Occupational health risks and working conditions of sanitation workers in Nepal: a cross-sectional study

Abstract

Background

The rapid increase in global consumption has led to a corresponding rise in waste production, intensifying the burden on sanitation workers (SW), whose work is of great importance to public health and prevents environmental pollution by ensuring proper waste disposal.

The aim of this study was to assess the working conditions and explore the health status of SW in Nepal, and to identify risk factors associated with occupational health issues.

Methods

A cross-sectional study was conducted among 737 SW included by convenience at a health camp for SW performed in five major municipalities of Nepal. A structured questionnaire was employed to collect data on sociodemographic factors, working conditions, and health issues. The SW were categorized into four groups based on type of work they performed, 1) collectors, who collect the garbage from the streets and load onto transport vehicles, 2) sweepers, who sweep the streets with brooms, 3) pickers, who sort through the garbage at the dumping sites, and 4) a reference group, that consisted of drivers and administrative staff without strenuous work or direct exposure to garbage. Data was introduced to STATA 18 and analyzed using Pearson’s chi-square test and logistic regression.

Results

The median age of participants was 32 years, and most participants were male, the educational level was low. Compared to the reference group, pickers had less access to personal protective equipment (PPE) and practiced poorer hygiene. Pickers experienced more symptoms related to dental problems and skin cuts, while collectors experienced more skin cuts and musculoskeletal problems. Sweepers reported lower rates of heat stress and occupational training than the remaining groups.

Conclusions

Pickers reported the highest risk of skin cuts and dental problems. Collectors reported higher risk of skin cuts and musculoskeletal problems. Pickers also reported the lowest rates of access to PPE, sanitary materials, and water, while collectors reported greater access to relevant vaccinations. These findings indicate unequal access to preventative measures across all types of SW.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-25194-7.

Introduction

Working conditions play a significant role in shaping an individual’s health, influencing the immediate risk of workplace injuries and the long-term development of physical and mental health issues [1]. Sanitation workers (SW) are essential to public health, as waste management helps prevent the spread of diseases and environmental contamination, but they risk their own health in return [2]. SW are subjected to physically strenuous labor and hazardous environments, which expose them to a range of health risks. Research has established that SW are especially vulnerable to respiratory, dermal, gastrointestinal, and various other health issues [2–4]. SW are often exposed to unsanitary conditions, harmful chemicals, and biological waste and as global consumption and urbanization increases, waste production is projected to increase from 2.01 billion tons to reach 3.4 billion tons annually by 2050, further intensifying the demands on waste management and the SW [5, 6]. This burden is especially severe in low-income countries like Nepal, where limited infrastructure, funding, and health frameworks can result in inadequate interventions to protect their health and safety [6].

There is limited comparative data assessing different sub-groups of SW in Nepal and other developing countries. Previous studies have shown for many years that sanitation work is filled with exposures that pose a health risk to the workers, like inflammation of the skin or respiratory system due to inhalation of bio-aerosols, and musculoskeletal pain due to carrying heavy loads and handling them at awkward positions [7]. Studies exploring the specific exposures faced by workers in developing countries, on the other hand, are scarce. However, detailed investigations into the specific exposures and vulnerabilities these workers face in low-income countries and how their health is impacted by this are increasing. Understanding the occupational hazards and health risks associated with sanitation work is essential for reducing occupational diseases and injuries.

This study aims to assess the occupational and general health conditions of SW in Nepal, and to identifying key factors contributing to these conditions. By examining the specific symptoms that SW experience, this study seeks to explore if and how these symptoms vary among different types of workers within the sector, as each type of SW performs distinct tasks in various environments and are thereby exposed to unique risks. Highlighting these differences can inform targeted interventions to protect their health.

Method

Study design and setting

This cross-sectional study was conducted using a questionnaire targeting SW in five municipalities in Nepal: Bharatpur, Janakpur, Nepalgunj, Biratnagar, and Pokhara. Data collection was conducted over a six-month period, from late February 2024 to early July 2024.

Types of SW

In this study, participants self-categorized into one of six occupational groups: sweepers, collectors, drivers, pickers, mix-type, and others.

Sweepers clean the public streets using brooms. Sweepers are not vulnerable to heavy lifting and awkward postures, but they are exposed to fast-paced repetitive movements, and they walk among the waste barefoot or in sandals [8].

Collectors gather waste from the streets and assist with loading and unloading transport vehicles. They may lift heavy bundles far above their head many times a day to put them onto the vehicles that are of various heights, resulting in possibly lifting several tons a day [8].

Drivers move the waste from the cities to landfills with various vehicles such as handcarts, trucks and tractors [8].

Pickers sort through the waste in the landfills with their bare hands and search for recyclable items that can be resold to scrap dealers [8]. They collect this scrap into large sacks, that they carry manually. Furthermore, they walk in the landfill among moving vehicles and heavy machinery operated by other workers. They walk barefoot or in sandals on top of heaps of waste. This makes them vulnerable to cuts and injuries. Since they walk among the waste and are in close contact with it, they may also breathe polluted air for prolonged periods, without the use of face masks, exposing them to prolonged inhalation of polluted air. Alongside this, pickers and collectors are exposed to dust from the waste and risk of infection through their close contact with the waste [9].

Sweepers, collectors, and drivers usually work in the morning hours, whereas pickers work throughout the day.

Furthermore, participants could categorize themselves as mix-types, which meant they performed a combination of tasks from two or more of the aforementioned categories. These participants were not asked to specify what tasks they performed during their workday, making it difficult to assess their specific exposures and the extent of those.

Besides these, there was also an “others” group for participants who did not belong in any of the aforementioned categories. Most of the participants who categorized themselves as “others” specified that they work as administrative or maintenance staff for the municipality waste management. These workers do not perform waste handling and have no direct contact with the waste but are employed by the municipality for other tasks related to the management of waste.

Inclusion and exclusion criteria

This study intended to include all SW employed by the five selected municipalities: Janakpur, Nepalgunj, Biratnagar, Pokhara, and Bharatpur. Inclusion was done via an open invitation to SW at the municipalities to participate by simply attending the health camp. SW who were absent from work, on leave, or severely ill were excluded from participation. Exclusion was passive and based on availability as SW who were absent, on leave, or severely ill were not on site of the municipalities during the health camps. As informal workers were also included, the exact number of eligible participants remains unknown. As a result, the response-rate could not be determined. Each interview lasted approximately one hour and included a questionnaire and a physical examination. No follow-up assessments were conducted.

Sample size

The required sample size to achieve an optimal level of precision was calculated using Cochran’s sample size formula, n = z²pq/I². The z-value is a set value based on the desired confidence interval (CI), where we chose 95%, therefore, the corresponding z-value was 1.96. p estimates the proportion of the population that is eligible for inclusion in the study, which we assumed to be 50%, which equals p = 0.5. q is equal to 1-p, which in this case is 0.5. I is the tolerated margin of error, in other words, the desired precision. We chose to tolerate a 5% margin of error, therefore I = 0.05. By using this formula, the minimum sample size needed for this study was 385.

Participants

A total of 783 SW participated in the study: 305 sweepers (39%), 166 collectors (21.2%), 150 drivers (19.2%), 57 pickers (7.3%), 46 mixed-type workers (5.9%), and 59 other workers (7.5%), which included maintenance and administrative staff. In ambiguous cases, the participant was classified as mixed-type or other, depending on their own classification. The mixed-type category mostly included individuals who identified with multiple roles such as sweeper, collector, driver, or picker, while the other category mostly consisted of participants working at the waste facility in non-sanitation roles, such as administrative staff and office cleaners. Thus, the proportion of each category emerged naturally. This subjective self-classification introduces the risk of bias due to misclassification, which could potentially mitigate the findings of this study. The 46 mixed-type workers were excluded from data analysis, due to insufficient detail on what tasks they performed and to what degree, making it unclear whether they should be counted as part of another group, or as a separate worker category. The risk of bias due to misclassification and response bias due to the open invitation recruitment, means that these proportions do not necessarily reflect the real proportions of SW in each category. Especially pickers tend to be informal workers, which means they may be underrepresented in this study. This resulted in a sample size of 737 SW for data analysis. For the analysis, drivers, and the “others” group were combined to form a reference group (N = 209, 28.4%), based on the assumption that they experience the least strenuous working conditions and the least dangerous environment of the SW and therefore experience the least health issues of the SW groups. We do not assume that this can act as a control group on par with the background population, as participants of this group still work in proximity to the waste, but since they do not handle it directly and mostly sit down during their workday, we assume that they experience fewer symptoms due to waste handling and strenuous physical tasks.

Data collection

Data was collected using a structured questionnaire. The questionnaire was developed and deployed using REDCap. The questionnaire was developed in English and translated to Nepali then retranslated into English for data analysis. The questionnaire was developed specifically for this study and the companion study and was not adapted from any previously validated tool. Research experts were consulted to verify that the questionnaire was in accordance with research objectives, and it was pretested to ensure feasibility and applicability of the tool. However, it was not pretested or validated for linguistic and cultural nuances before or after translation into Nepali and back to English. The self-reported questionnaire was facilitated by enumerators, with backgrounds in health professions. Following the questionnaire, participants underwent a physical examination conducted by a medical doctor. Both enumerators and doctors investigated symptomatic history, but the doctors also conducted a physical examination to validate current symptoms. Based on these examinations, the attending doctor completed a separate health assessment form for each participant. The data from the physical examination was exclusively used in the companion study. Both enumerators and medical doctors were oriented on the objective of the study, tools, and techniques of data collection. Each participant was only interviewed by one enumerator, followed by one medical doctor. The questionnaire contained a variety of questions, some were yes/no, some were multiple-choice, and others were free text items. There were no comment boxes in which enumerators could elaborate or explain data. The questionnaire sought to explore symptoms from all organ systems. The questionnaire was programmed with conditional branching logic, such that only SW who responded affirmatively that they had experienced symptoms related to each organ system, would then be asked the follow-up questions, detailing specific symptoms and frequency of symptoms. We sought to explore as many aspects of working conditions and environment as possible to identify trends within or across groups. This included various themes like specifications of the tools and protective equipment the SW were provided and hygiene practices during and after the workday. In this study, we focused on health problems and the general working conditions to get an overall indication of what conditions might be considered for more detailed research later. The questionnaire itself is available as a supplementary file.

Data processing

The data from these questionnaires was also used in the companion study: “Musculoskeletal disorders and other occupational health outcomes among sanitation workers in Nepal: A community based cross-sectional survey exploring the risk factors, knowledge, and practices”, for which the questionnaire was originally developed. That study reports the findings of the objective physical examination part of the study, whereas this study focused on the self-reported health and occupational factors.

Data was screened for outliers and logical inconsistencies. Non-responses, outliers, and logical inconsistencies may be due to misunderstandings, errors, or refusal to respond to certain questions. These, along with any ambiguity, were counted as missing data and was excluded from analysis using pairwise deletion. Missing data were assumed to be missing completely at random, as there were no apparent trends as to what themes or participants contained missing data.

Study variables

For this study, the data collected from the questionnaire has been dichotomized.

For the sociodemographic data, the median age of the participants was 32 years. The data was then dichotomized into 32 years and below, or 33 years and above. The same procedure was done for the duration of employment as SW, in which the median duration was found to be 5 years. The median age at which participants started working as SW was 23 years. Starting employment as a SW before the age of 5 was excluded from this analysis, as it indicated that the participant may have misunderstood the question. The cut-off of 5 years of age was chosen as some SW might have been brought as children to help their parents, and we consider children less than 5 years of age to be incapable of performing the work, though they may still have been brought to the site by their parents before the age of 5. This question only refers to the age at which the participants started work as SW, not their current age, as the youngest participant was 15 years of age.

For marital status, single, divorced, and separated were grouped together as “unmarried”. For income sufficiency, the responses “insufficient” and “barely sufficient” were grouped together as “insufficient”, whereas the responses “sufficient” and “enough sufficient” were grouped together as “sufficient”. Alcohol and tobacco use were already dichotomized as yes/no responses, but tobacco was stratified by chewing or smoking tobacco. In this analysis, chewing and smoking tobacco have been grouped together, so that a yes-response in either chewing or smoking resulted in being analyzed as affirmative, even if the participant only used one of the two.

The ethnicity and religion of participants was not dichotomized, instead the overall characteristics of the participants was simply reported descriptively in the main text.

Data on working conditions (Table 2) were binary yes/no responses. So were the responses regarding whether the participants worked on temporary or permanent contracts and what PPE they used. The responses detailing where the participants received their PPE if not from their employer, were provided by free-text boxes. The median duration of occupational training was one day, but this was not stratified into above or below, as this was reported in the main text only. For the daily working hours and the amount of weekly working days, the results were stratified into above or below the amount that was determined by Nepali law.

Table 2.

Working conditions of sanitation workers in the five regions of Nepal

	Crude OR	[95% conf.	Interval]	P-value	Adjusted OR	[95% conf.	Interval]	P-value
	Do you have access to any PPE/protective gears?
Sweeper (N = 302)	1.18	0.81	1.72	0.382	1.42	0.89	2.27	0.137
Collector (N = 164)	1.33	0.85	2.07	0.210	1.34	0.83	2.14	0.228
Picker (N = 55)	0.43	0.24	0.79	0.006	0.48	0.25	0.91	0.024
	Have you been trained in any type of occupational safety during your stay?
Sweeper (N = 304)	0.90	0.62	1.30	0.558	0.61	0.38	0.97	0.037
Collector (N = 166)	0.81	0.52	1.25	0.337	0.79	0.50	1.25	0.318
Picker (N = 55)	0.61	0.31	1.18	0.142	0.52	0.26	1.07	0.074
	Are you entitled to any health insurance by your institution?
Sweeper (N = 301)	1.10	0.73	1.64	0.652	1.03	0.62	1.72	0.902
Collector (N = 163)	0.79	0.48	1.28	0.336	0.92	0.54	1.155	0.744
Picker (N = 47)	0.28	0.09	0.81	0.019	0.34	0.11	1.01	0.052
	Have you been provisioned with a tetanus vaccine?
Sweeper (N = 305)	1.03	0.73	1.47	0.849	1.16	0.75	1.78	0.511
Collector (N = 166)	1.84	1.21	2.79	0.004	1.87	1.21	2.91	0.005
Picker (N = 57)	1.03	0.57	1.85	0.915	1.04	0.56	1.94	0.895
	Have you been provisioned with a hepatitis B vaccine?
Sweeper (N = 305)	1.03	0.49	2.18	0.939	0.91	0.37	2.23	0.831
Collector (N = 166)	2.51	1.20	5.23	0.014	2.48	1.15	5.35	0.020
Picker (N = 57)	0.29	0.04	2.30	0.243	0.26	0.03	2.14	0.212
	Does your municipality provide sanitary materials like soap and water?
Sweeper (N = 304)	1.36	0.95	1.95	0.090	1.55	0.99	2.41	0.053
Collector (N = 166)	0.64	0.42	0.99	0.046	0.69	0.44	1.09	0.116
Picker (N = 50)	0.24	0.10	0.57	0.001	0.28	0.12	0.67	0.004
	Is drinking water easily available at the workplace?
Sweeper (N = 305)	1.07	0.73	1.57	0.733	1.13	0.71	1.81	0.598
Collector (N = 164)	0.66	0.43	1.02	0.062	0.72	0.46	1.14	0.157
Picker (N = 57)	0.40	0.22	0.73	0.003	0.48	0.25	0.91	0.024
	Do you eat at work?
Sweeper (N = 304)	0.94	0.66	1.34	0.732	1.04	0.67	1.56	0.874
Collector (N = 165)	0.96	0.64	1.45	0.857	1.02	0.66	1.58	0.930
Picker (N = 54)	2.17	1.17	4.02	0.014	2.45	1.28	4.69	0.007
	Do you wash your hands with soap/sanitizer before eating anything?
Sweeper (N = 303)	0.94	0.52	1.71	0.847	0.97	0.47	1.99	0.926
Collector (N = 164)	1.07	0.53	2.16	0.854	1.05	0.50	2.19	0.902
Picker (N = 56)	0.27	0.13	0.56	0.001	0.29	0.13	0.66	0.003
	Are there shadows or a cool place to rest during breaks?
Sweeper (N = 305)	1.18	0.83	1.69	0.355	1.14	0.74	1.76	0.560
Collector (N = 164)	0.90	0.60	1.36	0.630	0.87	0.56	1.34	0.518
Picker (N = 56)	1.33	0.73	2.44	0.353	1.41	0.74	2.69	0.293

For certain binary outcomes regarding working conditions we fitted a univariate and multivariable logistic regression model with occupational group as the main exposure and adjusted for age, sex, education, in come sufficiency, duration of employment, alcohol, and tobacco. The reference category was the reference group consisting of transporters and administrative workers. Significant results are shown in bold. This data shows pickers had significantly less access to PPE and health insurance. Collectors reported an increased tendency to having received relevant vaccines. Both collectors and pickers had less access to sanitary materials than the reference group and sweepers. Pickers also reported higher risk of eating while working and not washing their hands before eating

Data on the prevalence of all health issues (Table 3) were categorized as binary yes/no responses.

Table 3.

Health risks of sanitation workers in the five regions of Nepal

	Crude OR	[95% conf.		Interval]	P-value	Adjusted OR	[95% conf.	Interval]	P-value
	Have you experienced occupational health risks within the last year?
Sweeper (N = 305)	1.08	0.73		1.60	0.706	0.85	0.52	1.38	0.604
Collector (N = 165)	1.29	0.83		2.02	0.258	1.35	0.84	2.17	0.220
Picker (N = 55)	2.06	1.11		3.81	0.021	1.85	0.95	3.59	0.068
	Have you experienced skin problems within the last year?
Sweeper (N = 305)	0.87	0.48		1.58	0.648	0.82	0.39	1.72	0.601
Collector (N = 166)	1.59	0.85		2.95	0.144	1.66	0.85	3.24	0.136
Picker (N = 57)	1.90	0.84		4.32	0.123	1.95	0.80	4.74	0.141
	Have you experienced eye problems within the last year?
Sweeper (N = 305)	1.09	0.60		1.99	0.778	0.98	0.45	2.12	0.956
Collector (N = 166)	0.78	0.37		1.66	0.516	0.87	0.39	1.97	0.742
Picker (N = 57)	2.13	0.93		4.88	0.074	2.07	0.81	5.31	0.131
	Have you experienced musculoskeletal problems within the last year?
Sweeper (N = 305)	1.77	1.07		2.93	0.027	1.31	0.71	2.42	0.383
Collector (N = 166)	1.90	1.08		3.33	0.026	1.89	1.05	3.42	0.035
Picker (N = 57)	1.96	0.92		4.20	0.083	1.61	0.71	3.63	0.250
	Have you experienced dental problems within the last year?
Sweeper (N = 305)	1.24	0.41	3.76		0.703	0.84	0.22	3.22	0.795
Collector (N = 166)	1.27	0.36	4.45		0.712	1.59	0.43	5.94	0.487
Picker (N = 57)	5.71	1.74	18.75		0.004	5.84	1.55	22.02	0.009
	Have you experienced gastrointestinal problems within the last year?
Sweeper (N = 305)	1.17	0.58	2.38		0.661	0.65	0.26	1.62	0.359
Collector (N = 166)	0.66	0.26	1.70		0.394	0.53	0.19	1.48	0.225
Picker (N = 57)	1.14	0.36	3.63		0.827	0.78	0.22	2.74	0.703
	Have you experienced respiratory problems within the last year?
Sweeper (N = 305)	1.03	0.45		2.34	0.945	0.52	0.19	1.47	0.220
Collector (N = 166)	0.88	0.33		2.35	0.793	0.72	0.26	2.03	0.535
Picker (N = 57)	2.34	0.81		6.74	0.115	1.23	0.37	4.13	0.733
	Have you experienced heat stress symptoms within the last year?
Sweeper (N = 305)	0.84	0.48		1.48	0.550	0.44	0.21	0.92	0.028
Collector (N = 166)	1.06	0.56		1.99	0.866	1.07	0.54	2.10	0.844
Picker (N = 57)	2.75	1.33		5.69	0.006	1.93	0.86	4.36	0.112
	Have you suffered from skin cuts within the last 6 months?
Sweeper (N = 303)	1.26	0.77		2.05	0.355	1.49	0.85	2.95	0.164
Collector (N = 166)	6.87	4.19		11.25	< 0.001	6.29	3.74	10.56	< 0.001
Picker (N = 57)	8.57	4.45		16.52	<0.001	8.75	4.32	17.71	< 0.001
	Have you suffered from needle sticks within the last 6 months?
Sweeper (N = 26)	0.60	0.17		2.06	0.416	0.64	0.13	3.05	0.571
Collector (N = 23)	2.11	0.63		7.06	0.224	2.09	0.46	9.45	0.337
Picker (N = 10)	3.79	0.75		19.04	0.106	2.77	0.41	18.52	0.294
	Have you had any body pain within the last 6 months that made it difficult to work?
Sweeper (N = 59)	0.40	0.12		1.33	0.136	0.29	0.07	1.32	0.112
Collector (N = 35)	0.92	0.23		3.67	0.907	1.02	0.22	4.67	0.980
Picker (N = 12)	2.10	0.21		21.10	0.530	1.53	0.13	17.34	0.733
	Have you suffered a bone or ligament fracture in the last 6 months?
Sweeper (N = 303)	1.54	0.47		5.06	0.479	2.55	0.54	12.02	0.236
Collector (N = 165)	2.23	0.64		7.74	0.208	3.48	0.83	14.58	0.087
Picker (N = 56)	0.91	0.10		8.34	0.936	1.50	0.14	16.21	0.737

For each binary outcome regarding health we fitted a univariate and multivariable logistic regression model with occupational group as the main exposure and adjusted for age, sex, education, income sufficiency, duration of employment, alcohol, and tobacco. The reference category was the reference group consisting of transporters and administrative workers. Significant results are shown in bold. In the questionnaire, the term “occupational health risks” is the umbrella term that includes all the symptoms that are reported in the rest of the table, which makes “occupational health risks” the overall result showing if a group is more or less vulnerable. This data shows that pickers had an increased risk of occupational health issues overall, compared to the other groups, specifically heat stress, dental problems, and skin cuts. Collectors reported a significantly higher risk of musculoskeletal problems and skin cuts. Sweepers reported a significantly lower risk of heat stress compared to the other group

Statistical analysis

Data analysis was conducted using STATA 18 and compared the types of SW to determine whether significant disparities existed between the groups in social circumstances, working conditions, and health issues. Pearson’s chi-square test was used for binary outcomes to compare the socioeconomic factors of the groups, with significance set at p < 0.05. Logistic regression was conducted to explore health risks and working conditions among the different types of SW. The aim was to determine whether there were significant differences between the groups to provide insight that can help strategize targeted interventions. Crude odds ratio (OR) was calculated by using univariate logistic regression and adjusted OR was calculated using multivariate logistic regression, adjusting for sex, age, level of education, income sufficiency, duration of employment, and use of alcohol and tobacco as confounders. These confounders were selected due to their well-documented influence on health status and development through various outcomes [10–12].

Ethical considerations

The recruitment of participants and data collection was approved by the Nepal Health Research Council (NHRC). Written informed consent was obtained from all participants. For illiterate participants, enumerators read the consent form aloud and obtained a thumbprint from the participant instead of a signature. All participants were informed about the study’s objectives, data collection methods and potential risks to the participants. They were also informed of confidentiality and anonymity, and of their right to withdraw consent at any time. For participants between the ages of 15 and 18, written informed consent was obtained from a legally authorized adult representative and from the minor themselves.

Results

Sociodemographic characteristics

A total of 737 SW were included in the data analysis for this study. Non-response rates were not recorded as recruitment was conducted by open invitation.

Most participants were male, and all groups were predominantly male, except sweepers (Table 1). Most participants also belonged to lower castes. The median age of participants was 32 years, and collectors and pickers consisted of a higher percentage of younger SW, compared to sweepers and the reference group (Table 1). Most participants did not attend school beyond 5th grade, but the reference group participants reported a significantly higher rate of longer education compared to all other groups (Table 1). Approximately half of participants reported that their income was insufficient to meet their needs, although there was no significant difference between the groups (Table 1).

Table 1.

Sociodemographic characteristics of sanitation workers in the five regions of Nepal

	Reference (n = 209)	Collector (n = 166)		P-value	Sweeper (n = 305)		P-value	Pickers (n = 57)	P-value
Sex
Male	188 (90%)	163 (98.2%)		0.001	142 (46.6%)		< 0.001	42 (73.7%)	0.001
Female	21 (10%)	3 (1.8%)			163 (53.4%)			15 (26.3%)
Age
< 32	96 (45.9%)	112 (67.5%)		< 0.001	144 (47.2%)		0.775	39 (68.4%)	0.003
33+	113 (54.1%)	54 (32.5%)			161 (52.8%)			18 (31.6%)
Level of education
< Grade 5	92/207 (44.4%)	103 (62%)		0.001	258 (84.6%)		< 0.001	43 (75.4%)	< 0.001
Grade 6+	115/207 (55.6%)	63 (38%)			47 (15.4%)			14 (24.6%)
Marital status
Married	185/208 (88.9%)	120/165 (72.7%)		< 0.001	260/304 (85.5%)		0.260	43 (75.4%)	0.009
How long have you been a sanitation worker?
< 5 years	117/202 (57.9%)	97/165 (58.8%)		0.867	143/302 (47.4%)		0.020	36 (63.2%)	0.478
6 + years	85/202 (42.1%)	68/165 (41.2%)			159/302 (52.6%)			21 (36.8%)
Is your family income sufficient for food and daily activity?
Insufficient	95 (45.5%)	83/162 (51.2%)		0.269	163/302 (54%)		0.058	35/56 (62.5%)	0.023
Sufficient	144 (54.5%)	79/162 (48.8%)			139/302 (46%)			21/56 (37.5%)
Are you drinking alcohol?
Yes	101/208 (48.6%)		89 (53.6%)	0.331		80/304 (26.3%)	< 0.001	28 (49.1%)	0.940
Do you use tobacco (chewing and/or smoking)?
Yes	130/208 (62.5%)		114 (68.7%)	0.213		113/303 (37.3%)	< 0.001	39 (68.4%)	0.410

Table of sociodemographic characteristics stratified by type of sanitation worker compared to reference group. Analyzed with Pearson’s chi-square test. Data shown in total numbers, followed by percentage in parentheses. When data was missing, the amount of available data is specified in the particular box. The p-value reflects the difference between the reference group and each of the other groups separately, meaning that differences with a p-value < 0.05 are less likely to be coincidental. Significant disparities between the groups is shown in bold. This data shows that the sweepers are the only group not dominated by male workers, and they also report lower rates of alcohol and tobacco use compared to the other groups. The reference group had significantly higher levels of education, while the pickers reported significantly lower income sufficiency

Sweepers reported the least use of alcohol and tobacco (Table 1).

The median age of participants was 32 years (range: 15–75 years). The median age at which participants started working as SW was 23 years (range: 5–68 years). The median duration of employment as SW was 5 years (range: <1–44 years).

In Nepal, the most common ethnicity is Chhetri-Brahman with 27.8% of the general population belonging to this group, which is regarded as a high-ranking caste [13, 14]. However, among the participants of this study only 7.1% belonged to this group. Rather, 27.3% of the reference group and 51.8% of pickers consisted mostly of Janajati, while 35.6% of collectors and 53.3% of sweepers were Dalits, both of which are regarded as lower castes [14].

The religious majority of Nepal is Hindu, making up 81.2% of the population, while the 93.2% of the participants of this study are Hindu [13]. The remaining participants identified as Buddhist, Christian, Muslim, or other.

15.9% of the reference group were formally employed on permanent contracts, the same applied to 8.5% of collectors, 24.7% of sweepers and 5.4% of pickers. The rest of the participants were employed informally or by temporary contracts.

Additional sociodemographic characteristics of the participants are presented in Table 1.

Risk assessment for hazardous working conditions

Key findings on working conditions are presented in Table 2. The main findings of this analysis show that pickers had significantly less access to PPE and health insurance. Collectors reported an increased tendency to having received relevant vaccines. Both collectors and pickers had less access to sanitary materials than the reference group and sweepers. Pickers also reported higher risk of eating while working and not washing their hands before eating (Table 2).

Collectors and pickers reported that 69.9% and 42.1% of the participants had access to gloves, respectively. Furthermore, 40.4% of pickers had access to masks, and 15.8% had access to eye protection, whereas 69.3% of collectors had access to masks and 22.9% had access to eye protection. The sweepers and reference group displayed PPE access levels similar to the collectors. Differences in access to gloves and masks between pickers and the reference group were statistically significant, whereas the difference in access to eye protection was not.

The participants who received PPE from other organizations than their employer primarily received them from various NGOs, but a few SW reported self-providing PPE.

Among participants who received training on occupational safety procedures, the median duration of the training was 1 day (range: 1->10).

Nepali labor law mandates a maximum 8-hour workday and at least one weekly day off [15]. Despite this, 75.5% of references, 72.6% of collectors, 76.5% of sweepers, and 59.7% of pickers reported working 7 days a week. This difference was significant between pickers and the reference group. Additionally, 42.5% of the reference group, 50.9% of collectors, 5.9% of sweepers, and 33.3% of pickers, worked more than 8 h per day. Statistical analysis revealed the difference in sweepers’ daily working hours compared to the reference group to be significant.

Risk and prevalence of occupational complaints

The OR of work-related health problems across the three groups of SW relative to the comparison group are presented in Table 3 while the prevalences within each group is stated below. The main findings of the logistic regression present that pickers had an increased risk of occupational health issues overall, compared to the other groups, specifically heat stress, dental problems, and skin cuts. Collectors reported a significantly higher risk of musculoskeletal problems and skin cuts. Sweepers reported a significantly lower risk of heat stress compared to the other group (Table 3).

Musculoskeletal complaints were most common among pickers (21.1%, P = 0.079), followed by collectors (20.5%, P = 0.024) and sweepers (19.3%, P = 0.026), while only 12% of the reference group reported such problems. Dental issues showed a similar pattern, affecting 12.3% of pickers (P = 0.001), compared to 3% of sweepers (P = 0.702) and collectors (P = 0.711) and 2.4% of the reference group.

Skin cuts were especially frequent among collectors (54.2%, P = < 0.001) and pickers (59.7%, P = < 0.001), whereas sweepers (17.8%, P = 0.355) and the reference group (14.7%) reported much lower rates. Other skin-related symptoms were reported by 8.9% of sweepers (P = 0.647), 15.1% of collectors (P = 0.142), 17.5% of pickers (P = 0.118), and 10.1% of the reference group.

Eye problems were reported most often by pickers (17.5%, P = 0.070), with lower frequencies among sweepers (9.8%, P = 0.778), collectors (7.2%, P = 0.515), and the reference group (9.1%). Gastrointestinal symptoms were reported by 7.2% of sweepers (P = 0.661), 4.2% of collectors (P = 0.391), 7% of pickers (P = 0.827), and 6.2% of the reference group. Respiratory complaints were uncommon overall, but still highest among pickers (10.5%, P = 0.106), compared with 4–5% in the other groups (sweepers P = 0.945, collectors P = 0.793).

Heat stress affected about one-quarter of pickers (26.3%, P = 0.005), compared with 9.8% of sweepers (P = 0.550), 12.1% of collectors (P = 0.866), and 11.5% of the reference group. Serious injuries such as ligament tears or bone fractures were rare, ranging from 1.8% among pickers (P = 0.936) to 4.2% among collectors (P = 0.197), with sweepers (3%, P = 0.475) and the reference group (2%) falling in between.

When considering overall occupational health risks, pickers again reported the highest prevalence (43.6%, P = 0.020), followed by collectors (32.7%, P = 0.258) and sweepers (28.9%, P = 0.706). The reference group reported a lower prevalence of 27.3%.

Some health outcomes were assessed only among participants who had already reported related problems in the past year. For instance, needle-stick injuries within the last six months were common among those with skin cuts, affecting 70% of pickers (P = 0.097), 56.5% of collectors (P = 0.222), and 26.9% of sweepers (P = 0.414), compared with 38.1% of the reference group. Similarly, among participants with musculoskeletal pain, body pain in the last six months was reported by the vast majority: 91.7% of pickers (P = 0.523), 82.9% of collectors (P = 0.907), 67.8% of sweepers (P = 0.128), and 84% of the reference group. These subgroup findings should therefore be interpreted with caution.

Discussion

This study aimed to examine differences in health and working conditions among various types of SW. The key findings indicate that pickers and collectors are the most vulnerable groups in terms of health risks, with pickers reporting the least access to preventive measures and hygiene resources compared to all other groups. Sweepers reported fewer symptoms of heat stress than the other SW types and had the lowest reported rate of occupational training.

Health risks

One notable finding in this study was that pickers reported a significantly higher prevalence of dental issues. This may be linked to their prolonged direct exposure to the waste and airborne pollutants, which some evidence indicates can impact dental health [16]. Furthermore, pickers reported the lowest income sufficiency, which may hinder their ability to afford dental care or maintain adequate oral hygiene, potentially exacerbating their dental issues.

A review study has shown that airborne pollutants generated from decaying waste are also associated with gastrointestinal infections, especially through hand-to-mouth contact [17]. The levels of air pollutants produced by the waste may be exacerbated in high temperatures and when the waste remains unsorted for prolonged periods of time [17]. Pickers may be more vulnerable to this due to their close contact with the waste and the low access to and use of soap and sanitizer they reported. Despite these risks, our study did not reveal significant differences in the risk of gastrointestinal infections between the groups. The review study exclusively included studies from countries that met EU standards of practices related to municipal waste management. Thus, the levels of airborne pollutants may differ substantially from those in Nepal, which may mean that SW in Nepal are either all affected or protected due to differences in waste management practices.

Studies that are similar to our study have identified respiratory issues as a highly prevalent symptom, with up to 70% of SW experiencing this [18–20]. Surprisingly, respiratory issues were uncommon in this survey as only 4.2–10.5% of SW reported experiencing any respiratory symptoms. This discrepancy may be due to the short median duration of employment of five years in this study, which may be insufficient exposure for chronic respiratory conditions to manifest. However, one study found a high prevalence of respiratory symptoms even among SW with employment durations of just seven years which suggests that additional factors may contribute [19]. The difference is unlikely to stem from tobacco use alone as the rates of tobacco use in this study were similar to those in a study that reported higher prevalence of respiratory issues [18]. Another study found that respiratory symptoms were associated with lack of training and PPE usage among others [20]. It is possible that contractual employment in our sample afforded better PPE access and occupational training, partially mitigating exposure though the quality of this remains unverified. The study measured respiratory symptoms using spirometry, which objectively confirms the prevalence of respiratory problems among their population. It also enables the discovery of subclinical symptoms which further strengthens the results whereas we did not objectively confirm the symptoms that were reported by the participants. Thus, our study may have underreported respiratory problems as participants with subclinical symptoms have most likely categorized themselves as healthy in that aspect.

Our analysis revealed a significantly higher unadjusted risk of musculoskeletal pain among sweepers and collectors compared to the reference group. However, only collectors retained a significantly higher risk after adjustment. A descriptive study found that up to 49% of SW experienced musculoskeletal pain in the upper body [21]. The population of the descriptive study corresponded to the collectors and pickers of our study, but 88% of the participants of the descriptive study were female, whereas only 26.3% of pickers and 1.8% of collectors in our study were female [21]. We did not measure the daily weight load of male and female SW, but evidence indicates that the different workloads relative to body weight are unlikely to explain this variation between these two studies as weight load relative to body weight has no significant effect [22]. There may be unknown intrinsic factors associated with being female, that exacerbates the risk of musculoskeletal pain, as a study in Greece found that female workers were at higher risk of experiencing musculoskeletal disorders [23]. Another study focusing on male collectors also found that up to 89% of participants experienced upper body pain, especially in the upper back and shoulders [24]. In contrast, our participants reported a much lower prevalence of musculoskeletal pain. Among those who did report musculoskeletal pain, many noted that it affected their ability to work. The difference in prevalence may be due to different assessment methods. In this study musculoskeletal pain was categorized as a binary yes/no response potentially underestimating prevalence by failing to capture mild symptoms. Other studies used specific and more detailed assessments, yielding more comprehensive data on the participants’ health status. Consequently, this study may underestimate the prevalence and risk of musculoskeletal pain, as participants may have reported only severe cases that impaired their ability to work. Some studies found that walking long distances or for long periods of time during the workday was associated with higher risk of musculoskeletal disorders [25, 26]. This may partly explain the higher risk for sweepers and collectors that our study found, as they move long distances during their workday, while the reference group are stationary at an office or move by vehicle, and the pickers usually sit for a long time while they sort the waste.

Our study found that collectors and pickers were at greater risk of sustaining skin cuts, which may be due to their more direct handling of waste materials compared to the other groups. For pickers, this elevated risk of skin cuts may be explained by their limited access to PPE. Additionally, pickers reported the lowest prevalence of relevant vaccinations, though this was not significant, which increases their vulnerability to infections following skin cuts and needle sticks. Furthermore, the pickers displayed lower access to health insurance, which may impair their ability to seek treatment for work-related injuries and infections. The lower access to health insurance was not significant after adjusting. In contrast, collectors reported a significantly higher vaccination rate against relevant diseases, which may provide greater protection in the event of skin cuts and needle sticks. Though it was not significant, sweepers reported the lowest risk of occupational health problems overall, and indicated the lowest risk of almost all health outcomes, despite having comparable access to PPE to the collectors. Only the lower risk of heat stress was significant. This difference may be an indication that sweepers’ tasks are less hazardous than those of the other groups. This may be due to limited direct contact with the waste, as their work primarily involves handling it with brooms. The lower risk of body pain and physical injuries may be due to task characteristics such as lighter workloads and minimal heavy lifting. Notably, sweepers were the only group that predominantly consisted of women, which may reflect gendered task allocation based on perceived physical demand or cultural permissions.

Pickers also reported greater exposure to heat stress, although this association did not remain statistically significant after adjustment. This trend may be influenced by their work schedules, which often extend into the hottest parts of the day, in contrast to other SW groups who typically work in the cooler morning hours. Additionally, pickers reported significantly less access to drinking water which may increase their susceptibility to heat-related symptoms.

Preventive materials

A qualitative study has found that access to PPE does not guarantee its use. Discomfort from the heat or weight of the equipment or interference with task performance seems to discourage consistent use of PPE [27]. Furthermore, the PPE that participants described in the qualitative study was mostly improvised from materials found among the waste [27]. Cross-sectional study also indicate that even when SW have access to and use PPE it might be inadequate materials or hygienic maintenance or it may not be used due to discomfort [19, 28]. While many participants in this study reported access PPE, we did not assess the quality or the manner of usage. Improvised solutions may have been mistaken for proper PPE in self-reports, leading to an overestimation of actual protection. Notably, only 40–69.7.7% of PPE access was provided by the employer, meaning that the rest was provided by NGO’s or self-sourced, suggesting varying quality and reliability. Although one study is qualitative and the others are cross-sectional, they all focus on informally employed SW in Nepal. Thus, it may provide an accurate depiction of PPE use among the participants in our study as most were employed on temporary contracts. The qualitative study included waste collectors, which may correspond to collectors and pickers in our study, while the cross-sectional studies seemingly included all types of SW, though random sampling was impossible and there may be inclusion bias. Therefore, whether this tendency to use non-standardized PPE is associated with specific types of SW is unknown. It is also unknown whether this occurs with formally employed SW.

Our findings indicate that pickers are significantly more likely to consume food during work hours compared to other groups. This may partly be explained by work schedule differences, as pickers typically work throughout the day while others primarily work in the morning and may return home to eat. However, this behavior increases the pickers’ risk of ingesting harmful pathogens, especially considering their limited access to soap and sanitizer. The lack of adequate hygiene facilities raises the likelihood of consuming contaminated food, particularly in a work environment heavily contaminated by waste materials.

Strengths and limitations

In this study, participants self-categorized into different types of SW, without detailed assessment of their actual work tasks. Neither enumerators nor researchers reclassified participants based on job descriptions, which introduces the potential for misclassification and overlap between SW categories. If significant misclassification occurred, the statistical disparities between the groups may have been mitigated as the groups may be more similar in the data. Despite this limitation, most outcomes followed the anticipated trend with the reference group and sweepers appearing less vulnerable to occupational health risks, while pickers and collectors seemed to be more vulnerable.

The absence of a general background population for comparison limits our ability to evaluate whether SW experience poorer health outcomes relative to the broader population of Nepal. Consequently, we could not assess how factors like smoking, literacy, or income sufficiency differ between SW and the general population. This restricts the contextualization and broader interpretation of our findings.

As participants were recruited through open invitation, there is a risk of non-response bias. SW who chose not to attend may systematically differ from those who did. They may be healthier or less interested in health services, or they may be more affected by symptoms and not have the capacity to attend. This introduces concerns regarding external validity, as the sample may not be representative of all SW in Nepal.

This study relies on self-reported data of experiences up to one year prior, which introduces the possibility of recall bias. Participants may have underreported or overreported symptoms depending on memory accuracy, misattribution of symptoms, or concerns about confidentiality and employer retaliation. In some cases, response bias may have been present if participants exaggerated health complaints due to awareness of the study objectives. Clinical validation of symptoms from the physical examination was not included in this manuscript as the examination could not confirm symptoms that were not currently present. It is also likely that the study underestimated the true prevalence of occupational health issues due to the healthy worker effect as severely ill or inactive SW were excluded, and their absence may have biased the results toward healthier outcomes. However, it is unclear whether the illnesses and absences that resulted in exclusion were due to occupational conditions and injuries.

The six-month data collection period introduces the possibility of seasonal bias. For instance, participants surveyed during warmer months may have reported more heat stress symptoms, while those interviewed during cooler months may have underreported such experiences.

Although enumerators were trained to conduct the survey, interpersonal variation in data collection introduced inconsistencies and missing data. Some responses were incomplete, ambiguous, or left blank, which we treated as missing values. Some responses appeared unrelated to the intended question, suggesting that certain participants may not have fully understood the question. Such answers have been categorized as missing data. However, even when answers appeared relevant, particularly in binary yes/no responses, we cannot be certain that participants fully understood the questions. Additionally, it is unknown how much explanation enumerators provided during interviews, or whether recorded responses were the participant’s own or interpretations by the enumerator. The questionnaire was originally designed in English, translated into Nepali for data collection, and then retranslated into English for analysis. This process may have introduced linguistic and cultural nuances that affected how questions were understood and answered. Seemingly inconsistent responses may reflect translation-related differences in interpretation rather than actual misunderstanding.

The logistic regression analysis was adjusted for sex, age, level of education, income sufficiency, duration of employment, alcohol and tobacco use. However, some residual confounding is likely, as these characteristics were all dichotomized. A more refined adjustment using continuous or ordinal variables would likely yield a more nuanced understanding of the relationships between health outcomes and these factors, especially for behaviors like alcohol and tobacco use, which typically demonstrate dose-response associations, and for education and income sufficiency, which often correlates with health literacy and healthcare access.

To control for statistical differences between the participants socioeconomic and demographic confounders were included in the analysis. Significant baseline differences between SW groups necessitated adjustment for lifestyle factors such as alcohol and tobacco use, which are known independent risk factors for various health issues, especially gastrointestinal and respiratory symptoms. Level of education and income sufficiency were incorporated due to their well-established protective associations with general health outcomes [10–12]. Age and duration of employment were included to account for the cumulative effect of time on health, as both are associated with increased risk of morbidity. Finally, sex was included to adjust for potential biological differences in health outcomes between males and females. This ensures that the findings of this study are due to the tasks performed by SW rather than due to the characteristics of the SW themselves.

Conclusion

This study found that among the four groups of SW, pickers and collectors were the most vulnerable to occupational health risks. Pickers reported a higher prevalence of skin cuts, dental problems, and limited access to PPE, hygiene materials, and water, indicating critical gaps in occupational safety. Collectors exhibited a significantly elevated risk of musculoskeletal pain and skin cuts despite similar access to preventive and sanitary measures as the reference group. This vulnerability is likely due to an inherently greater risk associated with working in close contact with the waste and the physical demands associated with this.

These findings indicate the need for equal access to preventative measures, such as PPE, vaccinations, and sanitary materials across all SW groups, along with tools to alleviate musculoskeletal pain.

Municipalities should establish routine health screenings, ensure the mandatory provision of appropriate PPE, and expand access to health insurance for all SW. Future research should compare SW with the general population to better contextualize health risks and explore task-specific exposures to inform development of effective occupational tools and strategies. These data can help advocate for improved governmental policies and municipal waste management practices ensuring the occupational health and safety of SW in Nepal.

Abbreviations

SW

Sanitation worker

PPE

Personal protective equipment

OR

Odds ratio

NHRC

Nepal Health Research Council

Authors’ contributions

The study protocol was written by KR, EJ, DN, GKS and the NEDS team. Data collection was performed by the NEDS team and local health workers. SHK prepared the manuscript. SK, DN and EJ provided comments.

Funding

Open access funding provided by University of Southern Denmark. The development and execution of the questionnaire and health camps, along with the companion study, was funded by Danish civil society namely Civil Society in Development (CISU) with technical support from Danish Society of Occupational and Environmental Medicine (DASAM). The funding organization has no role in the design of the study, collection & analysis of data and drafting of the manuscript, this was the role of authors. The writing of this manuscript has not been funded.

Data availability

The dataset used for the study is not available publicly but can be made available upon reasonable request with permission from the Nepal Development Society and Civil Society in Development (CISU). To request dataset, please e-mail Sara Hagedorn Kragh at [hagedornjensensara@gmail.com](mailto: hagedornjensara@gmail.com). Data is provided within the manuscript. The questionnaire is available as a supplementary file.

Declarations

Ethics approval and consent to participate

Ethical clearance for the study was approved by the ethical review board (ERB) of the Nepal Health Research Council (NHRC) [Ref No. 1260]. Since the study involved the participation of humans (study participants) in the health camps, we fully adhered to the National Ethical Guidelines for Health Research in Nepal 2022. The participant information sheet and consent form were provided to read or was verbally read by the research assistant. The objective of the study and the potential risks to the participants was clearly explained prior to obtaining the written/verbal informed consent from all study participants and the legally authorized representatives of minors and illiterate participants. They were also informed of confidentiality and anonymity, and of their right to withdraw consent at any time.Approval letters were obtained from each of the municipalities before the conduction of the health camp. Confidentiality and privacy were maintained thoroughly via de-identification of data. This study adhered to the principles of the Declaration of Helsinki throughout the entire development, execution, and processing of the health camps and data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.