Guadalupe-Fernandez et al.

Environmental Health (2021) 20:108

https://doi.org/10.1186/s12940-021-00774-3

REVIEW Open Access

Industrial odour pollution and human

health: a systematic review and meta-
analysis
Victor Guadalupe-Fernandez1,2, Manuela De Sario1, Simona Vecchi1, Lisa Bauleo1* , Paola Michelozzi1,
Marina Davoli1 and Carla Ancona1

Abstract
Objective: To conduct a systematic review to evaluate the association between residential or occupational short-
and long–term exposure to odour pollution from industrial sources and the health status of the exposed
population.
Methods: The searches were conducted in Medline, EMBASE and Scopus in April 2021. Exposure to an
environmental odour from industrial sources in population resident near the source or in workers was considered.
We considered outcomes for which there was a biological plausibility, such as wheezing and asthma, cough,
headache, nausea and vomiting (primary outcomes). We also included stress-related symptoms and novel
outcomes (e.g. mood states). Risk of bias was evaluated using the OHAT tool.
For primary outcomes, when at least 3 studies provided effect estimates by comparing exposed subjects versus not
exposed, we pooled the study-specific estimates of odour-related effect using random effects models.
Heterogeneity was evaluated with Higgins I2.
Results: Thirty studies were eligible for this review, mainly cross-sectional (n = 23). Only one study involved school-
age children and two studies involved workers. Only five studies reported odour effects on objective laboratory or
clinical outcomes. Animal Feeding Operations and waste were the most common industrial sources.
The overall odds ratios in exposed versus not exposed population were 1.15 (95% CI 1.01 to 1.29) for headache (7
studies), 1.09 (95% CI 0.88 to 1.30) for nausea/vomiting (7 studies), and 1.27 (95% CI 1.10 to 1.44) for cough/phlegm
(5 studies). Heterogeneity was a moderate concern. Overall, the body of evidence was affected by a definitely high
risk of bias in exposure and outcome assessment since most studies used self-reported information.
Conclusions: Findings underline the public health importance of odour pollution for population living nearby
industrial odour sources. The limited evidence for most outcomes supports the need for high quality
epidemiological studies on the association between odour pollution and its effects on human health.
Keywords: Odour pollution, Residential exposure, Respiratory effects, Systematic review, Risk of bias

* Correspondence: [email protected]
1
Department of Epidemiology of the Lazio Regional Health Service, ASL
Roma 1 (Italy), Via Cristoforo Colombo, 112, 00147 Rome, Italy
Full list of author information is available at the end of the article

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 2 of 21

Introduction composed by complex mixtures of different volatile

Odour emissions from industrial sites constitute a major chemical compounds. Besides, the sensitivity of people
health issue both for neighbouring residents and and odours responses are different among individuals,
workers, mainly due to the olfactive nuisances generated hindering efforts to monitor and assess its health effects.
during industrial production processes [1–3]. Neverthe- In view of the above, it is considered that odour analyt-
less, little evidence is available on the impact of olfactory ical tools are not sufficiently accurate [1, 5, 13–15].
nuisance, compared to a large number of studies on the However, there are some predictive and observational
toxicity of the chemicals emitted by industrial plants approaches that have been used to estimate population
such as wastewater treatment, livestock operations, com- odour exposure, such as atmospheric dispersion models
posting facilities, landfills, paper and pulp mills or petro- [2], distance to the source [12], frequency of odour
chemical industries. Odour pollution is regulated events per year, sniff tests [1], chemical compounds ana-
differently worldwide, and it is addressed at a national or lysis [16], population complaints monitoring (mean an-
municipal level by different policy frameworks [2, 4]. noyance response or percent highly annoyed residents)
The olfactory function plays an important role in the [3, 10].
detection of hazards in the environment, with the upper As a result, the overall impact on communities of
respiratory tract usually being the first point through odour emissions remains unclear and there has been a
which air pollutants enters the human body. Olfactory rising number of concerns and complaints regarding
receptors of the nasal epithelium may detect odorant their possible health effects, ending up increasing the
compounds inducing sensations in different ways. At el- quantity of studies performed on this topic lately [4, 10].
evated concentrations, odorant receptors may send sig- We conducted a systematic review to evaluate the as-
nals via the olfactory and trigeminal nerve to the brain sociation between residential or occupational short- and
causing different reactions, also known as subjective long–term exposure to odour pollution from industrial
symptoms. Odour sensations processed in the central sources and the health status of the exposed population.
nervous system may induce pleasant reactions, positive
mood and emotions, but also negative responses includ- Methods/design
ing irritation, pain, sneezing, salivation, and vasodilation, Protocol and registration
ultimately resulting in nasal obstruction, bronchocon- Methods and inclusion criteria were registered for
striction, mucus secretion and inflammation. Malodours, PROSPERO (registration number: CRD42018117449).
mould or bad air quality have also been considered as The systematic review was reported according to the
environmental triggers of headaches, eyes irritation, and Preferred Reporting Items for Systematic Reviews and
unusual tiredness [3, 5–10]. It is also important to note, Meta-Analyses (PRISMA) statement guidelines [17].
that individuals’ sensory responses can vary due to
physiological factors, age or sex, persistent exposure, Eligibility criteria
perceived health risk, and various social factors [3, 5– Eligibility criteria were defined based on the PECO state-
11]. Odour-related symptoms seem more common in ment for the key elements (population, exposure, com-
subjects with odour intolerance [5]. In fact, odour seems parator and outcome). The population of interest were
to not have an effect per se, but it is mediated by per- people of any age living near industrial sources or
sonal perception or annoyance [7, 10]. Annoyance is a workers exposed to odour pollution in their workplace.
psychological symptom that can be related to poor qual- We limited the definition of an industrial source as all
ity of life or negative mood states. areas hosting production and processing plants and facil-
Several studies measure odour annoyance and monitor ities for chemicals, petrochemicals, manufacturing, waste
community impact by self-reporting of somatic symp- or water disposal and/or treatment, cement, power gen-
toms, as well as objective health effects, commonly in- eration, mining and metals, and we included other activ-
cluding respiratory inflammation and dysfunctions ities, such as production in industrial installations of
diagnosed by physicians. The population’s characteristics pulp and paper, textile, slaughterhouses and livestock
and health status have traditionally been considered in operations. We excluded studies which assessed the ef-
surveys and structured interviews when approaching fects of exposure to indoor pollution sources. Studies
odour assessment [12]. were included if they captured exposure to an environ-
Estimations of odour frequency, intensity and hedonic mental odour from industrial sources including both ob-
tone in the environment differ substantially among jective and subjective measures. Nevertheless, we
countries, according to their odour regulations [1–4, 12] excluded studies that mainly were focused on malodor-
and there are no standardized methods for population ous toxic compounds emissions since it would be diffi-
and exposure assessment to be used for environmental cult to disentangle the odour effect from the toxic one.
epidemiology studies. Odours emissions are generally The comparison group represented any alternative to
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 3 of 21

the exposed group; this was the minimum criterion for Data collection process and data items
inclusion. For studies that met inclusion criteria, two review au-
We considered primary outcomes for which there was thors independently extracted data using a data extrac-
a biological plausibility with the exposure, such as tion form. Disagreements about the extracted
wheezing and asthma, cough, headache, nausea and information were resolved by discussion with the in-
vomiting. Odour annoyance has been considered both as volvement of the research team when necessary. We
a surrogate for exposure and outcome, having a strong contacted three authors for further information. All au-
association with odour intensity, hedonic tone and mod- thors responded, one of them provided numerical data
elled odour from dispersion models [3]. We also consid- that had only been presented graphically in the pub-
ered secondary outcomes such as respiratory, stress- lished article, one provided a digital poster while the one
related symptoms and other stress-related consequences remaining author could not provide the requested
(e.g. cardiovascular sleep disorders), and also novel out- information.
comes (e.g., mood) [5, 7, 18]. There was no prior restric- Furthermore, the reviewers extracted data on study
tion on the method used for outcome measurement. We year and design from each study, sampling time frame,
excluded studies based on comparisons between odour region or country where the study was performed, sam-
exposure and odour discrimination and hedonic ratings. ple size (target, enrolled, follow-up rates) and character-
We included both observational and experimental study istics of the population, description of the reference or
designs evaluating short- and long–term effects of odour control group, exposure definition (data sources) and as-
pollution with an estimate health effect. sessment (e.g. distance from the facility, odour annoy-
ance using a 5-point-likert scale or dispersion modelling
odour assessment), health outcomes collected (methods
Information sources and search
used to measure the outcome), statistical approach per-
A preliminary search was conducted in bibliographic da-
formed by the authors to analyse the data, confounders
tabases to identify subject terms and free terms relevant
or co-exposures (methods used to measure them and
to the review question. Afterwards we developed a com-
how they were considered in analysis), type of effect
prehensive systematic search strategy using a combin-
measure (Risk Ratio, RR; Prevalence Ratio, PR; Odds Ra-
ation of Medical Subject Headings (MeSH) terms and
tio, OR; beta coefficients; absolute and relative change)
free text terms. We revised the strategy appropriately for
and the 95% confidence interval (CI). When more than
each database to take account of differences in con-
one effect measure was available from the same paper,
trolled vocabulary and syntax rules. We implemented
the following sequential but alternative criteria (if the
our search on April 2021, in Medline (via OVID, 1946
first does not apply, the second works and so on) were
to search date) and EMBASE (1947 to search date). To
applied to choose the estimate to be extracted: that from
identify additional studies, we screened the references
the best adjusted model; the most significant one; the
list of the included studies and searched the related arti-
largest effect size. Information on funding and conflict
cles publication, through Scopus (2004 to search date).
of interest by the authors of the studies was extracted
We set no date, and geographiclimits in our search strat-
and considered when available.
egy. We searched for grey literature by examining differ-
ent university libraries, and national/government/NGO
Risk of bias assessment in individual studies
reports. Furthermore, we contacted experts seeking add-
The risk of bias (RoB) of included studies were inde-
itional information about unpublished and published
pendently assessed by two reviewers. Disagreements
studies. The Ovid search string is presented in Add-
were discussed and resolved with a third author by con-
itional file 1.
sensus. We used the National Toxicology Program/Of-
fice of Health Assessment and Translation (NTP/
Study selection OHAT) Risk of Bias Rating Tool for Human and Animal
We uploaded search results into a reference manage- studies adapted to the review question (Program) [19,
ment software (EndNote, Clarivate Analytics) to man- 20]. The tool considered nine domains: assessment of
age the screening and coding process. Two reviewers exposure, assessment of outcome, confounding (three el-
independently screened titles and abstracts of the re- ements), selection bias, performance bias, attrition/ex-
cords obtained from the searches (VFG, MDS). The clusion bias, outcome reporting bias and inappropriate
full texts of potentially eligible studies were retrieved statistical methods as an additional category for other
for evaluation and inclusion. Any discrepancy regard- potential threats to internal validity. Assessment of con-
ing inclusion or exclusion of a particular study be- founding was based on three elements: 1) the design or
tween reviewers was resolved through discussion by a analysis accounting for confounding and modifying vari-
third reviewer (AC). ables, 2) the adjustment for other concurrent exposures
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 4 of 21

3) the confounding variables measured reliably and consist- which at least 3 studies provided effect estimates by compar-
ently. The first two elements were evaluated according to the ing exposed subjects versus not exposed, we pooled the
minimum set of confounders and co-occurring exposures study-specific estimates of odour-related effects. Effect esti-
considered a priori as relevant: sex, age, educational level/ so- mates using different metrics (e.g. beta coefficients for unit
cioeconomic status (SES)/ employment status, smoking sta- increase in odour or risk ratio across multiple exposure cat-
tus (active/passive) and any co-exposures (noise, traffic egories) were not included in the meta-analysis. We pooled
pollution, air pollution, indoor odour). estimates using random effects models (Restricted Maximum
According to the OHAT risk-of-bias (RoB) tool, for each Likelihood REML Method) [21]. Heterogeneity was evalu-
specific domain, a risk of bias “definitely low”, “probably ated with the I squared statistic [22], where 25%, 50% and
low”, “probably high”, and “definitely high” was assigned and 75% indicate a low, medium, high heterogeneity respectively.
each paper was classified accordingly. We classify individual To assess if exposure assessment (subjective vs objective)
studies into an overall quality category, i.e. tiers from 1 was a potential explanatory factor for the heterogeneity, a
(higher quality) to 3 (lower quality). The entire body of evi- stratified analysis was planned. We planned to assess the
dence was rated and grouped as having “not likely”, “serious” publication bias only for at least 10 effect estimates.
or “very serious” risk of bias, based on the RoB across studies A narrative synthesis of the results was carried out for
and classification tiers. Confidence ratings were integrated secondary outcomes.
on a standard evidence profile table. Meta-analyses were carried out in STATA software
version 14.0.
Data synthesis
Data patterns were explored and evaluated. Outcome- Results
specific odour-related effects were extracted from each Our search identified 5770 records after the removal of
study into evidence tables. For primary outcomes, in cases in duplicates. Of these, 5695 were discarded on the basis of

Fig. 1 Systematic review on industrial odour effects on health - PRISMA flow diagram
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 5 of 21

title and abstracts. No study was identified from grey lit- [37] and van Kersen [29] included NH3 exposure as a
erature sources. Seventy-five records were subsequently proxy of odour exposure. In one study, two different ex-
included in the full-text evaluation. From those, a total posure measures were used, distance and odour fre-
of 30 studies were included in the final synthesis with quency measured by a group of trained panellists [47].
two additional records identified through reference list In 16 studies [9, 23–25, 27–29, 35, 37, 38, 41, 42, 44, 45,
of the studies (Fig. 1). 47, 48], perceived level of exposure was rated labelling
Figure 2 shows the geographical distribution of the different scales (Likert-type scales and other alternatives)
studies by country, with most sudies placed in Europe. through questionnaires/interviews. Several studies used a
Study size ranged from 15 to 58,169 subjects. The ma- dichotomous exposure of odour annoyance and/or
jority of the studies had a cross-sectional design (n = 23), odour perception, defined as presence/absence [9, 24,
while seven were panel studies [23–29]. 29, 35, 36, 39, 41, 43, 45] .
The characteristics of the included studies, ordered by Retrospective and self-reported information on out-
study design and by publication year (newer to older), comes, questionnaire-based, was the most widely used
are summarized in Table 1 and additional information method for measuring primary outcomes. Most studies
are reported in Additional files 2 and 3. were related to both acute (e.g. symptoms, worsening of
Only one study [43] involved a sample of school-age disease) and chronic outcomes (e.g. prevalence of dis-
children (age range: 12–14 years). We observed a large eases), with different timing of data collection, with past
heterogeneity in terms of type of industrial source, study year prevalence in some studies [30, 36, 39, 43, 45, 49]
population, measurements for exposure and outcome or past 2 years [37, 40], or past 6 months prevalence
(i.e. objective or subjective) and type of outcomes. Re- [38], past 1 month [32, 33, 37], or current symptoms [9,
garding industrial source of exposure, 13 studies were 29, 31, 34, 42, 43, 45, 47].
conducted on Animal Feeding Operations (AFOs), 10 On the contrary, the seven panel studies focused on
studies on waste (both solid and liquid waste), 2 were on short-term or acute outcomes, that varied on a daily
multiple sites, and 6 were on other industrial exposure base, such as symptoms of disease [24, 27], or mood [25,
(e.g. paper, petrochemical plant). 26] or biological parameters such as lung/bronchial
Ten studies used distance from the source as proxy of function [27, 29], immune function and allergy [23],
odour exposure [26, 30, 31, 39, 40, 43, 45, 47, 50, 51]. blood pressure [28]. In addition, also a cross-sectional
Boers et al. 2016 estimated odour exposure using the study [42] reported objective outcomes (bronchial hyper-
Stacks dispersion model [34]. Lipscomb et al. [49] de- responsiveness to methacholine, IgE concentration). In
fined a measure of exposure based on odour zones some studies information on timing of outcome data
adopted from an earlier survey. In addition, Blanes-Vidal collection was not provided [9, 46–48, 50, 51].

Fig. 2 Geographic distribution of the included studies on industrial odour effects on health
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 6 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
Cross-sectional studies
Kret 2018 [30], Waste (landfill) N = 343 adults Distance (km) Questionnaire: Model: n.a Matching for
USA, households within a self-reported Effect percentage of
Cross-sectional 3.2-km radius prevalence of estimated: white population
(173 exposed; 170 diseases and 12 n.a and for 25+
non-exposed) months symptoms; Weighted population with
odour annoyance (5- prevalence education level at
point Likert scale) (95%CI) least high school.
Groups: No effect estimate.
Odour nuisances
Lower respiratory
symptoms
Upper respiratory
symptoms
Gastrointestinal
symptoms
Mucus irritation
General ill feeling
Hayes 2017 [31], Wastewater N = 153 residents Questionnaire Questionnaire: Model: Social readjustment
Australia, treatment Plant within a 3-km radius (presence/absence Self-reported ANOVA scale by Holmes
Cross-sectional on two exposed of bad smells and psychological Effect and Rahe 1967
(with a history of high odours impacting symptoms past estimated: added as covariate
or low number of community) week; odour None
complaints) and one annoyance (10-point
control sites scale)
Groups:
Mood states
Tjalvin 2015 [32], Chemical N = 284 workers in Questionnaire: Subjective Health Model: Age, gender,
Norway, Industry 2008 and 203 in 2012 Workers exposure Complaints (SHC) Linear mixed smoking habits,
Cross-sectional (Chemical (exposed workers history score effects educational level.
explosion in an employed in 2008 Groups: models with
Industrial and/or clean-up General ill feeling random
harbour) workers, proximity to Gastrointestinal intercept and
the explosion ≤1 km; symptoms slope
control workers) range Upper respiratory Effect
of age 18–67 symptoms estimated:
Immune function Mean
Cardiovascular difference
problems
Tjalvin 2017 [33], Chemical N = 486 workers Questionnaire: Questionnaire: Model: Age, gender,
Norway, Industry employed in 2008 Low/high odour Subjective Health Linear mixed smoking habits,
cross-sectional (Chemical (18% present during score (% of months Complaints (SHC) effects educational level,
(repeated survey explosion in an the explosion), each participant score previous models with absence/presence
of Tjalvin 2015) Industrial in 2010 (n = 379), noticed the odour month; Impact of random during the
harbour) 2012 (n = 252) in 2008) Event Scale-Revised intercept and explosion (> 1 km
Adults aged 18–67 (IES-R) previous 7 slope or ≤ 1 km)
years days Effect
Groups: estimated:
General ill feeling Mean
Mood states difference
Boers 2016 [34], Animal feeding N = 582 residents Calculated Questionnaire: Model: Age, educational
Netherlands, operations living near livestock exposure: 98th Self-reported odour multivariate level, indoor air
Cross-sectional farms percentile of odour annoyance (4-point logistic pollution, asthma,
Mean age = 51 years concentrations scale) regression or lower back pain
old (SD 13) (OUE/m3) from Groups: analysis
(part of the Stacks dispersion Odour nuisances Effect
population included model estimated:
in Hooiveld 2015) ORs(95%CI)
Hooiveld 2015 [35], Animal feeding N = 753 adults, Questionnaire: Questionnaire: Model: Smoking status,
Netherlands, operations residents with Self-reported odour Self-reported Multiple growing up at farm,
Cross-sectional asthma or lower annoyance (yes/no) symptoms last ordinal age, gender,
back pain month; general logistic, nationality, marital
health (5-point Likert logistic and status, educational
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 7 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
scale from bad to Poisson level, asthma or
very good) regression lower back pain,
Groups: analysis. other
Gastrointestinal Effect environmental
symptoms estimated: annoyances (noise,
General ill feelings ORs(95%CI) traffic and air
Lower respiratory pollution)
symptoms
Upper respiratory
symptoms
Mood states
Baldacci 2015 Waste N = 1407 residents Questionnaire: Questionnaire: Model: Gender, age,
[36], Italy, (incinerator) within 4-km radius Self-reported odour Self-reported Multivariate residence/
Cross-sectional from the incinerator annoyance (no, symptoms past 12 logistic incinerator distance,
and a control group. slightly annoying, months. regression educational level,
Mean age 44.4 very annoying) Groups: analysis working position,
(SD 22.1) Lower respiratory Effect smoking status,
symptoms estimated: passive smoking,
Upper respiratory ORs(95%CI) residential time,
symptoms occupational
exposure.
Blanes-Vidal 2015 Waste N = 454 Residents NH3 concentration: Questionnaire: Model: Age, gender,
[37], Denmark, (biodegradable) from six study areas loge (NH3 Self-reported multivariate smoking habit, job,
Cross-sectional in Denmark. exposure), NH3 symptoms past 2 logistic time spent at home
Mean age 54 (SD 14) exposure levels (< 2, years, odour regression per week, existence
2–3, > 3 μg/m3), annoyance ((no, analysis. of household
Questionnaire: slightly, moderately, Effect residents below 18
Self-reported odour very, extremely) estimated: years old, years
annoyance (no, Groups: ORs(95%CI) living in the region,
slightly, moderately, Odour nuisances and acute and
very, extremely) Gastrointestinal chronic respiratory
symptoms conditions
General ill feeling
Mood states
Wing 2014 [38], Sewage Sludge N = 158 adults, Questionnaire: Questionnaire: Model: Age, gender, race,
USA, and Animal residents living near Self-reported odour Self-reported Linear and educational level,
Cross-sectional feeding liquid TSS, 85 living annoyance past six symptoms past six poisson smoking status,
operations near cake TSS, and months (none/faint months regression passive smoking,
188 living in and moderate/ Groups: analysis agricultural
comparison areas strong/very strong) Gastrointestinal Effect chemical odours
symptoms estimated: and odours from
Mucus irritation Mean factor burning
General ill feeling score
Lower respiratory differences
symptoms (95%CI) and
Upper respiratory PRs (95%CI)
symptoms
Skin disorders
Aatamila 2011 Waste (Landfills N = 1142 residents Distance zone (< Questionnaire: Model: Model 1: adjusted
[39], Finland, and within a 5-km radius 1.5, 1.5–3, > 3 km) Self-reported Logistic for sex, age,
Cross-sectional composting of six different Questionnaire: symptoms past 12 regression educational level,
sites) biowaste sites odour perception months analysis Socio economic
Range of age: (4-point scale) Groups: Effect level and smoking
25–64 years stratified into Gastrointestinal estimated: Model 2:
sensitive vs not symptoms ORs(95%CI) additionally,
sensitive, odour Mucus irritation adjusted for odour
annoyance (4-point General ill feeling sensitivity
scale) categorized Lower and upper
as annoyed vs not respiratory symptoms
annoyed Skin disorders
Herr 2009 [40], Waste N = 477 residents Distance (km): Questionnaire: Model: Adjusted for age,
Germany, (composting living “near” two EnvExp2 (odour- Self-reported Logistic gender, and
Cross-sectional sites) composting sites. only exposed) and symptoms past 2 regression educational level
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 8 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
(263 EnvExp2 and a control group years analysis
214 control group). Groups: Effect
Individuals aged Gastrointestinal estimated:
≥16 years old symptoms ORs(95%CI)
General ill feeling
Lower respiratory
symptoms
Mood states
Skin disorders
Cardiovascular
symptoms
Sucker 2008 [41], Industrial sites N = 1434 adults from Questionnaire: Questionnaire: Model: Noise disturbance,
Germany, each household Log-values of odour Odour annoyance; Logistic length of residence,
Cross-sectional (the homemaking frequency Intensity self-reported health regression quality of
or the person (6-point scale from complaints analysis. residential area,
spending most of “very slight” to “ex- Groups: Effect tenant or owner,
the time at home) tremely strong”), Odour nuisances estimated: single/multiple
Hedonic tone (9- General ill feeling ORs(95%CI) houses, average
point scale with Mucus irritation time at home,
values ranging from perceived health,
“-4” “extremely un- smoking habit,
pleasant” through gender, age, marital
“0” “neither pleasant status, educational
nor unpleasant” to level
“+ 4” “extremely
pleasant”)
Radon 2007 [42], Animal feeding N = 5556 Residents Questionnaire: Questionnaire: Model: Age, sex, active and
Germany, operations from four rural Self-reported odour Self-reported Logistic and passive smoking,
Cross-sectional town with high annoyance (4-point symptoms during linear educational level,
density of AFOs Likert scale from the week. regression number of siblings
Mean age 33.6 “not at all” to Clinical analysis and parental
(SD 7.4) “strongly”) measurements: Effect allergies. FEV1
Specific IgE to estimated: additionally,
common allergens ORs(95%CI) adjusted for passive
> 0.35 IU/mL, smoking during
bronchial childhood
hyperresponsiveness
to methacholine,
forced expiratory
volume in 1 s (FEV1)
Group:
Lower and upper
respiratory
symptoms
Immune function
and allergy
Mirabelli 2006 Animal feeding N = 58,169 students Questionnaire: self- Questionnaire: Model: Age, race,
[43], USA, operations of 265 schools reported indoor Current and past 12- Random- socioeconomic
Cross-sectional within 3 miles of at and outdoor odours month self-reported intercepts status, smoking,
least one AFO source from schools (binary respiratory symp- binary regres- school exposures
Range of age: coded variable, “re- toms and medical sion analysis and household
aged 12–14 ported”/“no care Effect exposures
reported”) Groups: estimated:
Lower respiratory PRs (95%CI)
symptoms
General ill feeling
Radon 2004 [44], Animal feeding N = 2745 Residents Questionnaire: QoL questionnaire: Model: Age, gender,
Germany, operations living in rural towns Self-reported odour Physical SF-12 score, Multiple linear respiratory
Cross-sectional close to intensive annoyance (4-point emotional SF-12 regression symptoms,
animal production Likert scale from score analysis smoking, living on
Mean age “not at all” to Groups: Effect or close to a farm
32.7 (SD 7.7) “extremely”) General ill feeling estimated: and employment
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 9 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
Mood states β (SE) status.
Segala 2003 [45], Wastewater N = 2867 residents Distance zones (< Questionnaire: Model: Age, sex,
Canada, treatment plant from 8 nearby towns. 1.5, 1.5–3, 3–4.5 km) Self-reported Multivariate educational level,
Cross-sectional Distance zones: Questionnaire: symptoms past logistic active vs inactive,
3–4.5 km (N = 1003), Self-reported odour month and year regression smoking status,
mean age 47.5 (SD 15.2) tolerance (“tolerant”, Groups: analysis family size,
1.5–3 km (N = 1007), “moderately toler- Gastrointestinal Effect satisfaction with
mean age 48.2 (SD 67.7) ant”, “intolerant”), symptoms estimated: neighbourhood life
< 1.5 km (N = 857), odour annoyance Mucus irritation ORs(95%CI)
mean age 49.8 (SD 15.1) (“annoyed with im- Lower respiratory
pact on health”, symptoms
“annoyed without Upper respiratory
impact on health”, symptoms
“not annoyed”) General ill feeling
Cardiovascular
problems
Georgieff 1999 [46], Paper industry N = 538 Residents Questionnaire: Questionnaire: Model: n.a. None
Bulgaria, from Stamboliisky Self-reported un- Self-reported Effect
Cross-sectional town pleasant odour symptoms estimated:
Range of age: (yes/no) Groups: n.a.
16–60 years old General ill feeling Percentages
Lower respiratory (%) of
symptoms number of
Mood states reported
Immune function somatic
and allergy symptoms
Steinheider 1998 Nettetal study Nettetal study Nettetal study Questionnaire: Nettetal None
[47], Germany, Fertilisers (N = 250) 1) Distance from Self-reported study
Cross-sectional production Nörvenich study the odour source symptoms and Model:
plant (N = 322) Close: within 400– odour annoyance Analysis of
Nörvenich Adults aged 800 m (11-point graphic variance
study ≥18 years old Medium: 1600 m scale) Effect
Pig rearing Far (control area): 6 Control variables estimated:
facilty and 3.5 km (fever and asthma None
2) 11-point graphic attacks) Nörvenich
scale of Odour Groups: study
annoyance Odour nuisances Model:
Nörvenich study Gastrointestinal Linear
1) Log-values of symptoms regression
odour frequency General ill feeling analysis.
(odour hours/year). Lower respiratory Effect
34 observation symptoms estimated:
points; 2) 11-point Mood states β (SE)
graphic scale of
Odour annoyance
Steinheider 1993 Industrial sites N = 1539 adults, living Log-values of odour Questionnaire: Self- Model: Age, sex,
[48], Germany, 1) Duisburg- near of four cities in frequency (odour reported odour an- Multivariate educational level,
Cross-sectional chemical plant North Rhine-Westphalia. hours/year). noyance (11-point linear profession, length
2) Dortmund – Duisburg (N = 400), scale) regression of residence and
iron/steel plant Dortmund (N = 400), Groups: analysis perceived health.
3) Brühl – Brühl Odour nuisances Effect Dortmund, Brühl
castiron factory (N = 539), Rodenkirchen estimated: and Rodenkirchen
and sugar (N = 200) n.a. added coping
refinery strategies to the
4) model
Rodenkirchen –
oil refineries
Lipscomb 1991 Waste (McColl N = 193 residents living Exposure areas Questionnaire: Model: None
[49], USA, waste disposal nearby a disposal (high, medium, and Self-reported n.a.
Cross-sectional site) waste site low) based on an symptoms past 12 Effect
Adults ≥22 years old odour survey months estimated:
conducted in 1981 Groups: Crude
Odour nuisances PORs(95%CI)
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 10 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
Gastrointestinal
symptoms
Mucus irritation
General ill feeling
Lower respiratory
symptoms
Upper respiratory
symptoms
Mood states
Skin disorders
Immune function
and allergy
Shusterman 1991 Waste N = 2040 residents Self-reported Questionnaire: Model: None
[9], USA, living near three frequency of odour Self-reported n.a.
Cross-sectional hazardous waste sites perception (“none”, symptoms Effect
in Southern California “less than or equal Groups: estimated:
McColl. Acid petroleum to four times per Odour nuisances PORs(95%CI)
sludge (N = 670) month” and Gastrointestinal
Operating Industries. “greater than four symptoms
Municipal and sewage times per month” Mucus irritation
(N = 514) General ill feeling
Del Amo-Montrose.
Residues from
synthetic rubber
manufacturing
(N = 856)
Deane 1978 [50], Refineries and N = 291 Residents Exposure areas Questionnaire: Self- Model: n.a. None
USA, Cross-sectional other living in three estimated by reported symptoms Effect
petrochemical residential areas dynamic Groups: estimated:
industries nearby refineries and olfactometry: High Odour nuisances n.a.
petrochemical plants (Area I), Moderate Gastrointestinal Frequency of
(Area II), Low (Area symptoms self-reporting
III). Mucus irritation outcomes
General ill feeling
Lower respiratory
symptoms
Upper respiratory
symptoms
Mood states
Deane 1977 [51], Paper industry N = 140 Adults living Exposure areas: Questionnaire: Model: Analysis were
USA, Cross-sectional in three residential high (1–2 miles Self-reported n.a. stratified by odour
areas nearby a pulp southeast of the symptoms Effect annoyance and
mill mills), moderate (2– Groups: estimated: gender
3 miles east of the Odour nuisances Frequency of
mils), low (4 miles Gastrointestinal self-reporting
east of the mills) symptoms outcomes
Mucus irritation
General ill feeling
Lower respiratory
symptoms
Upper respiratory
symptoms
Mood states
Panel studies
Van Kersen Animal feeding N = 82 adults COPD NH3 concentration Questionnaire: Model: Adjustment for
2020 [29], operations non smokers residents (μg/m3), Self-reported Generalized ambient
Netherlands, Panel in the eastern part of Questionnaire: symptoms past 12-h estimated temperature,
(3 months) the province of Self-reported odour Clinical equations relative humidity
Noord-Brabant and annoyance (no, yes) measurements: (GEE) and day-in-study
the northern part of the Lung function assuming a (linear trend), PM10;
province of Limburg (forced expiratory first order Restriction to non-
(high prevalence volume or FEV1 and autoregressive smokers by study
of lifestocks) peak expiratory flow (AR1) design
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 11 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
rate or PEF) correlation
Groups: structure
Lower respiratory Effect
symptoms estimated:
Upper respiratory ORs (95%CI)
symptoms
Wing 2013 [28], Animal feeding N = 101 non-smoking Data-collection Systolic (SBP) and Model: Time-of-day (AM or
USA, Panel operations residents living within diary: Self-reported diastolic (DBP) blood Linear fixed- PM)
(2 weeks) 1.5 miles of an CAFOs odour annoyance pressure values effects
source (9-point Likert scale) Groups: models Effect
Adults aged ≥18 years old. Cardiovascular estimated:
Mean age 53.7 (19.2–89.5) problems β (SE)
Heaney 2011 [24], Waste (landfill) N = 23 adults, residents Questionnaire: 12- Questionnaire: Model: Time of day (AM/
USA, Panel within 0.75 miles of h of self-reported Self-reported Conditional PM)
(14 days) the landfill odour annoyance symptoms past 12-h fixed effects
(5-point Likert scale) Groups: logistic
Gastrointestinal regression
symptoms models
Mucus irritation Effect
General ill feeling estimated:
Lower respiratory ORs(95%CI)
symptoms
Upper respiratory
symptoms
Mood states
Skin disorders
Schinasi 2011 [27], Animal feeding N = 101 Non-smoking Questionnaire: 12- Questionnaire: Model: Time of day (AM/
USA, Panel (14 days) operations residents within h of self-reported Self-reported Conditional PM)
1.5 miles of an AFOs odour annoyance symptoms past 12-h fixed effects
source (9-point scale) Clinical logistic and
Mean age 53.7 (19.2–89.5) measurements: linear
Lung function regression
(forced expiratory analysis
volume or FEV1and Effect
peak expiratory flow estimated:
rate or PEF) β (SE)
Groups:
Gastrointestinal
symptoms
Mucus irritation
General ill feeling
Lower respiratory
symptoms
Upper respiratory
symptoms
Skin disorders
Horton 2009 Animal feeding N = 101 Non-smoking 12-h of self- Questionnaire: Model: Time of day (AM/
[25], USA, operations residents within reported odour an- Self-reported Logistic PM)
Panel (2 weeks) 1.5 miles of an AFOs noyance (9-point information on mixed models
source scale) mood states with random
Mean age 53.7 (19.2–89.5) Groups: intercepts.
Mood states Effect
Odour nuisances estimated: β
(SE) and
ORs(95%CI)
Avery 2004 [23], Animal feeding N = 15 residents within Questionnaire: Clinical Model: Day of data
USA, Panel (2 weeks) operations 2.4 km of an intensive Self-reported odour measurements: Hierarchical collection (1–14)
hog operation facility annoyance (9-point Log salivary IgA mixed models and time of day
Mean age 55.3 (SD 13.4). scale, coded as a concentration (μg/ Effect (AM/PM)
seven-level continu- ml) and secretion estimated:
ous variable) rate (μ g/ml) β (SE)
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 12 of 21

Table 1 Summary of characteristics of studies included in the systematic review and meta-analysis by study design (cross-sectional,
panel) and publication year (newer to older) (Continued)
Study, Country, Industrial Study population, Exposure Outcome Statistical Adjustment for
Study design source age group assessment assessment analysis confounders
Group:
Immune function
and allergy
Schiffman 1995 Animal feeding N = 88 Distance and Profile of Mood Model: adjusted by design
[26], USA, operations Exposure group,: duration: Exposed States (POMS) factors ANOVA (matching by
Panel (4 days) Mean age 52.0 ± 13.4 living an average of and Total Mood Effect gender, age, race,
Control group: 5.3 + 6.5 years near Disturbance (TMD) estimated: and education)
Mean age 51.7 ± 8.3 hog operations and score n.a.
comparison group Groups:
Mood states

Most cross-sectional studies took into account the po- measures was characterized by a high risk of bias in
tential confounding of age, sex, smoking status, educa- most studies since information was mostly self-
tional level and/or SES [32, 33, 35–39, 41–45]. Panel reported. Eight studies accounted for potential co-
studies [23–29] were adjusted only by time-varying vari- exposures, such as smoking, noise, indoor and/or out-
ables (e.g. time of the day when outcome was measured) door pollution and were judged at very low/low risk
because they do not need to adjust for individual con- of bias [24, 25, 27, 35, 38, 40, 41, 43].
founders since the study population serves as its own The risk of selection bias resulted to be definitely high
control. Eight studies [9, 26, 30, 46, 47, 49–51] did not in five studies because the control group could not be
account for any confounder and only one [30] reported defined as truly unexposed [31, 38, 41, 46, 49] or be-
to have matched exposed and control population by age, cause personal attitude towards livestock farming could
race and education level. One study on COPD patients have influenced participation [29]. The risk of selection
restricted the study population to non-smokers [29]. bias was probably high in most studies. Additionally, in
Figure 3 shows the results of evaluation of the risk of six studies [32–35, 43, 44] and in the two Steinheider’s
bias of the studies selected for the review. Overall, the study sites [47] no information was provided as to
body of evidence was affected by a definitely high risk of whether selection of study participants resulted in ap-
bias in exposure and outcome assessment since most propriate comparison groups (Fig. 3). Regarding the at-
studies used self-reported information. The study from trition bias, in 10 studies [25, 26, 32, 33, 36, 40, 45, 46,
Steinheider 1998 has been evaluated separately for the two 48, 49] and in the two Steinheider’s study sites [47] the
study sites (Norvenich study labelled as a) and Nettetal information about loss of participants was unclear or in-
study labelled as b)) [47]. Sixteen studies were classified in complete, hence they were considered at “probably high”
the worst quality level (3rd tier), 10 studies in the second risk of attrition bias. Missing values related to outcome
(2nd tier) and five studies in the first category (1st tier). variables in the study were treated in the analysis. Only
Confidence in exposure and outcome assessment was three studies were classified at “definitely low risk” of at-
very low in most studies. Only three studies were judged trition bias [23, 28, 39]. Six studies were judged at prob-
at low risk of bias since used objective outcome measures ably or definitely high risk of reporting bias [26, 29–31,
or only exposure from dispersion models [23, 28, 34]. 39, 40], and, additionally, two studies [36, 46] were at
As for confounding, adjustment with a minimum unclear risk since outcomes were not reported with suf-
set of potential confounders was achieved in most ficient detail in the short communications. A probably
studies for which the risk of bias was labelled low; 11 low risk of reporting bias was found in Sucker et al. [41],
studies that did not account for any confounders were after evaluating a previous publication of another part of
graded as “probably high” or “definitely high” risk of the results [52].
bias [9, 26, 30, 31, 34, 40, 46–51]. The second con- Regarding the additional element of appropriate statis-
founding element referred to the adjustment of other tical methods, nine studies were judged at high risk
environmental exposure and in this case most studies (probably or definitely) since they provided only a de-
did not adjust for concurrent exposures. For example, scriptive analysis [9, 26, 31, 42, 44, 46, 49–51] and in the
panel studies that only accounted for time of day two Steinheider’s study sites [47].
(morning /evening) were considered as “probably high Health outcomes were grouped as follows (Additional
RoB” [23, 24, 28], due to the lack of adjustment for file 2): general ill feelings (e.g. headache, sleeping prob-
time-varying air pollution or noise. The third con- lems), gastrointestinal symptoms (e.g. nausea/vomiting,
founding element regarding validity and reliability of reflux), lower and upper respiratory symptoms (e.g.
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 13 of 21

Fig. 3 Studies on industrial odour effects on health according to the NTP/OHAT risk of bias approach. * conference proceeding; **Steinheider
et al. 1998a Nörvenich study, Steinheider et al. 1998b Nettetal study

cough/phlegm, wheezing), immune function/allergy studied in Steinheider [47]. Only graphical results of
mucus irritation, skin disorders, mood states, cardiovas- Odds Ratios (and 95% confidence intervals) were provided
cular problems, and odour nuisances (e.g. odour annoy- for the association between NH3 exposure and prevalence
ance, risk perception). We ran meta-analyses for of symptoms in the study of van Kersen 2020 [29].
headache, nausea/vomiting and cough/phlegm. The Nineteen studies analysed general ill symptoms as
Additional file 3 reported also the results not included health outcome of odour related effects [9, 24, 27, 30,
in the meta-analyses of the association between residen- 32, 33, 35, 37–41, 44–47, 49–51] (Additional file 3). All
tial or occupational, short- and long–term exposure to studies were on adults. Two studies were conducted
odour pollution from industrial sources and the risk. No among workers [32, 33].
measure of association was available for five studies [9, Headache was the most common general ill symptom,
30, 46, 50, 51] and for one of the locations (Nettetal) being reported in 16 studies. Pooled analysis showed an
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 14 of 21

Fig. 4 Forest plot of study-specific and pooled Odds Ratio (OR) and 95% Confidence Intervals (95%CI) of residential exposure to odour and
headache in exposed versus non exposed subjects

increased risk of headache in exposed versus not exposed 1.27 to 10.5); odour intolerant vs tolerant (OR = 2.64; 95%
(OR = 1.15, 95% CI: 1.01 to 1.29) with moderate hetero- CI 2 to 3.5); group with complaints with impacts on health
geneity (I2 = 66%, p-value = 0.004) (Fig. 4). Among studies vs no complaint group (OR = 2.04; 95% CI 1.46 to 2.84).
that were not included in the meta-analysis (Additional Ten studies evaluated exposure to odour objectively
file 3), one study showed increasing headache prevalence [30, 32, 37, 39, 40, 45, 47, 49–51], reporting sparse evi-
[47] and two studies [37, 45] showed increasing risk in the dence of association for dizziness [40], sleeping difficul-
highest exposure categories: at extremely annoyed com- ties [47], fatigue [49], joint pain [39], fever past 12
pared to those who were not annoyed (OR = 3.65; 95% CI: months [39] and toothache [49].

Fig. 5 Forest plot of study-specific and pooled Odds Ratio (OR) and 95% Confidence Intervals (95%CI) of residential exposure to odour and
nausea/vomiting in exposed versus non exposed subjects
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 15 of 21

Among studies evaluating exposure subjectively [9, 24, There were no observed differences between groups
27, 35, 37–39, 45, 51], most consistent associations were for the gastrointestinal score among workers were ob-
found for dizziness [24, 35, 37, 45], unnatural fatigue served [32].
[37, 39, 45] and joint/muscular pain [39, 45]. Sixteen studies reported the association of lower re-
Among exposed workers, significant higher total sub- spiratory symptoms with odour pollution [24, 27, 29, 30,
jective health complaint (SHC) score [53] and the sub- 35, 36, 38–40, 42, 43, 45, 47, 49–51]. All studies were on
jective neurological complaints score were found in adults except one [43]. No study was conducted on
exposed workers than in controls and these associations workers.
lasted for at least 3 years after the pollution was re- Eleven studies reported cough and phlegm as odour-
moved [32, 33]. related symptoms [24, 27, 30, 35, 36, 38, 39, 45, 47, 50,
Fifteen studies reported gastrointestinal symptoms [9, 51]. Pooled analysis showed an Odds Ratio of 1.27 (95%
24, 27, 30, 32, 35, 37–40, 45, 47, 49–51]. All studies were CI: 1.10 to 1.44) (see Fig. 6), with moderate heterogen-
on adults. Only one study included workers [32]. eity (I2 = 53.8%, p-value = 0.043). Among studies that
The most frequent gastric symptom reported was nau- were not included in the meta-analysis (Additional file
sea/vomiting. Seven studies [9, 24, 35, 38–40, 49] were 3), self-reporting of cough/phlegm was significantly
feasible to meta-analysis (Fig. 5), showing an increased higher in the study of Segala et al. [45] in the highest ex-
risk of these symptoms (OR = 1.09; 95% CI: 0.88 to 1.30) posure categories: odour intolerant vs tolerant (OR =
with a low heterogeneity (I2 = 28.3%; p-value = 0.193). 2.35; 95% CI 1.75 to 3.15) and in the group with com-
Among studies not included in the meta-analysis (Add- plaints with impacts on health vs. no complaints (OR =
itional file 3), self-reporting of vomiting, nausea or 1.64; 95% CI 1.15 to 2.32) and in Aatamila et al. [39] the
retching was significantly higher for increase in odour fre- group of residents living closer to the waste site (dis-
quency in Nörvenich site [47] and in the study of Segala tance< 1.5 km: OR = 1.3; 95% CI 1 to 1.8). Cough was
et al. [45] in the highest exposure categories: odour in- significantly associated with odour frequency and even
tolerant vs tolerant (OR = 3.52; 95% CI 2.14 to 5.8) and in with odour annoyance after adjustment for odour fre-
group with complaints with impacts on health vs no com- quency; however, no direct link was revealed between
plaint group (OR = 2.11; 95% CI 1.13 to 3.94). Estimates of lower respiratory complaints and odour frequency after
the odour-nausea association tended to increase as the adjustment for odour annoyance [47]. Increasing reports
level of odour annoyance increased, but results were not of cough in the past 12 h related to 12-h mean odour
significant in Blanes-Vidal et al. [37]. were found in Schinasi et al. [27].
Among other gastric symptoms, eight studies mea- Among other respiratory symptoms, 10 studies re-
sured exposure objectively [32, 39, 40, 45, 47, 49–51]. ported exposure objectively [29, 30, 39, 40, 43, 45, 47,
High exposure to odours was associated with greater 49–51], mainly with distance as a proxy of exposure.
prevalence of loss of appetite (OR = 4.27; 95% CI: 1.43 to Only three studies reported significant findings [43, 47,
12.73) [49]. One study [47] showed a higher frequency 49] for wheezing, asthma and shortness of breath. Mira-
of gastric symptoms (disgust, loss of appetite, stomach belli et al. [43] found school proximity within 3 miles of
discomfort) when the frequency of odour exposure was a swine CAFO was related to higher physician-
increased. Another study [51] reported a significant diagnosed asthma (PR = 1.07; 95% CI: 1.01 to 1.14,
trend by area among women who had reported fre- mostly in non-allergic adolescents PR = 1.14; 95% CI:
quently or occasionally constipation. 1.01 to 1.26), asthma medication use (PR = 1.07; 95% CI:
Seven studies evaluated exposure subjectively [24, 27, 1.00 to 1.15), asthma-related visit to a physician or an
35, 38, 39, 45, 51]. Segala et al. [45] reported more fre- emergency department or hospitalization (PR = 1.06;
quent diarrhoea in people with self-reported odour in- 95% CI: 1.00 to 1.12), while for wheezing no clear associ-
tolerance (OR = 2.18, 95% CI: 1.43 to 3.33) or ation was found (Additional file 3) [43].
experiencing malodour-related health complaints (OR = Between studies evaluating exposure subjectively [24,
2.83, 95% CI: 1.82 to 4.4); however, the same study did 27, 29, 35, 36, 38, 39, 42, 43, 45, 49, 51], eight reported
not report any significant association in people with significant health effects [27, 35, 36, 38, 39, 42, 43, 45].
complaints that were not related to health (OR = 1.08, Most consistent estimates were reported for asthma [36,
95% CI: 0.74–1.58). Aatamila et al. [39] found an in- 42, 45], while associations with wheezing were weaker.
creased risk of diarrhoea in the group with odour per- Wing et al. [38] showed odour from livestock facilities
ception (OR = 1.3, 95% CI: 1 to 1.7) and with odour was significant related to difficulty breathing (PR = 1.52,
annoyance (OR = 1.2, 95% CI: 0.9 to 1.7). Statistically sig- 95% CI: 1.02 to 2.27) and increased the lower respiratory
nificant associations with stomach pain, gastrointestinal diseases score (mean difference = 0.28, 95% CI: 0.05 to
symptoms and constipation were reported in Hooiveld 0.5) for moderate/strong/very strong odour group. Ac-
et al. [35]. cording to Segala et al. [45], people complaining odour
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 16 of 21

Fig. 6 Forest plot of study-specific and pooled Odds Ratio (OR) and 95% Confidence Intervals (95%CI) of residential exposure to odour and lower
respiratory symptoms

intolerance had a higher prevalence of self-reported re- Regarding studies with subjective exposure [24, 27, 35,
spiratory infections (OR = 4.81, 95% CI: 3.24 to 7.14) or 38, 39, 45, 51], a significant effect of odour with an in-
COPD (OR = 2.95, 95% CI: 1.84 to 4.73), and similar creased risk for runny nose was found in only three [24,
findings were found for the group with complaints with 27, 45]. In Segala et al. [45] the higher risk was found
impacts on health vs. no complaints for COPD (OR = both in people with self-reported chemical intolerance
2.05; 95% CI 1.21 to 3.49). People complaining about (OR = 2.1, 95% CI: 1.59 to 2.78) and in people complain-
odours in terms of a health threat are found to be at a ing of malodour in terms of a health threat (OR = 1.69,
higher risk of enduring cough and COPD. Nonetheless, 95% CI: 1.22 to 2.32). A border line association was be-
the precision of the effect estimate is lower in this sense. tween cold/flu in last month and odour annoyance [35]
The included studies showed no association between (OR = 1.38, 95% CI: 0.97 to 1.99).
odour and chest pain in the included studies. In the only study conducted on workers, there were
Only three studies evaluated lung function and bron- no significant differences between the flu score in ex-
chial hyperresponsiveness [27, 29, 42]. A reduction in posed subjects and the control group [32].
PEF and FEV1 with increasing odour was suggested in Five studies evaluated the effect of odour on the im-
all studies [27, 42], however, 95% CIs included the null mune system and allergic sensitization by estimating IgE
value, except than for the association between evening and IgA concentration and an allergy score obtained by
PEF (lag 0) and odour annoyance in the van Kersen questionnaires, using self-reported exposure [23, 42] or
study [29]. In addition, no associations were seen be- objective exposure [32, 46, 49], but no association with
tween self-reported odour annoyance and bronchial increasing odour exposure emerged.
hyper-responsiveness to methacholine [42]. Twelve studies evaluated odour effect on mucous
Eleven studies [24, 27, 29, 32, 35, 38, 39, 45, 49–51] membrane irritation [9, 24, 27, 30, 35, 38, 39, 41, 45,
presented data regarding associations between odours 49–51] (Additional file 3). Six studies were conducted
and upper respiratory symptoms (Additional file 3). All on skin disorders [24, 27, 38–40, 49]. All studies were
studies were on adults. Only one study was conducted on adults. No study was conducted on workers.
on workers [32]. The symptoms considered in the studies were: eye irri-
Regarding studies with objective exposure [29, 32, 39, tation, sore throat/burning throat, nose irritation, gen-
45, 49–51], no consistent associations were found be- eral irritation symptoms, skin irritation/itchy eczema.
tween distance zones/exposure to NH3 and frequency of Six studies evaluated the occurrence of irritation
cold/flu, runny nose, nasal congestion and non-allergic symptoms objectively by distance zones [30, 39, 40, 45,
rhinitis [29, 39, 45, 49–51]. 49, 50]. Odour effects were found related to prevalence
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 17 of 21

of dry throat within the last 12 months [39, 49], nose ir- in subjects severely annoyed, and also in Horton et al.
ritation [39], and skin irritation [49]. [25], the latter association was consistent across odour
Regarding studies with subjective exposure [9, 24, 27, sources (livestock housings, slurry and manure, livestock
35, 38, 39, 41, 45, 51], significant findings were found for farming in general).
eye irritation/burning eye [9, 24, 27, 39, 45] and for sore
throat/dry throat/burning throat in five studies [9, 24, 27, Discussion
39, 45] (both odour tolerance and perception), for nose ir- This systematic review provides the state-of-art on the
ritation/burning nose in two studies [24, 27], for nose/eye health effects of odour from industrial sources. Meta-
irritation symptoms in one study [41], and for skin irrita- analysis results showed that residential odour exposure
tion/rash in three studies [24, 27, 38]. was associated to an increased risk of headache and
Thirteen studies considered that malodour may have an im- cough/phlegm, and to a borderline risk of nausea and
pact on mood [24, 25, 31, 33, 35, 37, 40, 44, 46, 47, 49–51]. vomiting. We found suggestive associations for the other
All studies were on adults. One study was on workers [33]. outcomes investigated (e.g. asthma, mucus irritation,
Six studies evaluated exposure objectively [37, 40, 46, mood states) but evidence is sparse. Only two studies
47, 49, 50]. Significant associations were only reported were carried out on occupational setting and they
for nervousness, and difficulty concentrating [49]. showed a statistically significant higher score of subject-
Nine studies evaluated exposure subjectively [24, 25, ive complaints, neurological complaints and post-
31, 33, 35, 37, 44, 47, 51]. Significant associations were traumatic stress symptoms in exposed workers than in
found for all mood outcomes in Horton et al. [25], for controls, and these associations persisted at least 3 years
nervousness, angriness, stress, unhappiness in Heaney after the pollution was removed [32, 33].
et al. [24], and for sadness and stress-related symptoms The associations with headache, cough/phlegm and
in Hooiveld et al. [35]. In Blanes-Vidal et al. [37], a nausea/vomiting have a biological plausibility. Unpleas-
dose-response association between odour annoyance and ant odours are able to modulate autonomic system re-
difficulty concentration was found. sponses, such as vagal nerve inducing nausea or
Considering the study on workers, participants in the vomiting [5]. Another mechanism involves stress, conse-
high odour score group reported a higher post-traumatic quent to environmental worry [18], and stress-related
score than those in the low odour score group, and these psychosomatic reactions such as chronic muscular
associations lasted for at least 3 years after the pollution tension, headaches, sleep disturbance. Chemicals re-
was removed [33]. sponsible for odour may cause irritation, supporting
Three studies evaluated the effects of odour on cardio- the higher risk for cough/phlegm. Eye and nose irrita-
vascular symptoms and blood pressure [28, 40, 45]. Each tion and asthma exacerbations can also be related to
unit of odour increase on an 8-point scale was associated this odour-related irritation but only limited evidence
with increases in diastolic blood pressure (mmHg) (OR = was found in this review. Our review confirms the
1.26; 95%CI: 1.08 to 1.47), but not in systolic blood strong association between odour and annoyance con-
pressure [28]. No significant association was found in firming the potential mediation role on odour-related
the other two studies [40, 45]. effects. We could not find any information on poten-
Ten papers [9, 25, 30, 34, 37, 41, 47, 49–51] investi- tial individual effect modifiers such as age, sex, educa-
gated odour nuisances in the population regarding to tional level [54].
their proximity to industries, odour perception, odour So far, only one other systematic review is available fo-
frequency or intensity, hedonic tone and NH3 exposure. cused only on exposure from Animal Feeding Operation
All studies were on adults. No study was carried out on proximity providing little evidence of association be-
workers. tween surrogate clinical outcomes and respiratory tract-
Regarding studies evaluating exposure objectively [30, related outcomes [55]. There is growing public attention
34, 37, 47, 49–51], odour annoyance showed a signifi- on the topic at an international level as documented by
cant association with odour frequency [47], with NH3 the non-negligible number of studies retrieved in this re-
concentration [37], as well as, with modelled odour ex- view. Nowadays, there is also an effort by a variety of
posure [34]. Moreover, three other studies showed a sig- countries to classify odour as an atmospheric pollutant
nificant increase in odour nuisances in the closest areas and regulate emissions by different policy frameworks
to the odour source [30, 49, 50]. worldwide [4].
Regarding studies evaluating exposure subjectively [9, Some limitations of our review should be mentioned.
25, 41], a significant dose–response association with Formal test for publication bias was not carried out due to
odour annoyance was found in Sucker et al. [41], con- the limited number of studies included in the meta-
sistent across the different exposure measure (odour fre- analysis, but we cannot exclude this kind of bias and pos-
quency, intensity, hedonic tone), aggravating the effect sibly other related biases (eg, language bias, citation bias,
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 18 of 21

multiple publication bias) [56]. However, we expect that affected by information bias. The European Standard
the comprehensive literature search, including grey litera- procedure for the measurement of odour concentration
ture, may have limited the impact of publications bias. uses a dynamic olfactometry assessed by a panel [59].
The inclusion of small studies (less than 100 subjects) in However, none of the studies adopted this measurement
our review suggests this bias is not a main concern. Meta- method, but two studies followed other systematic
analytical estimates are affected by a moderate degree of standard methods for the assessment of odour frequency
heterogeneity due to difference among studies in terms of through panellist testing and olfactometers [41, 47].
sources of exposure, population characteristics, study However, it should be considered that the methods for
length. An additional concern derives from the multiple assessing odour exposure should include also individual
hypothesis testing that increases the probability of false perceptions as effect modifier on odour impact on a
positive results due to the multiplicity phenomenon as population [12]. Odour perception, intolerance or an-
suggested also by other authors [55]. noyance or complaint [9, 35, 39, 40, 45] are adequate in-
Moreover, the associations between odour and head- dicators to this aim. Some of the included studies have
ache, nausea or cough need to be considered with cau- used distance as a proxy of odour exposure [30, 39, 40,
tion due to the overall low quality of the studies for 43, 45, 50, 51] and the Nettetal site studied in Steinhei-
methodological problems of the observational study der et al. [47]. In our results, no consistent evidence of
design. effects in the reporting of somatic symptoms was found
Most of included studies had a cross-sectional design by distance to the source of exposure. However, the bias
that can only provide a first hint of a hypothesized cause should be non-differential across outcomes leading in
of a disease, but not a proof of causality [57]. Six studies some cases to underestimate true associations. Another
used a panel approach, commonly used in air pollution exposure measure was the ammonia concentration in air
epidemiology [58], representing one of the best options [29, 37]. Although elevated levels of ammonia may cause
to study short-term health effects of odour although they irritative symptoms [60], the levels considered in the
can be affected by the drop-out bias and limited statis- studies are several orders of magnitude lower than ex-
tical power. posure limit in the workplace, 35 ppm for a short-term
We used the approach proposed by the US National (15-min) exposure limit in the workplace, about 3000
Toxicology Program [19, 20], one of the emerging ap- times higher than the maximum level reported in
proaches in the environmental (and occupational) health Blanes-Vidal [37] and 2 hundred times higher than max-
context, to evaluate the risk of bias of the body of evi- imum reported in van Kersen et al. [29]. However, due
dence. Overall, 15 out of the 29 studies had a high risk to the complexity of odour mixtures, the use of ammo-
of bias due to the limited confounding control, and ex- nia as a surrogate for odour pollution, as clearly stated
posure and outcome misclassification since most studies by the authors, represents a great limitation [37].
used self-reported information. On the contrary, five Included studies discussed the importance of using a
studies were at low risk of bias and the remaining nine standard objective method for exposure and outcome as-
showed an intermediate risk. sessment in environmental epidemiology [12, 33, 34, 37,
Regarding confounding, two aspects are worth of not- 41, 42], and some authors regarding other exposures
ing. In the present review, the most prevalent sources of mentioned dispersion modelling as a way out of this
odour were animal feeding operations and waste treat- methodological issue [61]. Boers et al. was the only study
ment sites. Therefore, exposure to air pollution from that used air dispersion modelling as proxy of exposure
these industrial activities can be common and adjust- [34]. Dispersion models include spatial characteristics
ment for concurrent environmental exposures is crucial (e.g. emissions, local meteorological conditions or topo-
to disentangle odour-related effects. Only few studies ad- graphical features, temperature, wind) which play a sig-
justed for these concurrent exposures such as noise, traf- nificant role in determining dispersion, concentration
fic, PM10, bioaerosol, pesticides [29, 35, 38, 41, 43], and intensity [61].
while another one stratified the study population to iso- Only 5 out of 30 studies used objective outcome mea-
late the odour-only exposed group [40]. One of the in- surements such as lung/bronchial function [27, 29], im-
cluded panel studies, without proper adjustment for mune function and allergy [23], blood pressure [28],
concurrent environmental exposures, was downgraded bronchial hyperresponsiveness to methacholine [42].
to a high risk of bias [26]. Another issue emerging from Most studies lacked medical objective assessments and
the review is that in many included studies, confounders generally depended on participants recall of symptoms
and co-occurring exposures were assessed by self-report. over different time periods that usually go from weeks to
Subjective exposure measures, such as odour rating over the last 12 months. If in the same study, both out-
and scores provided by participants, were used in most come and exposure were self-reported, it may have oc-
studies. Self-reported exposure is well known to be curred that exposed subjects, experiencing unpleasant
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 19 of 21

odours, were also more worried about their health and Supplementary Information
therefore more prone to the reporting of health symp- The online version contains supplementary material available at https://doi.
org/10.1186/s12940-021-00774-3.
toms than non-exposed subjects, creating the case for
differential misclassification of the outcome. Some stud- Additional file 1.
ies tried to reduce this bias by not mentioning odour when Additional file 2.
presenting the survey or by the use of memory aids to help Additional file 3.
remember symptoms [38, 39]. In addition, respondents
may be more likely to recall recent symptoms, also known
Acknowledgements
as seasonal bias, having difficulty in remembering past Authors would like to thank Zuzana Mitrova for her advices for the
events, related to the amount of time that has elapsed bibliographic search.
[30]. Response bias is a concern in most included surveys,
Authors’ contributions
both in terms of low participation rates and missing data VGF carried out the bibliographic search, the study selection, data extraction,
to specific questions. That is why, future studies should at- quality assessment and wrote the first draft of the manuscript. MDS helped
tempt to address this issue by ensuring adequate response in study selection, data extraction and contributed to the writing of the
manuscript. SV supervised the bibliographic search, the quality assessment
rates to the study or by controlling for non-response e.g. and contributed to the writing of the manuscript. LB did the statistical
by weighting methods [62]. analysis and the quality control of the database. CA obtained funding and
A recent European study collected all laws and regula- takes responsibility for the integrity of the data and the accuracy of the data
analysis and contributed to the writing of the manuscript. CA, MD and PM
tions in efforts toward the management of odour impact supervised the work and revised the manuscript. All authors have helped
in the communities, finding a heterogeneous picture (EU with data interpretation. The author (s) read and approved the final
Project D-NOSES). Europe has included odours in the manuscript.
European Directive on industrial emissions (Directive
Funding
2010/75/EU) but at national level, laws and environmen- The work was partially fund by the project “Italian Environment and Health
tal guidelines are in place only in some countries such as Network (RIAS)” – https://rias.epiprev.it/.
Italy (Legislative Decree 152/2006). However, no specific
public-health guidance is available. Wider considerations Availability of data and materials
The datasets used and/or analyzed during the current study are available as
of odour exposure are expected to increase with increas- Additional file 3.
ing urbanization [12], e.g., due to waste disposal sites or
intensive farming. It is clear that the effective prevention Declarations
and response to protect public health is a matter of ur-
Ethical approval and consent to participate
gency. Addressing the odour problem is also an equity Not applicable.
issue, since neighbouring residents of odour-polluted
sites are most likely low-income groups, as it happens Consent for publication
Not applicable.
for air pollution [63].
Competing interests
Conclusions The authors declare that they have no competing interests.
Findings from this systematic review underline the
Author details
public health importance of odour pollution for popu- 1
Department of Epidemiology of the Lazio Regional Health Service, ASL
lation living nearby industrial odour sources. The Roma 1 (Italy), Via Cristoforo Colombo, 112, 00147 Rome, Italy. 2Department
limited evidence for most outcomes supports the need of Preventive Medicine and Public Health, University of Valencia, 46010
Valencia, Spain.
for high quality epidemiological research to better
understand the association between odour pollution Received: 1 April 2021 Accepted: 22 July 2021
and its effects on human health. Exposure assessment
is crucial and should be improved to overcome the
References
lack of an objective and standardized method. Due to 1. Bax C, Sironi S, Capelli L. How can odors be measured? An overview of
the strong mediation by odour annoyance and lack of methods and their applications. Atmosphere. 2020;11(1):92. https://doi.org/1
evidence on individual effect modifiers, new studies 0.3390/atmos11010092.
2. Brancher M, Knauder W, Piringer M, Schauberger G. Temporal variability in
should include also these aspects, for example study- odour emissions: to what extent this matters for the assessment of
ing vulnerable groups, such as children or pregnant annoyance using dispersion modelling. Atmos Environ X. 2020;5:100054.
women, and workers. https://doi.org/10.1016/j.aeaoa.2019.100054.
3. Sucker K, Both R, Winneke G. Adverse effects of environmental odours:
Considering the growing efforts in regulating odour reviewing studies on annoyance responses and symptom reporting.
pollution, it is important to define standardized methods Water Sci Technol. 2001;44:43–51. https://doi.org/10.1016/j.
to estimate its effects on population health, and to pro- chemosphere.2016.11.160.
4. Brancher M, Griffiths KD, Franco D, de Melo Lisboa H. A review of odour
vide evidence-based guidance to bridge the gap also impact criteria in selected countries around the world. Chemosphere. 2017;
from a public health perspective. 168:1531–70. https://doi.org/10.1016/j.chemosphere.2016.11.160.
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 20 of 21

5. Dalton P. Odor, irritation and perception of health risk. Int Arch Occup Environ mood of nearby residents. Brain Res Bull. 1995;37(4):369–75. https://doi.
Health. 2002;75(5):283–90. https://doi.org/10.1007/s00420-002-0312-x. org/10.1016/0361-9230(95)00015-1.
6. Friedman DI, De Ver Dye T. Migraine and the environment. Headache J Head 27. Schinasi L, Horton RA, Guidry VT, Wing S, Marshall SW, Morland KB. Air
Face Pain. 2009;49(6):941–52. https://doi.org/10.1111/j.1526-4610.2009.01443.x. pollution, lung function, and physical symptoms in communities near
7. Nimmermark S. Odour influence on well-being and health with specific concentrated swine feeding operations. Epidemiology. 2011;22(2):208–15.
focus on animal production emissions. Ann Agric Environ Med. 2004;11(2): https://doi.org/10.1097/EDE.0b013e3182093c8b.
163–73. 28. Wing S, Horton RA, Rose KM. Air pollution from industrial swine operations
8. Shusterman D. Critical review: the health significance of environmental odor and blood pressure of neighboring residents. Environ Health Perspect. 2013;
pollution. Arch Environ Health. 1992;47(1):76–87. https://doi.org/10.1080/ 121(1):92–6. https://doi.org/10.1289/ehp.1205109.
00039896.1992.9935948. 29. van Kersen W, Oldenwening M, Aalders B, Bloemsma LD, Borlée F, Heederik
9. Shusterman D, Lipscomb J, Neutra R, Satin K. Symptom prevalence and D, et al. Acute respiratory effects of livestock-related air pollution in a panel
odor-worry interaction near hazardous waste sites. Environ Health Perspect. of COPD patients. Environ Int. 2020;136:105426. https://doi.org/10.1016/j.
1991;94:25–30. https://doi.org/10.1289/ehp.94-1567940. envint.2019.105426.
10. Sucker K, Both R, Winneke G. Review of adverse health effects of odours in 30. Kret J, Dalidowitz Dame L, Tutlam N, DeClue RW, Schmidt S, Donaldson K,
field studies. Water Sci Technol. 2009;59(7):1281–9. https://doi.org/10.2166/ et al. A respiratory health survey of a subsurface smoldering landfill. Environ
wst.2009.113. Res. 2018;166:427–36. https://doi.org/10.1016/j.envres.2018.05.025.
11. Jacob TJC, Fraser C, Wang L, Walker V, O’Connor S. Psychophysical 31. Hayes JE, Stevenson RJ, Stuetz RM. Survey of the effect of odour impact on
evaluation of responses to pleasant and mal-odour stimulation in communities. J Environ Manag. 2017;204(Pt 1):349–54. https://doi.org/10.101
human subjects; adaptation, dose response and gender differences. Int 6/j.jenvman.2017.09.016.
J Psychophysiol. 2003;48(1):67–80. https://doi.org/10.1016/s0167-8760(03 32. Tjalvin G, Lygre SH, Hollund BE, Moen BE, Bratveit M. Health complaints
)00020-5. after a malodorous chemical explosion: a longitudinal study. Occup Med.
12. Hayes JE, Stevenson RJ, Stuetz RM. The impact of malodour on 2015;65(3):202–9. https://doi.org/10.1093/occmed/kqu203.
communities: a review of assessment techniques. Sci Total Environ. 2014; 33. Tjalvin G, Mageroy N, Bratveit M, Lygre SH, Hollund BE, Moen BE. Odour as a
500–501:395–407. https://doi.org/10.1016/j.scitotenv.2014.09.003. determinant of persistent symptoms after a chemical explosion, a
13. Brattoli M, de Gennaro G, de Pinto V, Loiotile AD, Lovascio S, Penza M. longitudinal study. Ind Health. 2017;55(2):127–37. https://doi.org/10.2486/
Odour detection methods: Olfactometry and chemical sensors. Sensors. indhealth.2016-0155.
2011;11(5):5290–322. https://doi.org/10.3390/s110505290. 34. Boers D, Geelen L, Erbrink H, Smit LA, Heederik D, Hooiveld M, et al. The
14. Conti C, Guarino M, Bacenetti J. Measurements techniques and models to relation between modeled odor exposure from livestock farming and odor
assess odor annoyance: a review. Environ Int. 2020;134:105261. https://doi. annoyance among neighboring residents. Int Arch Occup Environ Health.
org/10.1016/j.envint.2019.105261. 2016;89(3):521–30. https://doi.org/10.1007/s00420-015-1092-4.
15. Gostelow P, Parsons SA. Sewage treatment works odour measurement. 35. Hooiveld M, van Dijk CE, van der Sman-De Beer F, Smit LAM, Vogelaar M,
Water Sci Technol. 2000;41(6):33–40. https://doi.org/10.2166/wst.2000.0090. Wouters IM, et al. Odour annoyance in the neighbourhood of livestock
16. Blanes-Vidal V, Suh H, Nadimi ES, Lofstrom P, Ellermann T, Andersen HV, farming - perceived health and health care seeking behaviour. Ann Agric
et al. Residential exposure to outdoor air pollution from livestock operations Environ Med. 2015;22(1):55–61. https://doi.org/10.5604/12321966.1141369.
and perceived annoyance among citizens. Environ Int. 2012;40:44–50. 36. Baldacci S, Maio S, Martini F, Silvi P, Sarno G, Cerrai S, et al. Odor annoyance
https://doi.org/10.1016/j.envint.2011.11.010. perception and health effects in an Italian general population sample. Eur
17. Moher D, Liberati A, Tetzlaff J, Altman DG, Altman D, Antes G, et al. Respir J. 2015;46:PA1115. https://doi.org/10.1183/13993003.congress-2015.
Preferred reporting items for systematic reviews and meta-analyses: the PA1115 European Respiratory Society.
PRISMA statement. PLoS Med. 2009;339(jul21 1):b2535. https://doi.org/10.113 37. Blanes-Vidal V. Air pollution from biodegradable wastes and non-specific
6/bmj.b2535. health symptoms among residents: direct or annoyance-mediated
18. Neutra R, Lipscomb J, Satin K, Shusterman D. Hypotheses to explain the associations? Chemosphere. 2015;120:371–7. https://doi.org/10.1016/j.
higher symptom rates observed around hazardous waste sites. Environ chemosphere.2014.07.089.
Health Perspect. 1991;94:31–8. https://doi.org/10.1289/ehp.94-1567952. 38. Wing S, Lowman A, Keil A, Marshall SW. Odors from sewage sludge and
19. Office of Health Assessment and Translation (OHAT), Division of the livestock: associations with self-reported health. Public Health Rep. 2014;
National Toxicology Program, National Institute of Environmental Health 129(6):505–15. https://doi.org/10.1177/003335491412900609.
Sciences. Handbook for Conducting a Literature-Based Health Assessment 39. Aatamila M, Verkasalo PK, Korhonen MJ, Suominen AL, Hirvonen MR,
Using OHAT Approach for Systematic Review and Evidence Integration. Viluksela MK, et al. Odour annoyance and physical symptoms among
https://ntp.niehs.nih.gov/ntp/ohat/pubs/handbookmarch2019_508.pdf. residents living near waste treatment centres. Environ Res. 2011;111(1):164–
Accessed June 2021. 70. https://doi.org/10.1016/j.envres.2010.11.008.
20. Rooney AA, Boyles AL, Wolfe MS, Bucher JR, Thayer KA. Systematic review 40. Herr CEW, Zur Nieden A, Kopka I, Rethage T, Gieler U, Eikmann TF,
and evidence integration for literature-based environmental health science et al. Assessment of somatic complaints in environmental health. Int J
assessments. Environ Health Perspect. 2014;122(7):711–8. https://doi.org/1 Hyg Environ Health. 2009;212(1):27–36. https://doi.org/10.1016/j.ijheh.2
0.1289/ehp.1307972. 008.01.003.
21. Deeks JJ, Altman DG, Bradburn MJ. Statistical methods for examining 41. Sucker K, Both R, Bischoff M, Guski R, Kramer U, Winneke G. Odor frequency
heterogeneity and combining results from several studies in meta-Analysis. and odor annoyance part II: dose - response associations and their
Syst Rev Health Care. London, UK: BMJ Publishing Group. 2008:285–312. modification by hedonic tone. Int Arch Occup Environ Health. 2008;81(6):
https://doi.org/10.1002/9780470693926.ch15. 683–94. https://doi.org/10.1007/s00420-007-0262-4.
22. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. 42. Radon K, Schulze A, Ehrenstein V, Van Strien RT, Praml G, Nowak D.
Stat Med. 2002;21(11):1539–58. https://doi.org/10.1002/sim.1186. Environmental exposure to confined animal feeding operations and
23. Avery RC, Wing S, Marshall SW, Schiffman SS. Odor from industrial hog respiratory health of neighboring residents. Epidemiology. 2007;18(3):300–8.
farming operations and mucosal immune function in neighbors. Arch https://doi.org/10.1097/01.ede.0000259966.62137.84.
Environ Health. 2004;59(2):101–8. https://doi.org/10.3200/AEOH.59.2.101-108. 43. Mirabelli MC, Wing S, Marshall SW, Wilcosky TC. Asthma symptoms among
24. Heaney CD, Wing S, Campbell RL, Caldwell D, Hopkins B, Richardson D, adolescents who attend public schools that are located near confined
et al. Relation between malodor, ambient hydrogen sulfide, and health in a swine feeding operations. Pediatrics. 2006;118(1):e66–75. https://doi.org/10.1
community bordering a landfill. Environ Res. 2011;111(6):847–52. https://doi. 542/peds.2005-2812.
org/10.1016/j.envres.2011.05.021. 44. Radon K, Peters A, Praml G, Ehrenstein V, Schulze A, Hehl O, et al. Livestock
25. Horton RA, Wing S, Marshall SW, Brownley KA. Malodor as a trigger of stress odours and quality of life of neighbouring residents. Ann Agric Environ
and negative mood in neighbors of industrial hog operations. Am J Public Med. 2004;11(1):59–62.
Health. 2009;99(Suppl 3):S610–5. https://doi.org/10.2105/AJPH.2008.148924. 45. Segala C, Poizeau D, Mace JM. Odors and health: a descriptive
26. Schiffman SS, Sattely Miller EA, Suggs MS, Graham BG. The effect of epidemiological study around a wastewater treatment plant. Rev Epidemiol
environmental odors emanating from commercial swine operations on the Sante Publique. 2003;51(2):201–14.
Guadalupe-Fernandez et al. Environmental Health (2021) 20:108 Page 21 of 21

46. Georgieff D, Turnovska T. Influence of odour immissions from cellulose-

paper industry on some health indicators. Folia Med. 1999;41:38–42.
47. Steinheider B, Both R, Winneke G. Field studies on environmental odors
inducing annoyance as well as gastric and general health-related
symptoms. J Psychophysiol. 1998;12(Suppl 1):64–79.
48. Steinheider B, Winneke G, Schlipkoter HW. Somatic and psychological
effects of malodours. A case study from a mushroom fertilizer production
plant. Staub. Reinhaltung der Luft. 1993;53:425–31.
49. Lipscomb JA, Goldman LR, Satin KP, Smith DF, Vance WA, Neutra RR. A follow-
up study of the community near the McColl waste disposal site. Environ
Health Perspect. 1991;94:15–24. https://doi.org/10.1289/ehp.94-1567957.
50. Deane M, Sanders G. Annoyance and health reactions to odor from
refineries and other industries in Carson, California 1972. Environ Res. 1978;
15(1):119–32. https://doi.org/10.1016/0013-9351(78)90085-3 Academic Press.
51. Deane M, Sanders G. Health effects of exposure to community odors from
pulp mills, Eureka, 1971. Environ Res. 1977;14(1):164–81. https://doi.org/10.1
016/0013-9351(77)90077-9.
52. Sucker K, Both R, Bischoff M, Guski R, Winneke G. Odor frequency and odor
annoyance. Part I: assessment of frequency, intensity and hedonic tone of
environmental odors in the field. Int Arch Occup Environ Health. 2008;81(6):
671–82. https://doi.org/10.1007/s00420-007-0259-z.
53. Ihlebæk C, Eriksen HR, Ursin H. Prevalence of subjective health complaints
(SHC) in Norway. Scand J Public Health. 2002;30(1):20–9.
54. Greenberg MI, Curtis JA, Vearrier D. The perception of odor is not a
surrogate marker for chemical exposure: a review of factors influencing
human odor perception. Clin Toxicol. 2013;51(2):70–6. https://doi.org/10.31
09/15563650.2013.767908.
55. O’Connor AM, Auvermann BW, Dzikamunhenga RS, Glanville JM, Higgins
JPT, Kirychuk SP, et al. Updated systematic review: associations between
proximity to animal feeding operations and health of individuals in nearby
communities. Syst Rev. 2017;6(1):86. https://doi.org/10.1186/s13643-017-04
65-z.
56. Sterne JAC, Sutton AJ, Ioannidis JPA, Terrin N, Jones DR, Lau J, et al.
Recommendations for examining and interpreting funnel plot asymmetry in
meta-analyses of randomised controlled trials. BMJ. 2011;343(jul22 1):d4002.
https://doi.org/10.1136/bmj.d4002.
57. Morgenstern H, Thomas D. Principles of study design in environmental
epidemiology. Environ Health Perspect. 1993;101(Suppl 4):23–38. https://doi.
org/10.1289/ehp.93101s423.
58. Dominici F, Sheppard L, Clyde M. Health effects of air pollution: a statistical
review. Int Stat Rev. 2003;71(2):243–76. https://doi.org/10.1111/j.1751-5823.2
003.tb00195.x.
59. Swedish Institute for Standards. Standard - Air quality - Determination of
odour concentration by dynamic olfactometry SS-EN 13725. 2003. https://
www.sis.se/en/produkter/environment-health-protection-safety/air-quality/
other/ssen13725ac2006/. Accessed June 2021.
60. Agency for Toxic Substances and Disease Registry of the U.S., Department
of Health & Human Services. Toxic Substances Public Health Statement:
Ammonia. 2004. https://www.atsdr.cdc.gov/toxprofiles/tp126-c1-b.pdf.
Accessed June 2021.
61. Hodgson S, Nieuwenhuijsen MJ, Colvile R, Jarup L. Assessment of exposure
to mercury from industrial emissions: comparing “distance as a proxy” and
dispersion modelling approaches. Occup Environ Med. 2007;64(6):380–8.
https://doi.org/10.1136/oem.2006.026781.
62. Tolonen H, Honkala M, Reinikainen J, Härkänen T, Mäkelä P. Adjusting for
non-response in the Finnish drinking habits survey. Scand J Public Health.
2019;47(4):469–73. https://doi.org/10.1177/1403494819840895.
63. Hajat A, Hsia C, O’Neill MS. Socioeconomic disparities and air pollution
exposure: a global review. Curr Environ Health Rep. 2015;2(4):440–50.
https://doi.org/10.1007/s40572-015-0069-5.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.