Task 1 – Desk-based literature review of the human health impacts of spreading sewage sludge on non-agricultural land
August 2008
Project partners: SNIFFER, SEPA, NIEA, Scottish Government, Forestry Commission, Health Protection Scotland

Use of the report
The technical report has been developed through a collaborative project, managed and facilitated by SNIFFER and has involved the members and partners. It provides background information, within the confines of the project brief, to support and inform member organisations and others.
Whilst the document is considered to represent the best available scientific information and expert opinion available to the consultant at the stage of completion of the report, within the confines of the specification given, it does not represent the final or policy positions of SNIFFER or any of its partner agencies, and it recognises that the historic practices regarding sewage sludge recycling discussed are not current practice within the UK.

Background to research

Although there is a great deal of research and scientific data on sewage sludge application to land, much of this relates to modest application rates on agricultural land.  In recent years there has been a substantial increase in the amount applied to forestry, and to former opencast coal sites in the UK for purposes of land restoration.  Application rates of sewage sludge have been considerably higher than traditionally practiced, and sewage sludge has been applied using different techniques.  There is concern that poorly managed practices could result in risks to human health, water, air and soil quality and biodiversity.  Public and political interest is high and this project will address the urgent need to review this activity and develop decision support systems and guidelines to ensure that the activities will not affect public health or adversely affect the environment.  The report outlines the findings of the detailed desk-based literature review carried out relating to potential health impacts.

Objectives of research

In detail, the project aims to:
The objective of the report is to fulfil the first aim in relation to the human health effects of spreading sewage sludge on non-agricultural land. However, the literature on the health status of sewage exposed populations proved to be small and the majority of studies are of occupationally exposed populations.  Consequently the original approach to this task has been limited and only broad estimates of risk have been achieved at this first stage.

Key findings and recommendations

Contents of sewage

  1. There is a variable literature on the content of sewage sludge which might inform on potential health effects.  The units of measurement vary across studies making direct comparisons difficult.  In addition, some comes from abroad which may not be directly usable in the UK setting.  However, some broad comments can be made.
  2. Untreated sewage sludge is capable of harbouring sometimes high levels of a wide range of bacteria (including drug-resistant forms), viruses and parasites and while differing forms of sewage treatment reduces levels of each of these, the effect is of varying degree depending on factors such as initial concentrations and resistance of the organism to treatment.
  3. Chief amongst these organisms, with respect to risk to humans, are Salmonella spp, E. coli, Campylobacter, Giardia and Ascaris.  Hepatitis viruses can be found in sewage as do a range of other viruses if looked for, and it is likely that intensive searching for specific organism may often result in positive finds even if at low levels which may not be important in terms of risk to health.
  4. Sewage sludge is a significant source of cadmium and lead and to a lesser extent mercury and arsenic.  Many chemicals, including a wide range of pharmaceutical agents, can be found in sewage at variable concentrations.  Treatment of sewage can result in dilution or concentration of these substances depending on their chemical interactions and properties.  It is difficult to identify a “best treatment” which would reduce chemical content across the board.
  5. While radioactivity can be detected in sewage sludge (from a range of sources including medical treatment and from industrial sources) levels are likely to be low although no repeated data on content in sewage sludge could be identified in the time available for this project.
  6. Treatment processes reduce the level of pathogenic micro-organisms (including drug-resistant strains) and PAHs in sewage sludge only partially, and the pathogenic organisms that survive treatment are often of human origin.  However, natural attenuation in soil of organisms from applied sewage sludge can occur although to a variable extent where different factors (e.g. site) will affect persistence.  Clear identification of surviving organisms derived from non-human animals has not been clearly defined.

  7. Exposure

  8. A key step in understanding and evaluating the risks posed by the application of sewage sludge to land involves creating a framework to identify the exposure compartments and transfer processes that have the potential to bring about contact between humans and the hazardous materials present in the sewage sludge.
  9. We have created a basic compartmental model to describe the potential for exposure among three population groups: those occupationally exposed during the application process; the local population that live close to the sewage application area; and recreational users of the land area at a time-point after application.  Although there are many problems in extrapolating from occupational to bystander exposures, because of the lack of literature in the latter situation, it is both necessary and inevitable that the occupational setting be considered in this report especially as most of the health literature on sewage sludge is from occupationally exposed populations.
  10. The three main exposure pathways are by inhalation, ingestion and trans-dermally.  Of these, in the context of sewage sludge to populations exposed non-occupationally, potentially the most important is the ingested route either directly, from water run-off or from food grown on sewage treated soil.  These are theoretical concepts because there are no data from sewage application to land which would allow estimation of risks.

  11. Occupational exposures

  12. Many studies in this area, for both occupationally and non-occupationally exposed populations, suffer from small sample sizes which is predictable as many studies were undertaken in response to a real or perceived problem in a defined population.  This limits the ability of such studies to identify small levels of risk in the exposed populations and thus runs into the problem of believing that effects do not occur whereas they do but at low level.  The only way this can be dealt with is by larger studies which might need to incorporate multiple populations with similar exposures, with adequate measures of exposure.  With these caveats some broad conclusions can be made on the literature available but it is not possible to identify sufficient information to derive quantitative estimates of risk.
  13. Sewage workers experience a wide range of exposure to airborne endotoxin, a pro-inflammatory molecule produced from bacterial degradation.  Flu-like illness and lower respiratory tract symptoms are associated with endotoxin exposure, sometimes with high odds ratios, one study showing a dose response relationship between exposure and symptoms.  
  14. The literature on the prevalence and incidence of symptoms in sewage treatment plant workers suggests a wide variety of health complaints, with significant odds ratios in the range of 2.2 to 9.4 which provides strong evidence of an effect on health in these workers.  The commonest are gastrointestinal symptoms, diseases of the airways and skin problems.  However, methods of data collection for the presence of symptoms varies, notably in terms of time frame (e.g. ever/never or during a certain time before the investigation).  There is no clear evidence of effects on lung function.  

    In some studies no effects on any health outcome were detected which might represent a true effect, perhaps due to specific work practices or differing relevant exposures, but more likely is due to reporting bias with subjects underreporting symptoms for fear of loss of employment especially if they had not been blinded to the objective of the study.  This is a reasonable conclusion given the high odds ratios found in other studies.

  15. Studies on infections and infestations in association with exposure to sewage show that workers exposed to raw sewage are at most risk.  The risk of infection is largely dependent on the prevalence of infections in the populations served by the treatment plants and on the treatment processes employed. Hepatitis A and giardiasis are the best recognised infections, but the association between hepatitis B markers and sewage exposure suggests that waste water workers should be vaccinated against both hepatitis A and B.  

  16. The limited evidence suggests there is potential for the development of cancer among sewage exposed workers, in particular urothelial tumours and primary liver cancer.  The findings of increased levels of urinary mutagens among sewage workers provides some exposure based support for this conclusion.  

  17. There are no data sufficient which enables quantitative risk assessment of health risks from occupational exposure to sewage (with no studies on sewage applicators) and, in addition, no data which would allow assessment of the presence or otherwise of thresholds of effect for specific agents.  There are no studies where measured exposures were compared to measured outcomes where both are numerically quantified.

  18. Non-occupational exposures

  19. Non-occupational exposures, either as residents or as by-standers, have been studied in a limited way and mostly in countries very different from Scotland and Northern Ireland (e.g. China, Kuwait, India). This literature suggests risks for HEV infection and perhaps Salmonellae and inhaled viruses but with more clear associations with Giardia and worm infestations.   The literature does, however, contain one formal trial of exposure (Dorn, et al.1985) which showed no increases in markers of ill health in the exposed population although the odds ratios were generally positive.  This suggests that if there is an effect, larger studies would need to be conducted although the content of the sewage sludge, poor application approaches and high application rates could result in a more clear cut risk.  There is no strong evidence for chemicals causing ill effects in these populations.

  20. In the majority of situations the infective dose of a pathogen cannot be stated as a precise number of pathogens guaranteed to cause disease in the host, although in some outbreaks of E. coli O157 as few as 10 to 100 organisms caused disease.  However, higher exposures may not always result in disease.  Resistance or immune responses to any pathogen vary from person to person depending on, for example, the current health of the person or their antibody levels to the pathogen under consideration i.e. susceptible individuals or those with compromised health. Further, strain to strain variability in pathogenicity exists.

  21. Overall, this limited literature is conflicting but shows no consistent effect on health from living near sewage treated land.  The Dorn study is the most rigorous and, being negative, generally supports this view. However, to address the issue more completely would require substantial research investment, requiring better formulated studies involving multiple sites over time to capture both short and long term effects.  A shorter term study to capture the risk of infections and infestations would bring results within about three years but would likely be very costly.

  22. There is theoretically a risk from exposure to foods grown on land subjected to sewage sludge application particularly from those foods eaten raw with no chance of cooking destroying bacterial or other pathogen content.  However, again, the literature gives no clear estimate of likely risk other than in the broadest terms and much of this relies on knowledge of what infections might result from known exposures in other settings.  However, some pathogens can be taken up by plants through their roots (e.g. salad crops eaten uncooked) and this might be of wider concern than health impacts from local populations if shown to be a real risk for the UK setting.

  23. Odour

  24. Odour is the main complaint of populations non-occupationally exposed to sewage sludge.  Ambient odours may produce health effects by direct irritation and/or toxicological routes, through responses which genetically coded or learned aversions or may be due to co-pollutant exposure(s) that is part of the odorant mixture.  It is also generally accepted that unpleasant odours act as warning signs or as indicators of potential risk to health but not as direct triggers to ill health.  

  25. The nose is a very sensitive organ being able to detect very low concentrations of some aromatic compounds.  The fact that a smell can be detected does not necessarily mean that the cause of that smell is adverse in terms of inducing an organic response in an individual.  But where there is a perception that the source of the smell is potentially toxic, the fact that a smell is detected is often taken by the public as proof that the cause of the odour is truly toxic.  Separating out these perceived effects from genuine effects can be impossible.  

  26. With respect to sewage sludge, four main groups of compounds contribute to the characteristic odours: sulphurous compounds, nitrogenous compounds, acids and aldehydes and ketones.  Some of these at high concentrations have the potential to cause health effects but there is no scientific literature which has specifically considered the effects of odour on health from exposure to sewage sludge.  It is likely that such odours will affect the quality of life to those regularly albeit intermittently exposed, although regular exposure may result in tolerance and loss of recognition of the odour.  In some cases this can be potentially dangerous as in occupational exposure to hydrogen sulphide where loss of appreciation of the presence of the gas can lead to increasing exposure.  In parallel, workers in smelly occupations regard this tolerance effect as a benefit as they continue to work in what others would regard an impossible environment, providing a continuing income.  As this aspect of sewage sludge seems to be most important to those exposed, studies in this area might be worthwhile although perceptions and pre-conceptions do make such studies difficult to design.  Such studies should involve expertise in odour research.

  27. Risk estimation of non-occupational exposure to sewage sludge

  28. Because the literature is so sparse and in particular the lack of studies which allow definition of dose response relationships for any potential exposure it has not been possible to use a source pathway receptor approach to the information available.  However, it should be appreciated that the time scale for this project was very short and it is likely that there is other literature which may prove helpful in specific areas.
  29. However, we are able to rank risks from non-occupational exposure to sewage sludge in the broadest terms as follows:

  30. Category Risk
    Infections: Salmonella, Giardia, E coli, Hepatitis A, B and E Moderate to low
    Infestations Low
    Chemicals (inc pharmaceuticals) Low
    Metals Low
    Radioactivity  Very low
         Table: Broad qualitative estimates of risk to health from exposure to sewage sludge treated land

  31. The authors have therefore come to the conclusion that there are potential risks to human health that may arise from the treatment of fields with sewage sludge or from its run off but in many situations effects are poorly understood mechanistically. However, from the information currently available the authors are not in a position to quantify these risks.

  32. In the case of exposure to chemicals, effects such as reproductive outcomes are theoretical problems with no human epidemiological data to support this, although as these have not formally been looked for this cannot be regarded as an absence of an effect.  Effects from infections such as enterobacteria are more clear (especially perhaps Salmonellae) but again how big the issue is as a contributor to the overall burden of these infections across Scotland cannot be defined.
  33. The authors believe, therefore that two options are open at this stage: either to accept there is a risk and attempt to find ways to reduce exposure (the precautionary approach) or to collect more specific information and define more clearly the true risks.  Both approaches together might also be appropriate.  Defining more clearly the risk will be expensive and difficult but some suggestions for ways of improving the evidence base are summarised in the following research recommendations.

    1. There is an absence of data on airborne soil and water run-off levels of microbes, chemicals and metals from sewage treated land in the UK.  To some extent the second phase of this study may address this but will not provide information on conventional use of sewage sludge to land as the target sties are those where sewage is being used as land fill.  Provision of such data is essential in determining quantitative risk assessment.  In particular:

    2. Knowledge of soil-based and run-off contaminant concentrations would then allow some estimation of likely exposures to the range of individuals exposed to sewage sludge. It is important to know:

    3. There may be some risk to health (specifically from infections) to populations living near land to which sewage sludge has been applied but from our examination of the scientific literature it is not possible to quantify the degree of risk.  The best paper in this area suggests no such risk but exposure situations will vary from place to place and it is our feeling that such risks are both plausible and possible.  To define this risk studies would be needed of populations living near sewage application sites.  These will be logistically difficult and expensive especially if hoping to define some of the less likely risks to more unusual health end-points.  Such studies would need to investigate multiple communities as each one alone would not provide sufficient statistical power.

    4. Identification and recording of use of sewage sludge application in Scotland, with indications of how well adherence to application guidelines was followed would allow study of health effects in the local communities in relation to exposure to such land.  This is logistically difficult at the record keeping level but is relatively easy to do from the health point of view using Geographic Information Systems (GIS) processes, provided that records of target conditions such as episodes of infectious diarrhoea (especially Salmonella) or pneumonia can be drawn from routinely collected data.

    5. An alternative approach would be to study occupationally exposed workers and from the exposure-response data gathered, extrapolate the degree of risk to those non-occupationally exposed populations who are less exposed.  This is the approach taken by EPAQS, but is not without its difficulties as a range of assumptions have to be made.  However, such an approach would also provide HSE with health and exposure data for this particular work force for which limited UK information is currently available. However, it should also be recognised that workers exposed to low levels of pathogens may build up immunity leading to under estimates in the health of the general population.

    6. The most useful way of addressing these issues is to take an integrated approach to the various aspects of emission, exposure and effect data.  We believe that identifying key target agents and key potential health outcomes as markers of these interactions would enable decisions to be made on approaches to reducing exposures to levels deemed appropriate, on the assumption that if there are risks of lesser degree from other agents, these will automatically fall in line with those which exert a greater effect.  This approach is implicit in the approach to air quality control.

    7. In addition to this we would recommend that further research is carried out to identify those factors that influence the perceptions of a local community with respect to the health risks associated with sewage sludge application. Other studies have shown that the acceptability of any given risk is often influenced by the likelihood and degree of the risk; the reversibility of the health effect; the knowledge or familiarity of the community with the health effect; whether the risk is voluntarily accepted or involuntarily imposed; whether the community is compensated for their exposure to the risk; the advantages of the policy or activity; and the risks and advantages of any alternative action. It is essential that those involved in policy relating to sewage sludge application have as full an understanding of these areas as possible and engage in clear risk communication to the communities most affected.

Key words: sewage sludge, health, respiratory disease, infections, cancer

Copies of this report are available from the Foundation, in electronic format on CDRom at 20.00 + VAT or hard copy at 25.00, less 20% to FWR members.

N.B. The report is available for download from the SNIFFER Website 

All rights reserved.  No part of this document may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of SNIFFER.