristics Referents n = 27 Workers All n = 121 Sewage plant n = 46 Sewage net system n = 75 Male, % 96 94 89 97 Age, yeara 40.5 (7.7) 44.6 (9.5) 42.9 (11.0) 45.7 (8.3) Atopy positive, % 29,6 22 27,3 18,9 Smoking, % 11,1 20,7 17,4 22,7 Body mass index, BMIa 28.4 (5.2) 28.6 (5.1) 27.7 (4.8) 29.2 (5.2) aAverage (SD) View Large Table 1. Characteristics of the study population. Characteristics Referents n = 27 Workers All n = 121 Sewage plant n = 46 Sewage net system n = 75 Male, % 96 94 89 97 Age, yeara 40.5 (7.7) 44.6 (9.5) 42.9 (11.0) 45.7 (8.3) Atopy positive, % 29,6 22 27,3 18,9 Smoking, % 11,1 20,7 17,4 22,7 Body mass index, BMIa 28.4 (5.2) 28.6 (5.1) 27.7 (4.8) 29.2 (5.2) Characteristics Referents n = 27 Workers All n = 121 Sewage plant n = 46 Sewage net system n = 75 Male, % 96 94 89 97 Age, yeara 40.5 (7.7) 44.6 (9.5) 42.9 (11.0) 45.7 (8.3) Atopy positive, % 29,6 22 27,3 18,9 Smoking, % 11,1 20,7 17,4 22,7 Body mass index, BMIa 28.4 (5.2) 28.6 (5.1) 27.7 (4.8) 29.2 (5.2) aAverage (SD) View Large Spirometry Pulmonary function was measured by spirometry, which included measures of forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC) (Spirare SPS 330, Diagnostica, Oslo, Norway) according to the joint European Respiratory Society (ERS)/ATS guidelines (Miller et al., 2005). The same technician performed all tests in each of the test locations. Predicted values were based on European values for a reference population (Quanjer et al., 1993). Spirometric airflow limitation was defined by a FEV1/FVC ratio <70% and FEV1 <80% of predicted value (European Respiratory Society [ERS], 1993). Blood sampling and protein analysis in blood Blood samples were collected from workers and referents at approximately the same hour of the day, between 10 and 12 AM. Blood samples were collected in vacuum tubes containing ethylenediaminetetraacetic acid (EDTA) for plasma or without additives for serum (BD Diagnostic, Plymouth, UK). Plasma samples were used for measuring protein levels of surfactant protein D (SpD), club cell protein 16 (CC16), interleukin 8 (IL-8), intercellular adhesion molecule 1 (ICAM-1), and macrophage inflammatory protein (MIP)-1alpha (or CCL3) using a DuoSet enzyme-linked immunosorbent assay kit obtained from R&D systems (Stillwater, MN, USA). Serum was obtained after coagulation for 60 min at room temperature and centrifugation for 15 min at 1550 g. The serum samples were frozen in NUNC® cryotubes at −25°C no more than 2 h later and kept frozen until analysis. CRP was measured by an HS-MicroCRP assay with an immunoturbidmetric assay with latex-bounded anti-CRP antibody (Tina QuantRoche, Roche Diagnostic Corporation, Germany). Atopy for respiratory allergens were tested in serum using a Phadiatop test (FEIA, UniCAP system, Fürst Laboratory, Norway). Sampling strategy The measurements were performed in all eight sewage treatment plants with connected sewer net system. Personal samples were collected of workers equipped with a rucksack with two pumps connected to cassettes fixed to the straps of the rucksack in the breathing zone. One of the cassettes was for endotoxin analysis, the other for gravimetric determination of particulates and polymerase chain reaction (PCR) for determination of Salmonella and Yersinia spp. A H2S sensor was fixed on straps in the breathing zone of the worker. All exposure measurements were carried out from the beginning of the day and until lunchtime (sampling time 4–5 h) while most of the practical work was performed. Throughout the sampling period, the workers recorded their work operations, the use of personal protective device and breaks. The sampling was performed on the available workers during the sampling period (January to December 2015), both in the summer and winter period at each plant, in order to include seasonal variations. The number of samples performed at each plant is listed in Table 2. Table 2. Descriptive statistics for the concentration of endotoxin, dust particles, and calculated H2S index for the determinants work place, work operations, seasons, and degree of flushing. Work elements Exposure Endotoxin, EU m−3 Dust particles, mg m−3 H2S, index n Median (range) n Median (range) n Median (range) Work place Urban sewage plants VEAS 21 24 (3–303) 22 0.2 (0.1–1.9) 11 0.9 (0.3–36.7) Bekkelaget 14 50 (5–342) 13 0.3 (0.1–1.4) 11 1.0 (0.3–7.4) Rural plants and Støren 2 121 (57–185) 2 0.4 (0.3–0.5) 2 2.8 (0.3–5.5) net system Klæbu 8 77 (5–271) 9 0.3 (0.1–1.6) 8 8.4 (0.3–177.5) Selbu 8 119 (3–193) 8 0.2 (0.1–0.7) 7 13.1 (3.7–281.0) Steinkjer 8 7 (2–61) 26 0.2 (0.1–0.8) 15 4.4 (0.3–338.6) Urban network Oslo VAV 39 9 (2–304) 40 0.2 (0.1–1.8) 14 0.4 (0.3–75.1) sytem Trondheim VAV 60 18 (1–152) 14 0.4 (0.1–0.8) 25 0.7 (0.3–25.8) Work operations Inspection Urban plants 37 36 (3–342) 45 0.3 (0.1–1.9) 31 0.7 (0.3–36.7) Rural plants/network 23 30 (2–271) 22 0.2 (0.1–1.6) 25 5.5 (0.3–281.0) Urban network 34 7 (1–150) 19 0.3 (0.1–1.8) 12 0.5 (0.3–55.5) Septic/sludge 44 37 (2–304) 35 0.3 (0.1–1.5) 25 3.0 (0.3–338.6) Greace 10 13 (5–70) – – TV-inspection/ditches 12 9 (3–30) 13 0.2 (0.1–1.2) 4 5.1 (1.0–8.0) Season Winter 53 8 (2–193) 34 0.3 (0.1–1.8) 26 3.4 (0.3–281.0) Spring 18 56 (6–152) 14 0.4 (0.1–0.8) 14 0.7 (0.3–5.3) Summer 37 17 (1–304) 46 0.2 (0.1–1.6) 25 6.3 (0.3–338.6) Autumn 52 26 (2–342) 40 0.3 (0.1–1.9) 28 0.3 (0.3–55.5) Flushing No 51 12 (2–271) 51 0.3 (0.1–1.8) 34 0.6 (0.03–55.5) 1–3 times 50 38 (3–342) 47 0.3 (0.1–1.9) 31 1.2 (0.3–25.8) More than 3 times 56 17 (1–304) 36 0.2 (0.1–1.1) 28 7.6 (0.3–338.6) Work elements Exposure Endotoxin, EU m−3 Dust particles, mg m−3 H2S, index n Median (range) n Median (range) n Median (range) Work place Urban sewage plants VEAS 21 24 (3–303) 22 0.2 (0.1–1.9) 11 0.9 (0.3–36.7) Bekkelaget 14 50 (5–342) 13 0.3 (0.1–1.4) 11 1.0 (0.3–7.4) Rural plants and Støren 2 121 (57–185) 2 0.4 (0.3–0.5) 2 2.8 (0.3–5.5) net system Klæbu 8 77 (5–271) 9 0.3 (0.1–1.6) 8 8.4 (0.3–177.5) Selbu 8 119 (3–193) 8 0.2 (0.1–0.7) 7 13.1 (3.7–281.0) Steinkjer 8 7 (2–61) 26 0.2 (0.1–0.8) 15 4.4 (0.3–338.6) Urban network Oslo VAV 39 9 (2–304) 40 0.2 (0.1–1.8) 14 0.4 (0.3–75.1) sytem Trondheim VAV 60 18 (1–152) 14 0.4 (0.1–0.8) 25 0.7 (0.3–25.8) Work operations Inspection Urban plants 37 36 (3–342) 45 0.3 (0.1–1.9) 31 0.7 (0.3–36.7) Rural plants/network 23 30 (2–271) 22 0.2 (0.1–1.6) 25 5.5 (0.3–281.0) Urban network 34 7 (1–150) 19 0.3 (0.1–1.8) 12 0.5 (0.3–55.5) Septic/sludge 44 37 (2–304) 35 0.3 (0.1–1.5) 25 3.0 (0.3–338.6) Greace 10 13 (5–70) – – TV-inspection/ditches 12 9 (3–30) 13 0.2 (0.1–1.2) 4 5.1 (1.0–8.0) Season Winter 53 8 (2–193) 34 0.3 (0.1–1.8) 26 3.4 (0.3–281.0) Spring 18 56 (6–152) 14 0.4 (0.1–0.8) 14 0.7 (0.3–5.3) Summer 37 17 (1–304) 46 0.2 (0.1–1.6) 25 6.3 (0.3–338.6) Autumn 52 26 (2–342) 40 0.3 (0.1–1.9) 28 0.3 (0.3–55.5) Flushing No 51 12 (2–271) 51 0.3 (0.1–1.8) 34 0.6 (0.03–55.5) 1–3 times 50 38 (3–342) 47 0.3 (0.1–1.9) 31 1.2 (0.3–25.8) More than 3 times 56 17 (1–304) 36 0.2 (0.1–1.1) 28 7.6 (0.3–338.6) Descriptive statistics expressed with number of measurements and the corresponding median, minimum, and maximum values (range) of exposure to three different agents influencing the WWWs. View Large Table 2. Descriptive statistics for the concentration of endotoxin, dust particles, and calculated H2S index for the determinants work place, work operations, seasons, and degree of flushing. Work elements Exposure Endotoxin, EU m−3 Dust particles, mg m−3 H2S, index n Median (range) n Median (range) n Median (range) Work place Urban sewage plants VEAS 21 24 (3–303) 22 0.2 (0.1–1.9) 11 0.9 (0.3–36.7) Bekkelaget 14 50 (5–342) 13 0.3 (0.1–1.4) 11 1.0 (0.3–7.4) Rural plants and Støren 2 121 (57–185) 2 0.4 (0.3–0.5) 2 2.8 (0.3–5.5) net system Klæbu 8 77 (5–271) 9 0.3 (0.1–1.6) 8 8.4 (0.3–177.5) Selbu 8 119 (3–193) 8 0.2 (0.1–0.7) 7 13.1 (3.7–281.0) Steinkjer 8 7 (2–61) 26 0.2 (0.1–0.8) 15 4.4 (0.3–338.6) Urban network Oslo VAV 39 9 (2–304) 40 0.2 (0.1–1.8) 14 0.4 (0.3–75.1) sytem Trondheim VAV 60 18 (1–152) 14 0.4 (0.1–0.8) 25 0.7 (0.3–25.8) Work operations Inspection Urban plants 37 36 (3–342) 45 0.3 (0.1–1.9) 31 0.7 (0.3–36.7) Rural plants/network 23 30 (2–271) 22 0.2 (0.1–1.6) 25 5.5 (0.3–281.0) Urban network 34 7 (1–150) 19 0.3 (0.1–1.8) 12 0.5 (0.3–55.5) Septic/sludge 44 37 (2–304) 35 0.3 (0.1–1.5) 25 3.0 (0.3–338.6) Greace 10 13 (5–70) – – TV-inspection/ditches 12 9 (3–30) 13 0.2 (0.1–1.2) 4 5.1 (1.0–8.0) Season Winter 53 8 (2–193) 34 0.3 (0.1–1.8) 26 3.4 (0.3–281.0) Spring 18 56 (6–152) 14 0.4 (0.1–0.8) 14 0.7 (0.3–5.3) Summer 37 17 (1–304) 46 0.2 (0.1–1.6) 25 6.3 (0.3–338.6) Autumn 52 26 (2–342) 40 0.3 (0.1–1.9) 28 0.3 (0.3–55.5) Flushing No 51 12 (2–271) 51 0.3 (0.1–1.8) 34 0.6 (0.03–55.5) 1–3 times 50 38 (3–342) 47 0.3 (0.1–1.9) 31 1.2 (0.3–25.8) More than 3 times 56 17 (1–304) 36 0.2 (0.1–1.1) 28 7.6 (0.3–338.6) Work elements Exposure Endotoxin, EU m−3 Dust particles, mg m−3 H2S, index n Median (range) n Median (range) n Median (range) Work place Urban sewage plants VEAS 21 24 (3–303) 22 0.2 (0.1–1.9) 11 0.9 (0.3–36.7) Bekkelaget 14 50 (5–342) 13 0.3 (0.1–1.4) 11 1.0 (0.3–7.4) Rural plants and Støren 2 121 (57–185) 2 0.4 (0.3–0.5) 2 2.8 (0.3–5.5) net system Klæbu 8 77 (5–271) 9 0.3 (0.1–1.6) 8 8.4 (0.3–177.5) Selbu 8 119 (3–193) 8 0.2 (0.1–0.7) 7 13.1 (3.7–281.0) Steinkjer 8 7 (2–61) 26 0.2 (0.1–0.8) 15 4.4 (0.3–338.6) Urban network Oslo VAV 39 9 (2–304) 40 0.2 (0.1–1.8) 14 0.4 (0.3–75.1) sytem Trondheim VAV 60 18 (1–152) 14 0.4 (0.1–0.8) 25 0.7 (0.3–25.8) Work operations Inspection Urban plants 37 36 (3–342) 45 0.3 (0.1–1.9) 31 0.7 (0.3–36.7) Rural plants/network 23 30 (2–271) 22 0.2 (0.1–1.6) 25 5.5 (0.3–281.0) Urban network 34 7 (1–150) 19 0.3 (0.1–1.8) 12 0.5 (0.3–55.5) Septic/sludge 44 37 (2–304) 35 0.3 (0.1–1.5) 25 3.0 (0.3–338.6) Greace 10 13 (5–70) – – TV-inspection/ditches 12 9 (3–30) 13 0.2 (0.1–1.2) 4 5.1 (1.0–8.0) Season Winter 53 8 (2–193) 34 0.3 (0.1–1.8) 26 3.4 (0.3–281.0) Spring 18 56 (6–152) 14 0.4 (0.1–0.8) 14 0.7 (0.3–5.3) Summer 37 17 (1–304) 46 0.2 (0.1–1.6) 25 6.3 (0.3–338.6) Autumn 52 26 (2–342) 40 0.3 (0.1–1.9) 28 0.3 (0.3–55.5) Flushing No 51 12 (2–271) 51 0.3 (0.1–1.8) 34 0.6 (0.03–55.5) 1–3 times 50 38 (3–342) 47 0.3 (0.1–1.9) 31 1.2 (0.3–25.8) More than 3 times 56 17 (1–304) 36 0.2 (0.1–1.1) 28 7.6 (0.3–338.6) Descriptive statistics expressed with number of measurements and the corresponding median, minimum, and maximum values (range) of exposure to three different agents influencing the WWWs. View Large Personal sampling of airborne endotoxins, dust, and pathogens Exposure to endotoxins, dust particles, and bacteria (Salmonella and Yersinia spp.) was assessed in samples collected with inhalable personal air samplers with a 6 mm inlet (PAS6) cassettes (manufactured at the workshop of the NIOH, Oslo) (Van der Wal 1983). The cassettes were connected to two pumps (PS103) operated at a flow rate of 2.0 l min−1. The sampling time was approximately 4–5 h. Aerosols for the determination of dust and viable bacteria (Salmonella and Yersinia spp.) were collected on polycarbonate filters with pore size 0.8 µm (Poretics, Osmonics, Livermore, CA, USA), while endotoxins were collected on glass fibre filters (Whatman GF/A, Maidstone, VT, USA). The samples were stored at +4°C and transported to the laboratory by car the same or the following day for analyses. The dust mass was determined gravimetrically in a climate-controlled weighing room. Endotoxins were extracted with 0.05% (v/v) Tween 20 (Merck Schuchard) in pyrogen-free water, rocked vigorously at room temperature and centrifuged at 1000 g for 10 min (Douwes et al., 1995). The suspension was analysed by a quantitative kinetic chromogenic Limulus amoebocyte lysate assay according to the manufacturer’s instructions (Cambrex, NJ, USA). As a part of standard procedure airborne bacteria were pre-cultivated in order to reach a higher concentration of cells of the target pathogens (Salmonella and Yersinia spp.) suitable for detection by the PCR method (BAX System PCR Assay for Salmonella spp. and ISO/TS 18867:2015 for Yersinia spp.). Filters were put in 100 ml of buffered peptone water (BPW, at pH 7.2), shaken gently for 15–30 min and then the suspension was divided in two aliquots for further analysis. The buffer with material from the filters was incubated for growth as follows: Salmonella at 37°C for 16–20 h, Yersinia at 25°C for 20–24 h. The two cultures were used for DNA extraction with the paramagnetic bead method and then analysed based on Real Time PCR assay with Taqman. This analysis was performed by a commercial laboratory (Eurofins Steins Laboratorium A/S, Veje, Denmark). Calculation of H2S index H2S was measured using direct reading instruments with logging OdaLog L2/LL (ThermoFisher Scientific Inc. Australia) with a quantification range of 0.1–200 p.p.m., Dräger GasVisi X-am 500 and Dräger Pac 700 (Dräger Safety, Norway) and 0–100 p.p.m., respectively. The concentrations were measured every 15 s and the average of each interval of 15 s was recorded using data loggers. H2S exposure was registered by the number of peaks recorded above 0.1 p.p.m. (H2S01), 1 p.p.m. (H2S1), 5.0 p.p.m. (H2S5), and 10.0 p.p.m. (H2S10), as well as the duration of the peaks. To combine these measurement results into a single value, an H2S index was computed, as described earlier (Austigard et al., 2018): Hydrogen sulphide index:  =H2S01×0.1 + H2Sduration01 × 0.1 + H2S1+ H2S5×5 + H2Sduration5× 5 + H2S10×10 + H2Smax A background sulphide level was estimated as 0.4 √2−1 = 0.28, where 0.4 is the lowest calculated sulphide index, and all results with a 0 sulphide index were replaced with this value. Statistical analyses The variables were summarized with median, minimum, and maximum values. Thereafter, parametric statistical methods were used. Continuous variables were log-transformed to achieve normal distribution when the skewness exceeded 2.0 and, accordingly, all exposure and biomarker variables, except for ICAM-1, were log-transformed. One-way analysis of variance (ANOVA) was used when more than two groups were compared with a post hoc least significant difference test. Multiple linear regression (backward strategy) was used to assess associations between dependent and exposure variables simultaneously and to study interaction effects between exposures on lung function. General linear models of relevant parameters were used to calculate group estimates, adjusted for smoke, age, body mass index (BMI), and atopy. Exposure was modelled by linear mixed effect regression to account for the correlation between repeated measurements. Determinants of exposure were treated as fixed effects. The models were built stepwise including work operations, work place, season, and the amount of high-pressure flushing. Determinants that significantly improved the models as judged by a P-value <0.05 in likelihood ratio tests were kept in the models. Example of calculation: geometric mean (GM) exposure to endotoxin in a worker from rural plant/net in the autumn and at Selbu with more than 1 to 3 episodes of flushing during the work shift: 0.745 × 2.64  × 3.35  × 22.64 × 0.79= 117.8 EUm−3 AM (Arithmetic mean: exp[log-transformed exposure + 0.5(σ2B + σ2W)] (Friesen et al., 2006): Endotoxin  exposure = GM × e(0.5 × varians)= 117.8 × e0.595                               =117.8 × 1.81=213.6 EUm−3 where variance within workers (W) = 1.15 and variance between workers (B) = 0.04. All analyses were performed using IBM SPSS Statistics 20 (IBM, Armonk, NY, USA). Results Exposure levels of airborne endotoxins and dust particles In total, 160 airborne samples were collected during a period of 1 year (2015) and analysed for endotoxins and dust exposure. Because a limited number of electrochemical sensors were available, only 93 personal measurements of H2S were performed. The work place, typical work operations, seasons, and number of flushing operations during a day contributed to the explanation of the exposure concentration levels. A descriptive overview of the levels of exposure to endotoxins and dust particles and H2S index considering these determinants are shown in Table 2. The endotoxin exposure levels in the 160 observations obtained during handling of sewage at sewage plants or in the sewer net system were moderate with a highest concentration of 342 EU m−3 measured during 1–3 flushing operations in a working day. Inspection of the sewer net system (TV-inspection) [median 7 EU m−3 (1–150 EU m−3)] and grease handling [median 13 EU m−3 (5–70 EU m−3)] showed the lowest exposure levels during different working operations (P < 0.05). The exposure levels also varied by season, with the highest concentration during the spring time [median 56 EU m−3 (6–152 EU m−3), P < 0.05]. The inhalable dust exposure was low (0.1–1.9 mg m−3), with no significant differences between working operations. No viable Salmonella and Yersinia spp. were detected in any of the air samples (not shown). The correlations between the exposure variables endotoxin and dust were low (rs = –0.07–0.13). Estimated exposure levels based on a linear mixed effect model of the 160 observations (Table 5) of the determinants were used in the assessment of associations between exposure and health effects among 121 sewage plants and sewer net workers. Workers in the sewage plants were significantly more exposed to endotoxins compared to those working in the sewer net system [median 55 EU m−3 (4–262 EU m−3) and median 27 EU m−3 (1–304 EU m−3), respectively] but showed significantly lower values of H2S index than workers in the sewer net system [median 1.3 (0.5–9.3) and median 3.1 (3.1–78.1), respectively]. Table 5. Linear mixed effect model (MEM) of exposure determinants for endotoxin, dust particles (ln values), and H2S index. Determinants ln endotoxin ln dust particles H2S index MEM B (SE) eMEM B MEM B (SE) eMEM B MEM B (SE) eMEM B Intercept −0.295 0.742 0.745 −1.67* 0.28 0.19 5.02* 1.35 151.4 Work operation  Rural plant/net 0.97 0.60 2.64 0.16 0.34 1.17 −1.73 1.51 0.18  Septic 1.26* 0.46 3.53 0.15 0.30 1.16 −0.42 1.46 0.66  Sewage plants 1.23 0.77 3.42 −0.21 0.28 0.81 −2.55 1.52 0.08  Grease 1.01 0.64 2.75 –  Sewer net 0.32 0.44 1.38 0.02 0.30 1.02 0 0 1  TV inspec/ditches 0 0 1 0 0 1 Seasons  Autumn 1.21* 0.27 3.35 −1.36* 0.65 0.26  Summer 0.61* 0.29 1.84 −1.10 0.82 0.33  Winter 1.56* 0.36 4.76 −2.49* 0.88 0.08  Spring 0 0 1 0 0 1.00 Work place  Veas 2.02* 0.54 7.54 −1.14 1.10 0.32  Bekkelaget 2.26* 0.59 9.58 0.07 1.22 1.07  Oslo 2.15* 0.67 8.58 −2.80* 1.09 0.06  Støren 4.19* 1.04 66.02 −1.90 1.33 0.15  Klæbu 3.19* 0.75 24.29 −0.38 0.69 0.68  Selbu 3.12 0.74 22.64 0.57 0.90 1.77  Trondheim 1.65* 0.61 5.21 −2.07* 1.00 0.13  Steinkjer 0 0 1 0 0 1 Flushing  Not −0.78* 0.26 0.46 0.56* 0.21 1.75 −1.10* 0.47 0.33  1 to 3 times −0.23 0.26 0.79 0.61* 0.21 1.84 −1.16* 0.45 0.31  More than three times 0 0 1 0 0 1 0 0 1.00 Determinants ln endotoxin ln dust particles H2S index MEM B (SE) eMEM B MEM B (SE) eMEM B MEM B (SE) eMEM B Intercept −0.295 0.742 0.745 −1.67* 0.28 0.19 5.02* 1.35 151.4 Work operation  Rural plant/net 0.97 0.60 2.64 0.16 0.34 1.17 −1.73 1.51 0.18  Septic 1.26* 0.46 3.53 0.15 0.30 1.16 −0.42 1.46 0.66  Sewage plants 1.23 0.77 3.42 −0.21 0.28 0.81 −2.55 1.52 0.08  Grease 1.01 0.64 2.75 –  Sewer net 0.32 0.44 1.38 0.02 0.30 1.02 0 0 1  TV inspec/ditches 0 0 1 0 0 1 Seasons  Autumn 1.21* 0.27 3.35 −1.36* 0.65 0.26  Summer 0.61* 0.29 1.84 −1.10 0.82 0.33  Winter 1.56* 0.36 4.76 −2.49* 0.88 0.08  Spring 0 0 1 0 0 1.00 Work place  Veas 2.02* 0.54 7.54 −1.14 1.10 0.32  Bekkelaget 2.26* 0.59 9.58 0.07 1.22 1.07  Oslo 2.15* 0.67 8.58 −2.80* 1.09 0.06  Støren 4.19* 1.04 66.02 −1.90 1.33 0.15  Klæbu 3.19* 0.75 24.29 −0.38 0.69 0.68  Selbu 3.12 0.74 22.64 0.57 0.90 1.77  Trondheim 1.65* 0.61 5.21 −2.07* 1.00 0.13  Steinkjer 0 0 1 0 0 1 Flushing  Not −0.78* 0.26 0.46 0.56* 0.21 1.75 −1.10* 0.47 0.33  1 to 3 times −0.23 0.26 0.79 0.61* 0.21 1.84 −1.16* 0.45 0.31  More than three times 0 0 1 0 0 1 0 0 1.00 SE, standard error. *P < 0.05. View Large Table 5. Linear mixed effect model (MEM) of exposure determinants for endotoxin, dust particles (ln values), and H2S index. Determinants ln endotoxin ln dust particles H2S index MEM B (SE) eMEM B MEM B (SE) eMEM B MEM B (SE) eMEM B Intercept −0.295 0.742 0.745 −1.67* 0.28 0.19 5.02* 1.35 151.4 Work operation  Rural plant/net 0.97 0.60 2.64 0.16 0.34 1.17 −1.73 1.51 0.18  Septic 1.26* 0.46 3.53 0.15 0.30 1.16 −0.42 1.46 0.66  Sewage plants 1.23 0.77 3.42 −0.21 0.28 0.81 −2.55 1.52 0.08  Grease 1.01 0.64 2.75 –  Sewer net 0.32 0.44 1.38 0.02 0.30 1.02 0 0 1  TV inspec/ditches 0 0 1 0 0 1 Seasons  Autumn 1.21* 0.27 3.35 −1.36* 0.65 0.26  Summer 0.61* 0.29 1.84 −1.10 0.82 0.33  Winter 1.56* 0.36 4.76 −2.49* 0.88 0.08  Spring 0 0 1 0 0 1.00 Work place  Veas 2.02* 0.54 7.54 −1.14 1.10 0.32  Bekkelaget 2.26* 0.59 9.58 0.07 1.22 1.07  Oslo 2.15* 0.67 8.58 −2.80* 1.09 0.06  Støren 4.19* 1.04 66.02 −1.90 1.33 0.15  Klæbu 3.19* 0.75 24.29 −0.38 0.69 0.68  Selbu 3.12 0.74 22.64 0.57 0.90 1.77  Trondheim 1.65* 0.61 5.21 −2.07* 1.00 0.13  Steinkjer 0 0 1 0 0 1 Flushing  Not −0.78* 0.26 0.46 0.56* 0.21 1.75 −1.10* 0.47 0.33  1 to 3 times −0.23 0.26 0.79 0.61* 0.21 1.84 −1.16* 0.45 0.31  More than three times 0 0 1 0 0 1 0 0 1.00 Determinants ln endotoxin ln dust particles H2S index MEM B (SE) eMEM B MEM B (SE) eMEM B MEM B (SE) eMEM B Intercept −0.295 0.742 0.745 −1.67* 0.28 0.19 5.02* 1.35 151.4 Work operation  Rural plant/net 0.97 0.60 2.64 0.16 0.34 1.17 −1.73 1.51 0.18  Septic 1.26* 0.46 3.53 0.15 0.30 1.16 −0.42 1.46 0.66  Sewage plants 1.23 0.77 3.42 −0.21 0.28 0.81 −2.55 1.52 0.08  Grease 1.01 0.64 2.75 –  Sewer net 0.32 0.44 1.38 0.02 0.30 1.02 0 0 1  TV inspec/ditches 0 0 1 0 0 1 Seasons  Autumn 1.21* 0.27 3.35 −1.36* 0.65 0.26  Summer 0.61* 0.29 1.84 −1.10 0.82 0.33  Winter 1.56* 0.36 4.76 −2.49* 0.88 0.08  Spring 0 0 1 0 0 1.00 Work place  Veas 2.02* 0.54 7.54 −1.14 1.10 0.32  Bekkelaget 2.26* 0.59 9.58 0.07 1.22 1.07  Oslo 2.15* 0.67 8.58 −2.80* 1.09 0.06  Støren 4.19* 1.04 66.02 −1.90 1.33 0.15  Klæbu 3.19* 0.75 24.29 −0.38 0.69 0.68  Selbu 3.12 0.74 22.64 0.57 0.90 1.77  Trondheim 1.65* 0.61 5.21 −2.07* 1.00 0.13  Steinkjer 0 0 1 0 0 1 Flushing  Not −0.78* 0.26 0.46 0.56* 0.21 1.75 −1.10* 0.47 0.33  1 to 3 times −0.23 0.26 0.79 0.61* 0.21 1.84 −1.16* 0.45 0.31  More than three times 0 0 1 0 0 1 0 0 1.00 SE, standard error. *P < 0.05. View Large Measurements of H2S Twenty-nine percent of the 93 measurements of H2S were below the detection limit (0 and 0.1 p.p.m.). Nine percent of all H2S recordings showed peaks above 10 p.p.m.; in addition, 15% have peaks of 5–10 p.p.m., 35% have peaks of 1–5 p.p.m., and 65% have peaks of 0.1–1 p.p.m. Pump stations and small sewage plants, work in sewer net system, and collecting sewage from cesspools all had shifts with H2S peaks above 10 p.p.m. The job with most excessive exposure to H2S was collecting sewage from cesspools (273 p.p.m.), with 15% of all shifts having peaks above 10 p.p.m. Health effects Work-related symptoms In total, 148 WWWs, 121 exposed and 27 referents, participated in the health examination of the study. The prevalence of work-related symptoms was generally higher among sewage plant and sewer workers compared to referents, but only airway symptoms were significantly higher (33 and 11%, respectively). In particular, cough (57%), nausea (43%), and fever attack (35%) were often reported among exposed workers. There was no difference in reported symptoms between workers at sewage plants and the sewer net system. The same tendency was observed with symptoms from the CNS. Concentration difficulties (32%), forgetfulness (47%), and tiredness (46%) were the symptoms most often reported among the sewage plant and sewer workers. Spirometry Table 3 summarizes the lung function measures of referents, sewage plant, and sewer net workers. The predicted FEV1% and FVC% were lower for all exposed workers (92.6 and 97.5%, respectively) compared to the referents (102.0 and 106.4%), respectively, in analyses adjusted for smoking, atopy, age, and BMI. The FEV1% predicted was also lower among the sewage plant working group (93.0%) and sewer net working group (92.4%) compared to the referents. However, there were no significant differences in the lung function measures between the sewage plant and the sewer net working group. Nine of non-smoking workers (8%) showed moderate spirometric airflow limitation according to the ERS criteria (FEV1<80% of predicted value and FEV1/FVC ratio <70%). Table 3. Lung function in workers and referents. Parameter Referents (n = 27) Workers AM (SD) All (n = 119) Sewage plant (n = 45) Sewer net (n = 74) AM (SD) AM (SD) AM (SD) FEV1% of predicteda 102.0 (10.1)b,c,d 92.6 (14.6)b 93.0 (16.4)c 92.4 (13.4)d FVC% of predicteda 106.4 (11.5)b 97.9 (12.9)b 99.7 (13.1) 96.8 (12.7) FEV1/FVC 0.79 (0.1) 0.77 (0.1) 0.76 (0.1) 0.78 (0.1) Parameter Referents (n = 27) Workers AM (SD) All (n = 119) Sewage plant (n = 45) Sewer net (n = 74) AM (SD) AM (SD) AM (SD) FEV1% of predicteda 102.0 (10.1)b,c,d 92.6 (14.6)b 93.0 (16.4)c 92.4 (13.4)d FVC% of predicteda 106.4 (11.5)b 97.9 (12.9)b 99.7 (13.1) 96.8 (12.7) FEV1/FVC 0.79 (0.1) 0.77 (0.1) 0.76 (0.1) 0.78 (0.1) AM, arithmetic mean; SD, standard deviation. aPredicted using European normal values for sex, age and height bBetween referents and all workers. cBetween referents and plant workers. dBetween referents and net workers. *P < 0.05 (GLM test) adjusted for age, smoking, and BMI. View Large Table 3. Lung function in workers and referents. Parameter Referents (n = 27) Workers AM (SD) All (n = 119) Sewage plant (n = 45) Sewer net (n = 74) AM (SD) AM (SD) AM (SD) FEV1% of predicteda 102.0 (10.1)b,c,d 92.6 (14.6)b 93.0 (16.4)c 92.4 (13.4)d FVC% of predicteda 106.4 (11.5)b 97.9 (12.9)b 99.7 (13.1) 96.8 (12.7) FEV1/FVC 0.79 (0.1) 0.77 (0.1) 0.76 (0.1) 0.78 (0.1) Parameter Referents (n = 27) Workers AM (SD) All (n = 119) Sewage plant (n = 45) Sewer net (n = 74) AM (SD) AM (SD) AM (SD) FEV1% of predicteda 102.0 (10.1)b,c,d 92.6 (14.6)b 93.0 (16.4)c 92.4 (13.4)d FVC% of predicteda 106.4 (11.5)b 97.9 (12.9)b 99.7 (13.1) 96.8 (12.7) FEV1/FVC 0.79 (0.1) 0.77 (0.1) 0.76 (0.1) 0.78 (0.1) AM, arithmetic mean; SD, standard deviation. aPredicted using European normal values for sex, age and height bBetween referents and all workers. cBetween referents and plant workers. dBetween referents and net workers. *P < 0.05 (GLM test) adjusted for age, smoking, and BMI. View Large Biomarkers of inflammation The results are shown in Table 4. The acute phase protein CRP was elevated in all exposed workers compared to the reference group [median (range) 1.2 µg ml−1 (0.1–19.0 µg ml−1) and 0.8 µg ml−1 (0.1–5.0 µg ml−1) P < 0.05, respectively]. Workers at the sewer net system (0.3–19.0 µg ml−1) demonstrated higher CRP levels compared to workers at the sewage treatment plants (0.1–14.0 µg ml−1, P < 0.05). The concentration of CC16 was significantly lower among the sewer net workers compared to the referents [median 27 ng ml−1 (7–300 ng ml−1) and 34 ng/ml (5–152 ng ml−1), respectively, P < 0.05]. The level of adhesion molecule ICAM-1 in all exposed workers was close to significantly higher than the referents [median 98 ng ml−1 (25–237 ng ml−1) and 75 ng ml−1 (26–174 ng ml−1), respectively, P = 0.06]. No other of the proteins measured showed significant differences between workers and referents or between working groups. Table 4. The concentrations of biomarkers in sewage plant and sewer system workers and referents. Biomarkers Referents Workers na AM Med (range) All Sewage plants Sewer system n AM Med (range) n AM Med (range) n AM Med (range) CRP µg ml−1b,c,d 27 1.4b,c 0.8 (0.1–5.0) 118 2.6 1.2b (0.1–19.0) 44 1.9d 1.2 (0.1–14.0) 74 3cd 1.5 (0.3–19) ICAM-1 ng ml−1 27 84 75 (26–174) 118 97 90 (25–237) 44 88 83 (36–188) 74 102 98 (25–237) IL-8 pg ml−1 26 13 84 (6–56) 118 26 12 (4–389) 44 37 13 (4–389) 74 18 10 (4–278) Sp-D pg ml−1 27 14 8 (0.4–159) 118 14 7 (0.3–146) 44 12 7 (0.3–74) 74 14 7 (0.3–146) CC16 ng ml−1c 26 44 34 (5–152) 118 35 30 (7–131) 44 39 36 (8–98) 74 32b 27 (7–130) MIP1a pg ml−1 27 99 9 (0.1–748) 111 65 8 (0.3–735) 44 53 6 (0.1–735) 74 65 6 (0.1–650) MIP1b pg ml−1 26 25 22 (4–63) 110 22 71 (2–71) 43 22 18 (2–71) 68 21 19 (0.1–69) Biomarkers Referents Workers na AM Med (range) All Sewage plants Sewer system n AM Med (range) n AM Med (range) n AM Med (range) CRP µg ml−1b,c,d 27 1.4b,c 0.8 (0.1–5.0) 118 2.6 1.2b (0.1–19.0) 44 1.9d 1.2 (0.1–14.0) 74 3cd 1.5 (0.3–19) ICAM-1 ng ml−1 27 84 75 (26–174) 118 97 90 (25–237) 44 88 83 (36–188) 74 102 98 (25–237) IL-8 pg ml−1 26 13 84 (6–56) 118 26 12 (4–389) 44 37 13 (4–389) 74 18 10 (4–278) Sp-D pg ml−1 27 14 8 (0.4–159) 118 14 7 (0.3–146) 44 12 7 (0.3–74) 74 14 7 (0.3–146) CC16 ng ml−1c 26 44 34 (5–152) 118 35 30 (7–131) 44 39 36 (8–98) 74 32b 27 (7–130) MIP1a pg ml−1 27 99 9 (0.1–748) 111 65 8 (0.3–735) 44 53 6 (0.1–735) 74 65 6 (0.1–650) MIP1b pg ml−1 26 25 22 (4–63) 110 22 71 (2–71) 43 22 18 (2–71) 68 21 19 (0.1–69) AM, aritmetic mean; Med (range), median and range (lowest highest value). aNumber of measurements. bBetween referents and all workers. cBetween referents and net workers. dBetween sewage plant and net workers. P < 0.05. View Large Table 4. The concentrations of biomarkers in sewage plant and sewer system workers and referents. Biomarkers Referents Workers na AM Med (range) All Sewage plants Sewer system n AM Med (range) n AM Med (range) n AM Med (range) CRP µg ml−1b,c,d 27 1.4b,c 0.8 (0.1–5.0) 118 2.6 1.2b (0.1–19.0) 44 1.9d 1.2 (0.1–14.0) 74 3cd 1.5 (0.3–19) ICAM-1 ng ml−1 27 84 75 (26–174) 118 97 90 (25–237) 44 88 83 (36–188) 74 102 98 (25–237) IL-8 pg ml−1 26 13 84 (6–56) 118 26 12 (4–389) 44 37 13 (4–389) 74 18 10 (4–278) Sp-D pg ml−1 27 14 8 (0.4–159) 118 14 7 (0.3–146) 44 12 7 (0.3–74) 74 14 7 (0.3–146) CC16 ng ml−1c 26 44 34 (5–152) 118 35 30 (7–131) 44 39 36 (8–98) 74 32b 27 (7–130) MIP1a pg ml−1 27 99 9 (0.1–748) 111 65 8 (0.3–735) 44 53 6 (0.1–735) 74 65 6 (0.1–650) MIP1b pg ml−1 26 25 22 (4–63) 110 22 71 (2–71) 43 22 18 (2–71) 68 21 19 (0.1–69) Biomarkers Referents Workers na AM Med (range) All Sewage plants Sewer system n AM Med (range) n AM Med (range) n AM Med (range) CRP µg ml−1b,c,d 27 1.4b,c 0.8 (0.1–5.0) 118 2.6 1.2b (0.1–19.0) 44 1.9d 1.2 (0.1–14.0) 74 3cd 1.5 (0.3–19) ICAM-1 ng ml−1 27 84 75 (26–174) 118 97 90 (25–237) 44 88 83 (36–188) 74 102 98 (25–237) IL-8 pg ml−1 26 13 84 (6–56) 118 26 12 (4–389) 44 37 13 (4–389) 74 18 10 (4–278) Sp-D pg ml−1 27 14 8 (0.4–159) 118 14 7 (0.3–146) 44 12 7 (0.3–74) 74 14 7 (0.3–146) CC16 ng ml−1c 26 44 34 (5–152) 118 35 30 (7–131) 44 39 36 (8–98) 74 32b 27 (7–130) MIP1a pg ml−1 27 99 9 (0.1–748) 111 65 8 (0.3–735) 44 53 6 (0.1–735) 74 65 6 (0.1–650) MIP1b pg ml−1 26 25 22 (4–63) 110 22 71 (2–71) 43 22 18 (2–71) 68 21 19 (0.1–69) AM, aritmetic mean; Med (range), median and range (lowest highest value). aNumber of measurements. bBetween referents and all workers. cBetween referents and net workers. dBetween sewage plant and net workers. P < 0.05. View Large The concentration of CRP was negatively associated with the lung function variables FEV1% predicted (β = −7.7, R2 = 0.05, P < 0.01; Fig. 1) and FVC% predicted (β = −8.5, R2 = 0.08, P < 0.01), adjusted for BMI, age, and smoking. Figure 1. View largeDownload slide Relationship between serum concentration of CRP in 27 referents (open diamond symbol), 46 sewage plant workers (filled round symbol), and 75 sewer net workers (filled triangle symbol) and FEV1 in % of predicted (β = −7.7, R2 = 0.05, P < 0.01). Figure 1. View largeDownload slide Relationship between serum concentration of CRP in 27 referents (open diamond symbol), 46 sewage plant workers (filled round symbol), and 75 sewer net workers (filled triangle symbol) and FEV1 in % of predicted (β = −7.7, R2 = 0.05, P < 0.01). Associations between exposure and health effects The associations between the concentration of biomarkers, lung function variables, and exposure were studied using linear regression analyses. The concentration of ICAM-1 was negatively associated with the estimated exposure to H2S (H2S index) (β = −19.9, R2 = 0.09, P < 0.05). This association was also significant for sewer net workers only (β = −52.6, R2 = 0.07, P < 0.05). The absolute value of FEV1 (3.39 l, ±0.7) was inversely associated with endotoxin exposure (β = −0.22, R2 = 0.18, P < 0.05), and this association was close to significant regarding FEV1% of predicted value (P = 0.07). No interaction between the exposure to endotoxin and H2S index was observed on the association with any of the analysed lung function variables. Discussion Exposure to endotoxins during the handling of sewage was higher when working at the sewage treatment plants compared to the sewer net system. The exposure to H2S was higher when working in the sewer net system compared to the sewage plants, and alarmingly high when collecting sewage from cesspools. A work-related weak inflammatory effect on the airways was observed as CRP, particularly among the sewer net workers, increased with impaired lung function. Exposure to endotoxin was associated with lower FEV1. Furthermore, exposure to H2S was associated with ICAM-1. Because exposure levels to H2S and endotoxin exposure were not correlated, the observed exposure-response associations found may be regarded as independent. Overall, the exposure level of endotoxins was moderate, both when working at the sewage plants or in the sewer net system. Season and flushing as work operations were the most important determinants to exposure to both endotoxins and H2S. The endotoxin exposure level measured at the sewage plants is comparable to other studies of WWWs (1–4000 EU m−3) (Smit et al., 2005; Spaan et al., 2008; Heldal et al., 2010; Cyprowski et al., 2015). However, the occurrence of high levels of endotoxin exposure as measured earlier in Norwegian sewage plants (3200 EU m−3) was not observed in this study (Heldal et al., 2010). The exposure during work in the sewer net system has been less studied. In a French exploratory study of sewer net workers, the endotoxin measurements ranged from 8 to 420 EU m−3 and were thus comparable to the present study (1–304 EU m−3) (Duquenne et al., 2014). However, the number of measurements in that study were few (n = 39) and the sampling devices were not worn by sewer workers but by an independent person close to the worker, possibly resulting in a lower exposure. The exposure to H2S was in general low when working in the sewage plants and sewer net system but peaked to high concentrations when handling sewage and transporting sludge (5–270 p.p.m.). This result is comparable with the study of Lee et al. (2007), where 140 workers at sewage plants and sewer net system in USA were exposed to peak levels from 42 to 122 p.p.m. during collection and dewatering of sludge. Respiratory symptoms among WWWs may often be associated with exposure to endotoxins, while CNS symptoms are known to be associated with exposure to H2S. High prevalence of symptoms, both from the airways and CNS, were observed in this study. There was no significant difference in the prevalence of self-reported symptoms between sewage plant and sewer net system workers, although the sewage plant workers tended to report symptoms more often than the sewer net workers. A high percentage of workers reported a feeling of fever and influenza-like symptoms after work among both groups of workers (35%), and the prevalence was higher than earlier reported among 44 sewage plant workers in Norway (0%) and 2% among 371 Dutch sewage workers (Smit et al., 2005). CNS-related symptoms such as tiredness (46%) and concentration difficulties (32%) were also more often reported compared to 141 sewage plant workers in a Dutch study (6 and 1%, respectively) (Douwes et al., 2001) but at similar prevalences among 59 Swedish sewage plant workers (31 and 10%, respectively) (Thorn et al., 2004). However, exposure to H2S was not measured in these studies. In a case report, several workers in the sewer net system developed airway symptoms such as cough and sore throat, in addition to influenza-like symptoms, which persisted several weeks after incidences of probable high H2S exposure (Watt et al., 1997). However, personal exposure to H2S was not measured in that study, and exposure was recorded subjectively based on the smell of gas. In the same case report, an impact on the lung function (FEV1) was observed among several workers up to 1 year after exposure to highly odorant gas when removing a broken pipe (Watt et al., 1997), but no associations between the self-reported symptoms and exposure to endotoxins or H2S were observed. A lower level of the lung-specific pneumoprotein CC16 and a close to significant higher level of the adhesion molecule ICAM-1 were observed among the subgroup of sewer net workers compared to the referents. This result may indicate early signs of both lung-specific and systemic inflammation, which has been previously reported among sewage plant workers exposed to sewage dust (Heldal et al., 2013, 2016). A lower concentration of CC16 compared to the reference suggested that chronic exposure may compromise the synthesis or secretion of the protein. In addition, a dose-response relationship between exposure to bacterial cells and CC16 measured in blood taken the same day as the exposure was measured for that individual indicated a more acute response in a previous study (Heldal et al., 2013). Among the same sewage plant workers, an association between sewage dust exposure and elevated serum concentration of the adhesion molecule ICAM were observed, suggesting that the increase in ICAM-1 may be initiated by a local inflammatory response in the pulmonary tissue caused by exposure to bacteria and dust (Heldal et al., 2016). In the present study, we found associations between estimated exposure to H2S and serum concentration of ICAM-1, indicating that H2S may also contribute to the observed inflammatory response. ICAM-1 could play a pivotal role in acute and possibly in chronic inflammation (Beck-Schimmer et al., 2002; Courmier and Israël-Assayag, 2004). Animal studies have shown that ICAM-1 contributes in the development of acute lung inflammation by adhesion of neutrophils to alveolar epithelial cells after instillation of bacterial lipopolysaccharide (LPS) in rats (Beck-Shimmer et al., 1997, 2002). The most plausible causative agent of the observed non-specific inflammatory reaction is endotoxins (Rylander and Jacobs, 1999; Thorn 2001). The elevated level of CRP observed in this study, particularly among sewer net workers, suggests a systemic inflammatory response, which may originate from the airways. In agreement with an earlier study among sewage plant workers (Heldal et al., 2016), we found that the elevated level of CRP was associated with impaired lung function among the WWWs. However, no association between exposure to endotoxins and levels of CRP was observed among the WWWs in previous studies. A study among 33 greenhouse workers showed that CRP levels positively associated with endotoxin exposure (Madsen et al., 2016). These workers were more exposed to endotoxins (median 66 EU m−3, range 1–3100 EU m−3) compared to our study. On the other hand, neither elevated levels of CRP (GM 1.7 mg l−1) compared to the control group (1.3 mg l−1) nor association to the exposure were found among grain workers (n = 67) exposed to even higher levels of endotoxins (GM 777 EU m−3; geometric standard deviation [GSD] 5,53) (Straumfors et al., 2018). We observed an association between the relatively low exposure level of endotoxins and the absolute value of FEV1, indicating that the exposure during handling sewage had an impact on lung function. An impairment of lung function variables (FEV1% predicted and FVC% predicted) compared to the referents were observed earlier among workers exposed to sewage dust (Heldal et al., 2010), however, with no association to the exposure. A significant impact on the lung function (FEV1) related to the relatively moderate exposure level to endotoxin (1–214 EU m−3) was also reported among 78 sewer workers in a Polish study (Cyprowski et al., 2015), at an exposure level of endotoxin comparable to the present study (1–340 EU m−3). In this study, we have investigated a presumed interaction between combined exposure to endotoxins and H2S on the lung function among WWWs. The inflammatory airway effects of endotoxins are well known (Thorn 2001; Heldal et al., 2013) as is the impact on lung function being related to H2S exposure (Richardson, 1995). However, no such interactions were observed in our study. Conclusion In summary, the results from this study demonstrate that sewer net workers are more exposed to incidental H2S peaks than workers at sewage plants and less exposed to endotoxins. However, despite moderate levels of endotoxin and H2S exposure, the results indicate an impact of these agents on lung function and the adhesion molecule ICAM-1, respectively. The results also signify a low-grade systemic inflammation shown by an increased level of CRP, particularly among workers at the sewer net system, a relation which was associated with an observed lower lung function among the workers. Exposure to dust and gases among sewer and sewage treatment plant workers should be reduced in order to prevent lowered respiratory function and negative health effects from inflammation. References Adelson L , Sunshine I . ( 1966 ) Fatal hydrogen sulfide intoxication. Report of three cases occurring in a sewer . Arch Pathol ; 81 : 375 – 80 . Google Scholar PubMed Austigard ÅD , Svendsen K , Heldal KK . ( 2018 ) Hydrogen sulphide exposure in waste water treatment . J Occup Med Toxicol ; 13 : 10 . Google Scholar Crossref Search ADS PubMed Beauchamp RO Jr , Bus JS , Popp JA et al. ( 1984 ) A critical review of the literature on hydrogen sulfide toxicity . Crit Rev Toxicol ; 13 : 25 – 97 . Google Scholar Crossref Search ADS PubMed Beck-Schimmer B , Madjdpour C , Kneller S et al. ( 2002 ) Role of alveolar epithelial ICAM-1 in lipopolysaccharide-induced lung inflammation . Eur Respir J ; 19 : 1142 – 50 . Google Scholar Crossref Search ADS PubMed Beck-Schimmer B , Schimmer RC , Warner RL et al. ( 1997 ) Expression of lung vascular and airway ICAM-1 after exposure to bacterial lipopolysaccharide . Am J Respir Cell Mol Biol ; 17 : 344 – 52 . Google Scholar Crossref Search ADS PubMed Cyprowski M , Sobala W , Buczyńska A et al. ( 2015 ) Endotoxin exposure and changes in short-term pulmonary function among sewage workers . Int J Occup Med Environ Health ; 28 : 803 – 11 . Google Scholar Crossref Search ADS PubMed Douwes J , Mannetje A , Heederik D . ( 2001 ) Work-related symptoms in sewage treatment workers . Ann Agric Environ Med ; 8 : 39 – 45 . Google Scholar PubMed Douwes J , Versloot P , Hollander A et al. ( 1995 ) Influence of various dust sampling and extraction methods on the measurement of airborne endotoxin . Appl Environ Microbiol ; 61 : 1763 – 9 . Google Scholar PubMed Duquenne P , Ambroise D , Görner P et al. ( 2014 ) Exposure to airborne endotoxins among sewer workers: an exploratory study . Ann Occup Hyg ; 58 : 283 – 93 . Google Scholar PubMed European Respiratory Society (ERS). ( 1993 ) Standard lung function testing . Eur Respir J ; 6( Suppl. 16 ): 25 – 32 . Farahat SA , Kishk NA . ( 2010 ) Cognitive functions changes among Egyptian sewage network workers . Toxicol Ind Hyealth ; 26 : 229 – 38 . Google Scholar Crossref Search ADS Friesen MC , Demers PA , Spinelli JJ et al. ( 2006 ) From expert-based to quantitative retrospective exposure assessment at a Soderberg aluminum smelter . Ann Occup Hyg ; 50 : 359 – 70 . Google Scholar PubMed Heldal KK , Barregard L , Ellingsen DG . ( 2016 ) Biomarkers of inflammation in workers exposed to compost and sewage dust . Int Arch Occup Environ Health ; 89 : 711 – 8 . Google Scholar Crossref Search ADS PubMed Heldal KK , Barregard L , Larsson P et al. ( 2013 ) Pneumoproteins in sewage workers exposed to sewage dust . Int Arch Occup Environ Health ; 86 : 65 – 70 . Google Scholar Crossref Search ADS PubMed Heldal KK , Madsø L , Huser PO et al. ( 2010 ) Exposure, symptoms and airway inflammation among sewage workers . Ann Agric Environ Med ; 17 : 263 – 8 . Google Scholar PubMed Lee JA , Thorne PS , Reynolds SJ et al. ( 2007 ) Monitoring risks in association with exposure levels among wastewater treatment plant workers . J Occup Environ Med ; 49 : 1235 – 48 . Google Scholar Crossref Search ADS PubMed Lewis RJ , Copley GB . ( 2015 ) Chronic low-level hydrogen sulfide exposure and potential effects on human health: a review of the epidemiological evidence . Crit Rev Toxicol ; 45 : 93 – 123 . Google Scholar Crossref Search ADS PubMed Lim E , Mbowe O , Lee AS et al. ( 2016 ) Effect of environmental exposure to hydrogen sulfide on central nervous system and respiratory function: a systematic review of human studies . Int J Occup Environ Health ; 22 : 80 – 90 . Google Scholar Crossref Search ADS PubMed Lundholm M , Rylander R . ( 1983 ) Work related symptoms among sewage workers . Br J Ind Med ; 40 : 325 – 9 . Google Scholar PubMed Madsen AM , Thilsing T , Bælum J et al. ( 2016 ) Occupational exposure levels of bioaerosol components are associated with serum levels of the acute phase protein Serum Amyloid A in greenhouse workers . Environ Health ; 15 : 9 . Google Scholar Crossref Search ADS PubMed Melbostad E , Eduard W . ( 2001 ) Organic dust-related respiratory and eye irritation in Norwegian farmers . Am J Ind Med ; 20 : 690 – 2 . Melbostad E , Eduard W , Skogstad A et al. ( 1994 ) Exposure to bacterial aerosols and work-related symptoms in sewage workers . Am J Ind Med ; 25 : 59 – 63 . Google Scholar Crossref Search ADS PubMed Miller MR , Hankinson J , Brusasco V et al. ( 2005 ) ATS/ERS Task Force. Standardisation of Spirometry . Eur Retspir J ; 26 : 319 – 38 . Google Scholar Crossref Search ADS NEG . ( 2001 ) Hydrogen Sulphide, In Arbete och Hälsa, TNEGfCDoHrf Chemicals, Editor . Stockholm, Sweden : National Institute of Working Life . Neumann HD , Buxtrup M , Balfanz J et al. ( 2002 ) Belastungen durch biologische Asrbeitsstoffe bei der Kanalreinigung . Gefahrst Reinhalt Luft ; 9 : 371 – 80 . Quanjer PH , Tammeling GJ , Cotes JE et al. ( 1993 ) Lung volumes and forced ventilatory flows . Eur Respir J ; 6 ( Suppl. 16 ): 5 – 40 . Google Scholar Crossref Search ADS PubMed Richardson DB . ( 1995 ) Respiratory effects of chronic hydrogen sulfide exposure . Am J Ind Med ; 28 : 99 – 108 . Google Scholar Crossref Search ADS PubMed Rylander R . ( 1999 ) Health effects among workers in sewage treatment plants . Occup Environ Med ; 56 : 354 – 7 . Google Scholar Crossref Search ADS PubMed Smit LA , Spaan S , Heederik D . ( 2005 ) Endotoxin exposure and symptoms in wastewater treatment workers . Am J Ind Med ; 48 : 30 – 9 . Google Scholar Crossref Search ADS PubMed Spaan S , Smit LA , Eduard W et al. ( 2008 ) Endotoxin exposure in sewage treatment workers: investigation of exposure variability and comparison of analytical techniques . Ann Agric Environ Med ; 15 : 251 – 61 . Google Scholar PubMed Straumfors A , Eduard W , Heldal KK et al. ( 2018 ) Pneumoproteins and markers of inflammation and platelet activation in the blood of grain dust exposed workers . Biomarkers ; 23 : 748 – 55 . Google Scholar Crossref Search ADS PubMed Thorn J . ( 2001 ) The inflammatory response in humans after inhalation of bacterial endotoxin: a review . Inflamm Res ; 50 : 254 – 61 . Google Scholar Crossref Search ADS PubMed Thorn J , Beijer L . ( 2004 ) Work-related symptoms and inflammation among sewage plant operatives . Int J Occup Environ Health ; 10 : 84 – 9 . Google Scholar Crossref Search ADS PubMed Thorn J , Beijer L , Rylander R . ( 2002 ) Work related symptoms among sewage workers: a nationwide survey in Sweden . Occup Environ Med ; 59 : 562 – 6 . Google Scholar Crossref Search ADS PubMed Tvedt B , Skyberg K , Aaserud O et al. ( 1991 ) Brain damage caused by hydrogen sulfide: a follow-up study of six patients . Am J Ind Med ; 20 : 91 – 101 . Google Scholar Crossref Search ADS PubMed van der Wal JF . ( 1983 ) Comparative measurements of the total dust concentration at the work place with different samplers - part 1 . Staub-Reinhalt, Luft ; 43 : 292 – 4 [in German]. Watt MM , Watt SJ , Seaton A . ( 1997 ) Episode of toxic gas exposure in sewer workers . Occup Environ Med ; 54 : 277 – 80 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2019. Published by Oxford University Press on behalf of the British Occupational Hygiene Society. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Endotoxin and Hydrogen Sulphide Exposure and Effects on the Airways Among Waste Water Workers in Sewage Treatment Plants and Sewer Net System
JF - Annals of Work Exposures and Health (formerly Annals Of Occupational Hygiene)
DO - 10.1093/annweh/wxz020
DA - 2019-04-19
UR - https://www.deepdyve.com/lp/oxford-university-press/endotoxin-and-hydrogen-sulphide-exposure-and-effects-on-the-airways-MUIcWR7TVH
SP - 437
VL - 63
IS - 4
DP - DeepDyve
ER -