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, Study design	Industrial source	Study population, age group	Exposure assessment	Outcome assessment	Statistical analysis	Adjustment for confounders
Cross-sectional studies
Kret 2018 [30], USA, Cross-sectional	Waste (landfill)	N = 343 adults households within a 3.2-km radius (173 exposed; 170 non-exposed)	Distance (km)	Questionnaire: self-reported prevalence of diseases and 12 months symptoms; odour annoyance (5-point Likert scale) Groups: Odour nuisances Lower respiratory symptoms Upper respiratory symptoms Gastrointestinal symptoms Mucus irritation General ill feeling	Model: n.a Effect estimated: n.a Weighted prevalence (95%CI)	Matching for percentage of white population and for 25+ population with education level at least high school. No effect estimate.
Hayes 2017 [31], Australia, Cross-sectional	Wastewater treatment Plant	N = 153 residents within a 3-km radius on two exposed (with a history of high or low number of complaints) and one control sites	Questionnaire (presence/absence of bad smells and odours impacting community)	Questionnaire: Self-reported psychological symptoms past week; odour annoyance (10-point scale) Groups: Mood states	Model: ANOVA Effect estimated: None	Social readjustment scale by Holmes and Rahe 1967 added as covariate
Tjalvin 2015 [32], Norway, Cross-sectional	Chemical Industry (Chemical explosion in an Industrial harbour)	N = 284 workers in 2008 and 203 in 2012 (exposed workers employed in 2008 and/or clean-up workers, proximity to the explosion ≤1 km; control workers) range of age 18–67	Questionnaire: Workers exposure history	Subjective Health Complaints (SHC) score Groups: General ill feeling Gastrointestinal symptoms Upper respiratory symptoms Immune function Cardiovascular problems	Model: Linear mixed effects models with random intercept and slope Effect estimated: Mean difference	Age, gender, smoking habits, educational level.
Tjalvin 2017 [33], Norway, cross-sectional (repeated survey of Tjalvin 2015)	Chemical Industry (Chemical explosion in an Industrial harbour)	N = 486 workers employed in 2008 (18% present during the explosion), in 2010 (n = 379), 2012 (n = 252) Adults aged 18–67 years	Questionnaire: Low/high odour score (% of months each participant noticed the odour in 2008)	Questionnaire: Subjective Health Complaints (SHC) score previous month; Impact of Event Scale-Revised (IES-R) previous 7 days Groups: General ill feeling Mood states	Model: Linear mixed effects models with random intercept and slope Effect estimated: Mean difference	Age, gender, smoking habits, educational level, absence/presence during the explosion (>  1 km or ≤ 1 km)
Boers 2016 [34], Netherlands, Cross-sectional	Animal feeding operations	N = 582 residents living near livestock farms Mean age = 51 years old (SD 13) (part of the population included in Hooiveld 2015)	Calculated exposure: 98th percentile of odour concentrations (OUE/m3) from Stacks dispersion model	Questionnaire: Self-reported odour annoyance (4-point scale) Groups: Odour nuisances	Model: multivariate logistic regression analysis Effect estimated: ORs(95%CI)	Age, educational level, indoor air pollution, asthma, or lower back pain
Hooiveld 2015 [35], Netherlands, Cross-sectional	Animal feeding operations	N = 753 adults, residents with asthma or lower back pain	Questionnaire: Self-reported odour annoyance (yes/no)	Questionnaire: Self-reported symptoms last month; general health (5-point Likert scale from bad to very good) Groups: Gastrointestinal symptoms General ill feelings Lower respiratory symptoms Upper respiratory symptoms Mood states	Model: Multiple ordinal logistic, logistic and Poisson regression analysis. Effect estimated: ORs(95%CI)	Smoking status, growing up at farm, age, gender, nationality, marital status, educational level, asthma or lower back pain, other environmental annoyances (noise, traffic and air pollution)
Baldacci 2015 [36], Italy, Cross-sectional	Waste (incinerator)	N = 1407 residents within 4-km radius from the incinerator and a control group. Mean age 44.4 (SD 22.1)	Questionnaire: Self-reported odour annoyance (no, slightly annoying, very annoying)	Questionnaire: Self-reported symptoms past 12 months. Groups: Lower respiratory symptoms Upper respiratory symptoms	Model: Multivariate logistic regression analysis Effect estimated: ORs(95%CI)	Gender, age, residence/incinerator distance, educational level, working position, smoking status, passive smoking, residential time, occupational exposure.
Blanes-Vidal 2015 [37], Denmark, Cross-sectional	Waste (biodegradable)	N = 454 Residents from six study areas in Denmark. Mean age 54 (SD 14)	NH3 concentration: loge (NH3 exposure), NH3 exposure levels (< 2, 2–3, >  3 μg/m3), Questionnaire: Self-reported odour annoyance (no, slightly, moderately, very, extremely)	Questionnaire: Self-reported symptoms past 2 years, odour annoyance ((no, slightly, moderately, very, extremely) Groups: Odour nuisances Gastrointestinal symptoms General ill feeling Mood states	Model: multivariate logistic regression analysis. Effect estimated: ORs(95%CI)	Age, gender, smoking habit, job, time spent at home per week, existence of household residents below 18 years old, years living in the region, and acute and chronic respiratory conditions
Wing 2014 [38], USA, Cross-sectional	Sewage Sludge and Animal feeding operations	N = 158 adults, residents living near liquid TSS, 85 living near cake TSS, and 188 living in comparison areas	Questionnaire: Self-reported odour annoyance past six months (none/faint and moderate/strong/very strong)	Questionnaire: Self-reported symptoms past six months Groups: Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Skin disorders	Model: Linear and poisson regression analysis Effect estimated: Mean factor score differences (95%CI) and PRs (95%CI)	Age, gender, race, educational level, smoking status, passive smoking, agricultural chemical odours and odours from burning
Aatamila 2011 [39], Finland, Cross-sectional	Waste (Landfills and composting sites)	N = 1142 residents within a 5-km radius of six different biowaste sites Range of age: 25–64 years	Distance zone (< 1.5, 1.5–3, > 3 km) Questionnaire: odour perception (4-point scale) stratified into sensitive vs not sensitive, odour annoyance (4-point scale) categorized as annoyed vs not annoyed	Questionnaire: Self-reported symptoms past 12 months Groups: Gastrointestinal symptoms Mucus irritation General ill feeling Lower and upper respiratory symptoms Skin disorders	Model: Logistic regression analysis Effect estimated: ORs(95%CI)	Model 1: adjusted for sex, age, educational level, Socio economic level and smoking Model 2: additionally, adjusted for odour sensitivity
Herr 2009 [40], Germany, Cross-sectional	Waste (composting sites)	N = 477 residents living “near” two composting sites. (263 EnvExp2 and 214 control group). Individuals aged ≥16 years old	Distance (km): EnvExp2 (odour-only exposed) and a control group	Questionnaire: Self-reported symptoms past 2 years Groups: Gastrointestinal symptoms General ill feeling Lower respiratory symptoms Mood states Skin disorders Cardiovascular symptoms	Model: Logistic regression analysis Effect estimated: ORs(95%CI)	Adjusted for age, gender, and educational level
Sucker 2008 [41], Germany, Cross-sectional	Industrial sites	N = 1434 adults from each household (the homemaking or the person spending most of the time at home)	Questionnaire: Log-values of odour frequency Intensity (6-point scale from “very slight” to “extremely strong”), Hedonic tone (9-point scale with values ranging from “-4” “extremely unpleasant” through “0” “neither pleasant nor unpleasant” to “+ 4” “extremely pleasant”)	Questionnaire: Odour annoyance; self-reported health complaints Groups: Odour nuisances General ill feeling Mucus irritation	Model: Logistic regression analysis. Effect estimated: ORs(95%CI)	Noise disturbance, length of residence, quality of residential area, tenant or owner, single/multiple houses, average time at home, perceived health, smoking habit, gender, age, marital status, educational level
Radon 2007 [42], Germany, Cross-sectional	Animal feeding operations	N = 5556 Residents from four rural town with high density of AFOs Mean age 33.6 (SD 7.4)	Questionnaire: Self-reported odour annoyance (4-point Likert scale from “not at all” to “strongly”)	Questionnaire: Self-reported symptoms during the week. Clinical measurements: Specific IgE to common allergens > 0.35 IU/mL, bronchial hyperresponsiveness to methacholine, forced expiratory volume in 1 s (FEV1) Group: Lower and upper respiratory symptoms Immune function and allergy	Model: Logistic and linear regression analysis Effect estimated: ORs(95%CI)	Age, sex, active and passive smoking, educational level, number of siblings and parental allergies. FEV1 additionally, adjusted for passive smoking during childhood
Mirabelli 2006 [43], USA, Cross-sectional	Animal feeding operations	N = 58,169 students of 265 schools within 3 miles of at least one AFO source Range of age: aged 12–14	Questionnaire: self-reported indoor and outdoor odours from schools (binary coded variable, “reported”/“no reported”)	Questionnaire: Current and past 12-month self-reported respiratory symptoms and medical care Groups: Lower respiratory symptoms General ill feeling	Model: Random-intercepts binary regression analysis Effect estimated: PRs (95%CI)	Age, race, socioeconomic status, smoking, school exposures and household exposures
Radon 2004 [44], Germany, Cross-sectional	Animal feeding operations	N = 2745 Residents living in rural towns close to intensive animal production Mean age 32.7 (SD 7.7)	Questionnaire: Self-reported odour annoyance (4-point Likert scale from “not at all” to “extremely”)	QoL questionnaire: Physical SF-12 score, emotional SF-12 score Groups: General ill feeling Mood states	Model: Multiple linear regression analysis Effect estimated: β (SE)	Age, gender, respiratory symptoms, smoking, living on or close to a farm and employment status.
Segala 2003 [45], Canada, Cross-sectional	Wastewater treatment plant	N = 2867 residents from 8 nearby towns. Distance zones: 3–4.5 km (N = 1003), mean age 47.5 (SD 15.2) 1.5–3 km (N = 1007), mean age 48.2 (SD 67.7) <  1.5 km (N = 857), mean age 49.8 (SD 15.1)	Distance zones (< 1.5, 1.5–3, 3–4.5 km) Questionnaire: Self-reported odour tolerance (“tolerant”, “moderately tolerant”, “intolerant”), odour annoyance (“annoyed with impact on health”, “annoyed without impact on health”, “not annoyed”)	Questionnaire: Self-reported symptoms past month and year Groups: Gastrointestinal symptoms Mucus irritation Lower respiratory symptoms Upper respiratory symptoms General ill feeling Cardiovascular problems	Model: Multivariate logistic regression analysis Effect estimated: ORs(95%CI)	Age, sex, educational level, active vs inactive, smoking status, family size, satisfaction with neighbourhood life
Georgieff 1999 [46], Bulgaria, Cross-sectional	Paper industry	N = 538 Residents from Stamboliisky town Range of age: 16–60 years old	Questionnaire: Self-reported unpleasant odour (yes/no)	Questionnaire: Self-reported symptoms Groups: General ill feeling Lower respiratory symptoms Mood states Immune function and allergy	Model: n.a. Effect estimated: n.a. Percentages (%) of number of reported somatic symptoms	None
Steinheider 1998 [47], Germany, Cross-sectional	Nettetal study Fertilisers production plant Nörvenich study Pig rearing facilty	Nettetal study (N = 250) Nörvenich study (N = 322) Adults aged ≥18 years old	Nettetal study 1) Distance from the odour source Close: within 400–800 m Medium: 1600 m Far (control area): 6 and 3.5 km 2) 11-point graphic scale of Odour annoyance Nörvenich study 1) Log-values of odour frequency (odour hours/year). 34 observation points; 2) 11-point graphic scale of Odour annoyance	Questionnaire: Self-reported symptoms and odour annoyance (11-point graphic scale) Control variables (fever and asthma attacks) Groups: Odour nuisances Gastrointestinal symptoms General ill feeling Lower respiratory symptoms Mood states	Nettetal study Model: Analysis of variance Effect estimated: None Nörvenich study Model: Linear regression analysis. Effect estimated: β (SE)	None
Steinheider 1993 [48], Germany, Cross-sectional	Industrial sites 1) Duisburg- chemical plant 2) Dortmund – iron/steel plant 3) Brühl – castiron factory and sugar refinery 4) Rodenkirchen – oil refineries	N = 1539 adults, living near of four cities in North Rhine-Westphalia. Duisburg (N = 400), Dortmund (N = 400), Brühl (N = 539), Rodenkirchen (N = 200)	Log-values of odour frequency (odour hours/year).	Questionnaire: Self-reported odour annoyance (11-point scale) Groups: Odour nuisances	Model: Multivariate linear regression analysis Effect estimated: n.a.	Age, sex, educational level, profession, length of residence and perceived health. Dortmund, Brühl and Rodenkirchen added coping strategies to the model
Lipscomb 1991 [49], USA, Cross-sectional	Waste (McColl waste disposal site)	N = 193 residents living nearby a disposal waste site Adults ≥22 years old	Exposure areas (high, medium, and low) based on an odour survey conducted in 1981	Questionnaire: Self-reported symptoms past 12 months Groups: Odour nuisances Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Mood states Skin disorders Immune function and allergy	Model: n.a. Effect estimated: Crude PORs(95%CI)	None
Shusterman 1991 [9], USA, Cross-sectional	Waste	N = 2040 residents living near three hazardous waste sites in Southern California McColl. Acid petroleum sludge (N = 670) Operating Industries. Municipal and sewage (N = 514) Del Amo-Montrose. Residues from synthetic rubber manufacturing (N = 856)	Self-reported frequency of odour perception (“none”, “less than or equal to four times per month” and “greater than four times per month”	Questionnaire: Self-reported symptoms Groups: Odour nuisances Gastrointestinal symptoms Mucus irritation General ill feeling	Model: n.a. Effect estimated: PORs(95%CI)	None
Deane 1978 [50], USA, Cross-sectional	Refineries and other petrochemical industries	N = 291 Residents living in three residential areas nearby refineries and petrochemical plants	Exposure areas estimated by dynamic olfactometry: High (Area I), Moderate (Area II), Low (Area III).	Questionnaire: Self-reported symptoms Groups: Odour nuisances Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Mood states	Model: n.a. Effect estimated: n.a. Frequency of self-reporting outcomes	None
Deane 1977 [51], USA, Cross-sectional	Paper industry	N = 140 Adults living in three residential areas nearby a pulp mill	Exposure areas: high (1–2 miles southeast of the mills), moderate (2–3 miles east of the mils), low (4 miles east of the mills)	Questionnaire: Self-reported symptoms Groups: Odour nuisances Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Mood states	Model: n.a. Effect estimated: Frequency of self-reporting outcomes	Analysis were stratified by odour annoyance and gender
Panel studies
Van Kersen 2020 [29], Netherlands, Panel (3 months)	Animal feeding operations	N = 82 adults COPD non smokers residents in the eastern part of the province of Noord-Brabant and the northern part of the province of Limburg (high prevalence of lifestocks)	NH3 concentration (μg/m3), Questionnaire: Self-reported odour annoyance (no, yes)	Questionnaire: Self-reported symptoms past 12-h Clinical measurements: Lung function (forced expiratory volume or FEV1 and peak expiratory flow rate or PEF) Groups: Lower respiratory symptoms Upper respiratory symptoms	Model: Generalized estimated equations (GEE) assuming a first order autoregressive (AR1) correlation structure Effect estimated: ORs (95%CI)	Adjustment for ambient temperature, relative humidity and day-in-study (linear trend), PM10; Restriction to non-smokers by study design
Wing 2013 [28], USA, Panel (2 weeks)	Animal feeding operations	N = 101 non-smoking residents living within 1.5 miles of an CAFOs source Adults aged ≥18 years old. Mean age 53.7 (19.2–89.5)	Data-collection diary: Self-reported odour annoyance (9-point Likert scale)	Systolic (SBP) and diastolic (DBP) blood pressure values Groups: Cardiovascular problems	Model: Linear fixed-effects models Effect estimated: β (SE)	Time-of-day (AM or PM)
Heaney 2011 [24], USA, Panel (14 days)	Waste (landfill)	N = 23 adults, residents within 0.75 miles of the landfill	Questionnaire: 12-h of self-reported odour annoyance (5-point Likert scale)	Questionnaire: Self-reported symptoms past 12-h Groups: Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Mood states Skin disorders	Model: Conditional fixed effects logistic regression models Effect estimated: ORs(95%CI)	Time of day (AM/PM)
Schinasi 2011 [27], USA, Panel (14 days)	Animal feeding operations	N = 101 Non-smoking residents within 1.5 miles of an AFOs source Mean age 53.7 (19.2–89.5)	Questionnaire: 12-h of self-reported odour annoyance (9-point scale)	Questionnaire: Self-reported symptoms past 12-h Clinical measurements: Lung function (forced expiratory volume or FEV1and peak expiratory flow rate or PEF) Groups: Gastrointestinal symptoms Mucus irritation General ill feeling Lower respiratory symptoms Upper respiratory symptoms Skin disorders	Model: Conditional fixed effects logistic and linear regression analysis Effect estimated: β (SE)	Time of day (AM/PM)
Horton 2009 [25], USA, Panel (2 weeks)	Animal feeding operations	N = 101 Non-smoking residents within 1.5 miles of an AFOs source Mean age 53.7 (19.2–89.5)	12-h of self-reported odour annoyance (9-point scale)	Questionnaire: Self-reported information on mood states Groups: Mood states Odour nuisances	Model: Logistic mixed models with random intercepts. Effect estimated: β (SE) and ORs(95%CI)	Time of day (AM/PM)
Avery 2004 [23], USA, Panel (2 weeks)	Animal feeding operations	N = 15 residents within 2.4 km of an intensive hog operation facility Mean age 55.3 (SD 13.4).	Questionnaire: Self-reported odour annoyance (9-point scale, coded as a seven-level continuous variable)	Clinical measurements: Log salivary IgA concentration (μg/ml) and secretion rate (μ g/ml) Group: Immune function and allergy	Model: Hierarchical mixed models Effect estimated: β (SE)	Day of data collection (1–14) and time of day (AM/PM)
Schiffman 1995 [26], USA, Panel (4 days)	Animal feeding operations	N = 88 Exposure group,: Mean age 52.0 ± 13.4 Control group: Mean age 51.7 ± 8.3	Distance and duration: Exposed living an average of 5.3 + 6.5 years near hog operations and comparison group	Profile of Mood States (POMS) factors and Total Mood Disturbance (TMD) score Groups: Mood states	Model: ANOVA Effect estimated: n.a.	adjusted by design (matching by gender, age, race, and education)