Introduction

Sewage workers employed at wastewater treatment plants (WWTPs) are potentially exposed to a wide variety of organic and inorganic pollutants, to specific hazardous chemical agents used or produced in the treatment process, to human and animal infectious (bacteria, viruses, protozoa, helminths and fungi) and non-infectious (eg, endotoxins) biological agents.1 Waterborne and airborne exposure varies from hour-to-hour, day-to-day and season-to-season, and may vary depending upon the geographical location, the type and the capacity of the treatment facility, the performed activities at different locations within a WWTP, and the weather conditions.2–5 The worker's health can be endangered by inhalation, ingestion by hand-to-mouth contact (faeco-oral route), or skin absorption of the above-mentioned harmful agents. The work-related health problems of sewage workers have been reviewed previously: most frequently reported symptoms are gastrointestinal issues, fever, headache, dizziness and fatigue, while respiratory issues, skin rash, and irritation of eyes, nose and throat are also common.3 6–8 Abdominal symptoms are found in excess among sewage workers.7 9–13 The exact causes of these health problems are still not very well known. The gastrointestinal and respiratory symptoms may be related to the exposure to endotoxins from Gram-negative bacteria.3 The process of wastewater treatment does not necessarily destroy the micro-organisms; potential pathogenic or opportunistic micro-organisms can survive (or multiply) in sewage, sludge, or aerosolised air.3 6 7 The risk of contracting infectious diseases is an issue still under debate.14 15 There are studies reporting an increased prevalence of antibodies against hepatitis A, and risks of illness caused by Leptospira, Entamoeba histolytica and Giardia lamblia have been reported among sewage workers.2 3 6 16–18 A particular form of illness, probably of viral origin, characterised by general malaise, weakness, acute rhinitis and a fever called ‘sewage worker's syndrome’ was reported in 1976 by Rylander et al.19 In a list of potential and actual biological-related health risks of the wastewater industry published in 1987, Helicobacter pylori was not mentioned.17

Since 1983, when Warren and Marshall successfully cultured H pylori, this Gram-negative bacteria has been recognised as a major cause of several gastroduodenal diseases, including gastric and duodenal peptic ulcers, chronic gastritis and atrophic gastritis. Infection with H pylori has been associated with an increased risk of gastric adenocarcinoma and mucosa-associated lymphoid tissue (MALT) gastric lymphoma. The International Agency for Research on Cancer (IARC) has classified H pylori among the class I carcinogens. Various pathways of transmission are documented, favouring person-to-person mode of transmission mostly early in life. Faeco-oral, oro-oral and gastro-oral transmission routes have all been proposed, but waterborne transmission has also been suggested.20–24 If introduced into an aqueous environment, H pylori can survive for periods of time, depending upon a number of specific conditions of the environment (eg, temperature, pH level, chlorination).20 25 Detection of the bacteria in pretreated and post-treated wastewater suggests that wastewater treatment may be ineffective in removing H pylori from this aqueous environment.21–23

Studies from Sweden, Switzerland and Singapore have failed to show a higher prevalence of H pylori antibodies among sewage workers, but sample sizes used were small, the workers had different socioeconomic backgrounds, and the exposure data were not clearly defined.26–28

The Flemish wastewater treatment infrastructure consists of over 200 WWTPs, over 850 pumping stations and over 4000 km of sewers in operation with a combined treatment capacity of over 3 million m³/day. Since 1990 the supramunicipal part of the wastewater treatment is provided by the company Aquafin NV, which also has a role in the management and maintenance of the municipal wastewater treatment infrastructure for the Flemish Region of Belgium.

A preliminary study was performed in 2001 among 404 sewage and/or sludge exposed and non-exposed employees of this company, using a health and hygiene questionnaire. This study showed that there was a higher prevalence of abdominal symptoms, oesophagogastroduodenal endoscopies and peptic ulcers among sewage-exposed workers than among non-sewage exposed employees working for the same employer.29

These findings, together with the reports of frequent gastrointestinal symptoms (although H pylori infection is most often asymptomatic), and the evidence of the presence of viable but not culturable coccoid forms and DNA of H pylori in sewage,20 prompted us to study the potential occupational risk for H pylori among sewage workers. The hypothesis tested is whether WWTP workers are at risk to be infected by H pylori at work.

Materials and methods

Subjects

In the period September 2004 to June 2005, 431 employees of Belgian (Flemish) WWTPs scheduled for the annual examination by the occupational health doctor, were invited to take part in this study, of which 317 employees agreed to participate voluntarily (73.5%). Written informed consent was obtained from each subject.

Control group

Laboratory personnel, operators and maintenance workers of a Flemish pharmaceutical company with no occupational exposure to sewage, were chosen as control group because age structure, socioeconomic status and educational level, were expected to be comparable. Socioeconomic background is considered as one major determinant of H pylori seropositivity. Written informed consent was obtained.

Questionnaire

All participants received a structured questionnaire. Information was collected on: (1) demographic variables (gender, age, place of birth, educational level); (2) working history (current and previous jobs reflecting exposure to sewage or sludge; current and previous exposure (different tasks); compliance with hygienic measures); (3) (possible) risk factors for H pylori infection outside work (journeys to a tropical destination the last 10 years, number of persons living in same house or sharing the same bedroom during childhood); and (4) presence of gastrointestinal symptoms (during the last 3 months, history of peptic ulcer, oesophagogastroduodenoscopy, or eradication treatment for H pylori, family history of H pylori). Compliance with hygienic measures was measured by 23 questions, using a scoring system from 1 to 5 for every question (1 for least compliant and 5 for very compliant). The total score was the mean score on the 23 questions, and was expressed as the ‘hygienic index’ (HI). The different, shorter questionnaire for the control group included: (1) demographic variables; (2) working history (about current and previous jobs reflecting exposure to sewage or sludge; (3) (possible) risk factors for H pylori infection outside work; and (4) presence of gastrointestinal symptoms.

Serology

A blood sample for serum analysis was collected from each sitting, non-fasting worker by routine venopuncture at the time of the annual medical examination. The serum samples were stored after centrifugation at −20°C and analysed in batches. The sera were tested by ELISA for IgG antibodies using the Enzygnost Anti-Helicobacter II/IgG (Behring Diagnostics, Marburg, Germany) according to the manufacturer's instructions. A subject was considered as seropositive if IgG was ≥10 U/ml. The sensitivity and specificity of this full antigen ELISA are 85% and 90% respectively.30

Serology (ELISA test) was validated in a pilot study among 92 Belgian sewage workers of WWTPs with two other non-invasive diagnostic tests: faecal antigen test Premier Platinum (PP) HpSA (Meridian Diagnostics-Bioscience, Cincinnati, Ohio, USA) and ¹³C urea breath test (UBT).31 Validity of serology was consistent with published performance results.30

Statistical analyses

Data were analysed using the SPSS statistical software package (SPSS, Chicago, Illinois, USA) for Macintosh (V.10.0). The differences in the prevalence of H pylori infection and symptoms between sewage workers and controls were assessed using the Pearson χ² test or the Fisher exact test. A two-tailed t test was used to assess a difference in the means of the groups. The Mann–Whitney U test was used for non-normally distributed variables. The multiple logistic regression analyses were performed to compute adjusted odds ratios (OR) with 95% CI while controlling for variables with p<0.100. Age groups and variables associated with occupational exposure were entered into a binary multiple logistic regression analysis (backward stepwise likelihood ratio) with H pylori infection status as the dependent variable. The reference category for OR estimates was no or minimal occupational exposure. A p value less than 0.05 was considered to be significant for all the analyses. The CIs were calculated with Confidence Interval Analysis (CIA) software, which was provided with the book Statistics with Confidence.32

Approval

The study was approved by the management and the Committee for Prevention and Protection at Work of the employer, to minimise the non-response. The study protocol was also approved by the ethical committee of IDEWE Occupational Health Services (OG no. 117).

Results

The main demographic characteristics of the study group and the control group are given in table 1.

Study group

The mean (±SD) age of the study group was 39.67±8.31 years (range 20–61 years). A total of 12 of the participants (3.8%) were women. A total of 70.8% of the subjects were WWTP operators, 11.9% were maintenance workers and 17.3% worked at the central laboratory or performed other jobs at WWTPs. The mean age of operators was 39.16±8.21 years, maintenance workers 41.97±7.76 years, laboratory workers 33.33±9.39 years and for other jobs 41.83±8.81 years. Laboratory workers were younger, but not significantly, than the other job groups (p=0.21). Most women sewage workers were employed in the laboratory. The educational level of WWTP workers was significantly different than that of control group, mainly due to a higher proportion of technical educated workers (p<0.001). All laboratory workers, except one, and nearly half of staff with other jobs in WWTP hold university degrees or have had graduate school training. Nearly 75% of the operators, two-thirds of the maintenance workers, and nearly half of the workers with other jobs at WWTP, hold a technical degree. Four age groups were defined: <30 years (12.3%), 30–39 years (37.2%), 40–49 years (38.2%) and ≥50 years (12.3%). The mean period of exposure to sewage and/or sludge was 10.31±6.75 years. In all, 22.1% had one or more holiday journeys to a destination in Africa, Asia, or Latin America the last 10 years. All participants were born in Europe.

Control group

Participation in the control group was 99%. This group (n=250) was younger (mean age 36.07±10.86 years with range 19–67 years) and contained less men (53.6%). Gender has not been reported to be important in this context. Operators were younger with a mean age (±SD) of 34.71±10.84 years, 24 maintenance workers had a mean age of 43.13±9.34 years, and 26 persons with other jobs had a mean age of 40.16±8.81 years. The educational level of the control group was significantly different to the study group: fewer people had a technical degree but more had a humanities degree. The educational level of father was significant different (p=0.004), but the educational level of mother (as a proxy for socioeconomic status) was comparable with the study group. A total of 2% were born outside Europe.

Gastrointestinal symptoms and peptic ulcers

The prevalence of gastrointestinal symptoms (pyrosis, stomach ache, abdominal pain, dyspepsia, ructus) during the last 3 months, was higher among the sewage workers (37%) than among the control group (14.7%) (p=0.000).

The operators and maintenance workers of the WWTPs had a higher prevalence of gastrointestinal symptoms (39%) than laboratory workers and persons performing other job at WWTPs (23%) (p=0.186).

More sewage workers had been treated for peptic ulcer: 22 (6.9%) versus 12 persons of the control group (4.8%), but this difference was not statistically significant (p=0.162).

H pylori serology

Table 2 presents the overall H pylori IgG seroprevalence among sewage workers compared to the control group. A total of 16.7% of sewage workers (95% CI 12.6 to 20.8%) and 13.6% of controls (95% CI 9.3 to 17.9%) had a seropositive H pylori status, resulting in a crude Prevalence OR of 1.23 (95% CI 0.78 to 1.95).

The differences between the age-specific seroprevalence were not statistically significantly different (table 2).

Seroprevalence for H pylori increases with age in both groups (table 2). Mean age was significantly higher in H pylori-positive study group (42.8 years vs 39.1 years; p<0.003), as in the H pylori-positive control group (42.0 years vs 35.1 years; p<0.001). Statistical significant associations were found between a seropositive IgG status and lower educational level of father and mother in sewage workers group (p=0.01 and p=0.04 respectively), and also in the control group (p=0.01 and p<0.001 respectively) (data not shown). Also, after direct standardisation for age, no significant difference in seroprevalence was noted between exposed workers and controls (16.7% among study group versus 14.8% among control group).

Adjusting for the confounding effect of age, educational level of workers and both parents (covariates with a p value <0.10) resulted in an adjusted OR of 1.02 (95% CI 0.58 to 1.80) (p=0.93).

A history of gastrointestinal symptoms during the last 3 months according to serostatus among sewage workers and control group is shown in table 3. In both groups no statistical significant associations were found between H pylori prevalence and gastrointestinal symptoms, or with treatment for peptic ulcer in the past.

The evaluation of occupational risk factors for H pylori infection was performed based on the self-reported exposure (regularly, sometimes, seldom or never) at different workstations of WWTP (mechanical treatment, biological treatment, sewage sludge treatment, operational tasks and maintenance tasks). The highest emissions of bioaerosols are found at the sites of pretreatment and the primary clarifiers and moving mechanical equipments for water aeration.33 34

Table 4 shows that no statistically significant associations were demonstrated between the prevalence of H pylori infection and any job task or occupational exposure.

No significant association was found between HI and H pylori seroprevalence status. Mean score (±SD) of HI among H pylori seropositive workers was 3.64 (±0.52), versus 3.59 (±0.49) among seronegatives.

In order to explore the independent contribution of each significant occupational exposure variable on infection of H pylori, we used multiple logistic regression analysis to adjust the ORs for the confounding effect of age, job group, years of exposure, occupational exposure at the job, and results of the hygiene index (table 5).

In this analysis apart from age, no occupational exposures or self-reported compliance with hygienic measures showed an increased risk for H pylori infection.

Discussion

We hypothesised that sewage workers of WWTPs could have a work-related risk of H pylori infection and tested this hypothesis by comparing the seroprevalence with employees working in the pharmaceutical industry. Because we also assessed the self-reported exposure and hygienic measures, we could also compare groups with higher exposure in the group of sewage workers.

We did not find a higher seroprevalence of H pylori infection in sewage workers compared to non-exposed controls, which is consistent with previous studies among sewage workers at WWTPs in Sweden, Singapore and Switzerland.26–28 The overall prevalence of H pylori IgG antibodies among Swedish sewage workers did not differ from that of the 138 referents: adjusted OR was 0.90 (95% CI 0.53 to 1.5).26 In a study (1996–1997) among 416 sewage workers in Singapore the prevalence of H pylori IgG antibodies did not differ between sewage exposed (29.3%) and non-sewage exposed personnel (29.7%).27 In a prospective cohort study among sewage exposed municipal workers in Switzerland no occupational risk of acquiring H pylori infection was found during the first study year.28

We compared WWTP sewage workers with pharmaceutical workers. By choosing the control group, we expected same socioeconomic status. Unfortunately, educational levels as a proxy for socioeconomic status, were different. Educational level of the sewage workers was significantly higher. Educational level of mother was comparable, but educational level of fathers was significant different between the study group and the control group.

The importance of socioeconomic conditions on the acquisition of H pylori infection has been confirmed in a number of population-based studies. Social class factors have been associated with H pylori infection status. The infection is typically acquired in childhood or when children are in the family. Childhood socioeconomic conditions (social class) and childhood crowded living conditions (overcrowding) are known risk factors. Adult can be infected with bacteria, this has been reported to occur as a rate of around 0.3% to 0.5% per year.35

Persons with H pylori seropositive status were longer employed at WWTPs. Probably because of the strong correlation between years of employment at WWTPs and age (correlation coefficient=0.5; p<0.001). Sewage workers group is known to be job stable, with a low turnover.

We found no association between any self-reported gastrointestinal symptoms and H pylori infection as measured by serology test.

Our study found a higher prevalence of gastrointestinal symptoms and peptic ulcers among sewage workers at WWTPs than among pharmaceutical control group. In most individuals H pylori infection is asymptomatic. However, about 10% to 15% of infected individuals will some time experience peptic ulcer disease.

This study is potentially subject to differential and non-differential information bias because of use of self-reported data for gastrointestinal symptoms and exposure. The WWTP workers may be more prone to report their symptoms (recall bias), especially because the symptoms have been discussed in association with working with sewage in a health and hygiene questionnaire (2001), and a pilot study to evaluate three non-invasive diagnostic tests (2002–2003).29 31 There is a possibility that over-reporting to some extent can be responsible for the differences, although most studies have shown that gastrointestinal symptoms (stomach ache, nausea) are more common among WWTP workers.1 3 7 9

Friis et al also reported a higher prevalence of peptic ulcers among Swedish sewage workers (3.5% vs 2.2% among control group).12 But Jeggli et al found among Swiss sewage workers a peptic ulcer prevalence of 3.2%, and a histologically verified gastritis prevalence of 2.9%.28 The prevalence was not higher than among non-sewage exposed workers of which 3.5% had a history of peptic ulcer, and 2.8% had verified gastritis in the past. H pylori eradication did not occur more often in Swiss sewage workers.28

The fact that the results of the logistic regression analyses did not pointed out an increased risk for H pylori infection in one work area and compliance with the hygienic measures, suggest that there is no risk of H pylori infection in WWTP workers.

Recent studies have shown that H pylori can survive up to 96 h in water, showing that waterborne transmission is a real possibility.25 It has also been hypothesised that environmental reservoirs, like water, would act by providing a pool of genetic material during transient passages through the gastrointestinal tract.24 These passages could explain the high prevalence of gastrointestinal issues (including peptic ulcers) in our study in the exposed subjects compared to non-exposed persons.

These transient passages could also explain why the seroprevalence of H pylori is not different in exposed and non-exposed groups: these episodes are either too short for the infected person to seroconvert or, alternatively, the seroreversion rate is higher than the seroconversion rate; both mechanisms could play a role and reinforce each other. Seroconversion and seroreversion have been shown to occur at a rate of around 0.3% to 0.5% per year in young adults.35

The wastewater treatment plant workers are exposed to primarily Gram-negative bacteria via ingestion of liquid waste, hand-to-mouth contamination or inhalation of aerosolised micro-organisms. It is likely that the large variety of symptoms expressed by sewage workers are caused by diverse and several agents,1 some of which were probably not measured in our study: unknown microbiological agent(s) with potentially immunotoxic and/or allergenic properties or toxic chemical agent(s). Recently, a Norwegian study found no higher Salmonella seroprevalence among sewage workers.15 Gastrointestinal symptoms have been reported in connection with exposure to Gram-negative bacteria or endotoxins produced by a number of Gram-negative rods.9 Bacterial endotoxins, heat-stable proteins that are toxin specific for the cells of the intestinal mucosa, are present in the sewage environment and have been linked to the onset of gastrointestinal and also other symptoms (flu-like and respiratory) among sewage workers.2 7 9 11 13 37 38 High levels of exposure to airborne bacteria and endotoxins, were related to certain phases of the treatment process in studies in Finland (1994) and Sweden (1999).2 7 Laitinen et al (1994) and Thorn et al (2002) conclude that ambient bacteria and endotoxin exposure in sewage plants is complex, and level of exposure depends on activities.2 4 Prazmo et al (2003) found that limit values for bacteria and endotoxins were not exceeded in Polish WWTP.36 In a Dutch study by Smit et al (2005), endotoxins seem to play a causal role for wide range of work-related symptoms.37 Nausea and vomiting are normally not associated with endotoxin exposure.1 The study of Melbostad et al in Norway showed that workers with higher exposure levels to rod-shaped bacteria and total bacteria were reporting more respiratory symptoms as well as heachache, tiredness and nausea. But no such association was found for the exposure to endotoxins.10 Exposure to endotoxins could have been underestimated because sampling methods were not designed for outdoor sampling.1

Further investigations and field studies are necessary to determine possible other causes of the reported symptoms among sewage workers. Some of the symptoms may be preventable, eg, by introducing relatively simple workplace hygienic measures1 or immunisation(s). Use of respiratory protection equipment, rubber gloves and protective clothes are recommended.39 Hansen et al40 concluded that for preventive purposes, the worker's exposure to sewage and sludge should by minimised, an objective that may be achieved by giving information to the workers combined with a thorough hygienic scrutiny of every installation and work procedure. The use of company clothes that were washed daily was negatively associated with flu-like symptoms suggesting that good hygienic practice at the workplace may prevent some of these symptoms.1

Conclusions

In our study of sewage workers employed at wastewater treatment plants in Flanders (Belgium) the seroprevalence of H pylori was not different among sewage workers compared to non-exposed controls. However, building on hypothesis of Azevedo et al on transmission of H pylori, there is a possibility that a transient passage of H pylori is a cause of the increased prevalence of gastrointestinal symptoms among WWTP workers. Further investigations are necessary to find other specific occupational cause(s) of the more prevalent gastrointestinal symptoms among WWTP sewage workers.