Sociodemographic, environmental and labor conditions related to the presence of conjunctivitis and skin irritation in a group of informal street vendors in downtown Medellin, 2015-2019

Introduction: Occupational diseases are those that may have a causal relationship with occupational activity or environment. However, this definition does not specify how this disease would be identified and acknowledged for workers with subsistence jobs. Objectives: To determine sociodemographic, labor and environmental conditions that collaborate to explain the presence of eye and skin irritation among informal vendors in downtown Medellin. Methods: Descriptive cross-sectional study with analytical intention, primary sources of information, and administration of assisted survey including self-reported eye and skin irritation in 686 workers. Results: Predominantly male population with mean age of 50 (±11.8) years. A total of 23.5 % of workers had worked as vendors for their entire life; 37.5% had worked as vendors for 11 to 20 years as vendors; and 81.5%worked for more than 8 hours a day. According to 69.8% of the sample, poor air quality affected their work, and 80.0% believed that pollution was generated by vehicle fleet. The polluted areas generated unpleasant odors (50.2%) and air pollution (89.4%). The prevalence of eye and skin irritation was 65.4%. Not having working license and having fair, poor, or very poor job tenure significantly reduced the prevalence of eye and skin irritation. Conversely, exposure to polluted water, working in the professional for more than 30 years, male sex, and age older than 60 years were associated with increased prevalence of irritation. Conclusions: Eye and skin irritation was mainly explained by non-modifiable sociodemographic and labor conditions and by consumption of polluted water; however, implementing public health actions could reduce workers’ socio-environmental and labor vulnerability.

intrOductiOn Occupational diseases are defined as those resulting from the exposure to risk factors inherent to the occupational activity or environment, according to the article 4 of Law 1562, which came into effect in 2012 in Colombia, 1 as well as those that may present a causal relationship with these factors. However, this definition does not specify how these diseases would be identified and acknowledged in the informal sector of economy.
In Latin America, occupational diseases and accidents lead to high morbidity and mortality rates in the formal sector. 2 In Colombia, according to the 2018 report of the Federación de Aseguradores Colombianos, 3 "the rate of occupational diseases decreased 24.0% in the last eight years, reducing from 130.6/100,000 to 99.6/100,000". 3 Nevertheless, this information is not available yet for individuals working in the informal economy, among which morbidity and mortality rates could often result from their precarious labor conditions and exert a great impact on their physical and mental health status. 4 In July 2019, 47.5% of the Colombian worker population had informal jobs, 5 meaning that this form of employment is the means of subsistence for a high percentage of people. However, there is still little information on the morbidity conditions related to air pollution among these workers, particularly in the city of Medellín, 6 whose downtown district 7 is one of the areas with the greatest amount of informal workers and the highest environmental pollution of the metropolitan area of the Aburra Valley, as shown by the reports of Contraloría General de Medellín and of the Instituto de Hidrología y Meteorología (IDEAM). According to these reports, in 2016 the concentrations of particulate matter exceeded the maximum allowable threshold in 6 of the 9 monitoring stations that complied with the temporal representativeness criterion. The maximum category of the Air Quality Index (AQI) achieved during 2016, corresponding to hazardous levels for the health of sensitive groups, was reported in the monitoring stations Éxito San Antonio and Museo de Antioquía. However, the highest number of stations showed a moderate AQI level, ranging from 56.8% to 74.1%. 8,9 These locations concentrate a high amount of informal street vendors working in the downtown district, where they are exposed to situations and pollutants (high levels of noise, heat, particulate matter, and gases) 10 that may be affecting their physical health status, particularly because studies with this type of population have already shown that their working days are long and they have low monthly income. 6,11 One of the physical diseases that could arise among informal workers in downtown Medellin is eye and skin irritation (ESI), which could be related to environmental pollution. However, current evidence on the theme is not conclusive, because it is little explored within this population. Similarly, there is no conclusive information on how this type of health illnesses could be explained by the sociodemographic, labor, and environmental conditions in which workers develop their work, 12 which explains the relevance of providing evidence on this matter. Therefore, the present study aimed to determine which sociodemographic, working, and environmental conditions explain the presence of ESI in a group of informal street vendors working in downtown Medellin, in order to acknowledge modifiable and non-modifiable risk factors, which could be approached by actions and public health interventions that contribute to improve workers' physical health conditions, 13 and contribute to minimize their conditions of socio-environmental and labor vulnerability.

DeSiGn
The present study is a subproject of the doctoral thesis entitled "Condiciones ambientales, laborales, sociodemográficas, económicas y de percepción de salud que configuran la condición de vulnerabilidad laboral de un grupo de trabajadores informales 'venteros' del centro de Medellín, 2015-2019" and was approved by the institutional Ethics Committee of Universidad CES, Minutes 84, dated September 24, 2015.

POPulATiOn
Survey with 686 informal street vendors working in downtown Medellin.

VARiABleS
Outcomes: prevalence of ESI. The following • sociodemographic variables were considered explanatory variables: sex, age, and marital status. Labor conditions: • type of vendor, type of product, daily working hours, years working in the profession, working license, previous occupation, job tenure, conditions of working tools, order and cleanliness of point of sale and its surroundings, feeling of exaggerated heat or cold, use of chemicals, exposure to chemicals, possible contamination with microbes in the working area, personal protection methods, damaged walls and floor. Environmental conditions: time of day when the • respondent considers air pollution is greater, time of exposure to the pollutant, air quality affects work, point of sale near polluted area or site; waste accumulation in open spaces, waste accumulation in stationary waste buckets; waste accumulation in containers, sewage source, market place, vehicle fleet or restaurant, the polluted area or site generates: unpleasant odors, visual pollution, air pollution, water pollution, among others.

Pilot test, control of errors and BiASeS
A pilot test was conducted prior to administering the instrument, and the principal investigator and her field assistant were standardized for data collection. Selection bias was controlled by using a census sampling technique.

AnAlySiS
The nature and the measurement level of the variables were considered. A normality test was applied to assess data distribution for quantitative variables, and frequency and percentage were calculated for qualitative variables in univariate analyses. Non-causal associations between explanatory variables and prevalence of ESI were explored using the chi-square test, and prevalence ratio (PR) and its respective 95% confidence interval (95%CI) were used as epidemiological measurements. The types of products were reclassified into: seasonal and perishable products; beverages, appetizers, and desserts; fast food; goods and pots, with the latter being the reference category. Moreover, sociodemographic, labor, and environmental conditions were reclassified for bivariate and multivariate analyses.
Multivariate analysis with multiple logistic regression, • in order to identify the factors that contribute to explain ESI: variables were included in the model from the lowest to the highest p-value, according to the results of bivariate analysis and complying with the Hosmen-Lemenchow criterion (p < 0.25). Tests were applied considering a 95%CI and a significance level of 5.0%. Calculations were made using Epidat 3.1, 4.1 and SPSS® version 21 (licensed to Universidad CES). results sociodemograPhic and labor COnDiTiOnS Workers' mean age was 50 years (± 11.76), with a greater proportion of men (57.6%) and individuals without a partner (52.4%) ( Table 1).
With regard to labor conditions, 23.5 % (161) of participants had only worked as a street vendor during all their lives; most (37.5 %) had been working for 11 to 20 years; and most worked more than 8 hours a day (81.5%). Furthermore, 61.2% (420) had a working license and more than 75.0% considered they had job tenure. Study participants were mostly semi-stationary and itinerant vendors, and the highest proportion of workers (58.8%) sold goods and pots (Table 1).
When it comes to occupational risk factors, 40.4% (277) of participants considered that their working tools were not in good conditions; 35.6% (244) believed that there were damaged walls or floors in their workplace or in its surroundings; 6.1% (42) stated that their working area was not clean and tidy. Furthermore, 72.7% (499) of workers reported not using personal protection measures; 57.9% (397) believe that they could be contaminated with microbes in the workplace or in its surroundings; and 86.6% (587) stated that they were exposed to exaggerated heat or cold in the workplace (Table 1). * Type of vendor: stationary when the vendor remains at the same workplace 24 hours a day; semi-stationary when the workplace can be located in the public space only during vendor's working hours; and itinerant when the goods or the products sold by the vendor are carried on his/her body. † Type of product: goods and pots; seasonal and perishable products; beverages, appetizers, and desserts; and fast food. ‡ Other products: seasonal and perishable products; fast food; beverages, appetizers, and desserts.
environmental conditions in the Point of sale and its surroundings A total of 69.8% (479) believed that poor air quality affected their work, 56.4% (373) reported that morning and evening were the periods of the day with greater air pollution; 36.0% (220) observed waste accumulation in open spaces in the workplace or its surrounding; 34.9% (213) considered to be near wastewater (sewage); 80.0% (489) stated that the polluted area or site (contaminant source) was represented by el vehicle fleet. These pollutant sources mainly generated unpleasant odors (50.2%) and air pollution (89.4%) ( Table 2).

Prevalence of esi
A total of 54.4% of the workers reported that, in the six months prior to the survey, they had presented with some disease related to air contamination; of these, 65.4% (244) mentioned that they have developed ESI (Table 2).

sociodemograPhic and labor conditions associated with esi
There was a statistically significant association (p < 0.05) of ESI with working license and job tenure. That is, workers with working license had a higher prevalence of ESI than those with no license (PR = 0.81; 95%CI = 0.68-0.95); whereas, for each vendor with good and very good job tenure who had ESI, 0.81 workers with fair, poor, and very poor job tenure who had the same condition (PR = 0.81; 95%CI = 0.68-0.97) ( Table 3). Although no statistically significant association was observed between age and the outcome of interest, there was a higher prevalence of ESI (18%) among workers aged 60 years or older than among those aged from 18 to 29 years (PR = 1.18). There was also a higher prevalence of ESI among those who had worked in the profession from 21 to 30 years (PR = 1.22).
Furthermore, the prevalence of ESI was higher in men (PR = 1.13) ( Table 3).

environmental conditions associated with esi
Statistical significant associations (p < 0.05) were observed between waste accumulation in containers provided by Empresas Varias de Medellín (EEVV) in the public space and air pollution, meaning that, for each worker with ESI who were not near waste   (Table 4).
sociodemograPhic, labor and environmental conditions that exPlain esi among workers One of the statistically significant variables (p < 0.05) to explain a higher prevalence of ESI is not having working license, showing that the prevalence of ESI was 43.0% lower [adjusted prevalence ratio (PR adj ) = 0.57] among those who did not have such license, with a decrease of 24% [unadjusted prevalence ratio (PR unadj ) = 0.81; 95%CI = 0.68-0.95; PR adj = 0.57; 95%CI = 0.33-0.97) after adjustment for the remaining sociodemographic, labor and environmental conditions. Other variable that had a significant contribution to explain a lower prevalence of ESI was the fact of considering job tenure as fair, poor, and very poor, observing that there was a decrease in the prevalence of ESI after adjustment of this characteristic for the remaining variables, varying from 19.0% in unadjusted analysis to 43.0% after adjustment for the remaining study variables (PR unadj = 0.81; 95%CI = 0.68-0.97; PR adj = 0.57; 95%CI = 0.34-0.94) ( Table 5).
Air pollution and waste accumulation in containers provided by EEVV in the public space, which were  found to be significantly associated with a lower prevalence of ESI in bivariate analyses, lost their statistical significance and exhibited changes in their ability to explain a lower prevalence of ESI. Conversely, air pollution strengthened their association with a lower prevalence of ESI (PR unadj = 0.74; PR adj = 0.34), a situation contrary to that observed with waste accumulation in containers provided by EEVV in the public space, which had no longer a significant association with a lower prevalence of ESI (PR unadj = 0.48; PR adj = 0.85) in adjusted analysis after adjustment for the remaining sociodemographic, labor and environmental conditions, as observed in Table 5. Conversely, water pollution increased its explanatory ability in the multivariate model (PR adj = 2.75) compared with bivariate analyzes (PR unadj = 1.27), showing that the prevalence of ESI was 1.75 times higher among those exposed to this type of pollution (Table 5).
There was also a change in the results of the multivariate model with regard to length of time working in the profession, considering that unadjusted analyzed yielded a higher prevalence of ESI for workers with 21 to 30 years of profession (PR unadj = 1.22), whereas adjusted analysis yield a higher prevalence of ESI for those with more 30 years of profession (PR adj = 1.77), followed by those with 11 to 20 years (PR adj = 1.41), and those with 21 to 30 years (PR adj = 0.89). Thus, vendors working for more 30 years in the profession were shown to have a 77.0% higher prevalence of ESI compared with those working for shorter (5 to 10 years) in the profession after adjustment for the remaining variables included in the multivariate analysis (Table 5).
Finally, it was found that the prevalence of ESI for men increased by 27.0% after adjustment for the remaining variables included in the analysis (PR unadj =1.13; PR adj = 1.40), with a prevalence of ESI 40% higher than that of women (Table 5).

discussiOn
The prevalence of ESI in the study population was 65.4% (244), a result similar to that reported in a study in which occupational dermatitis was one of the most frequent work-related diseases, with a 60% likelihood of presenting with dermatitis in formal workers of developed countries 14 . It is also important to bear in mind that the most prevalent skin disease in the work context is contact dermatitis ("irritation"), with a prevalence ranging from 70 to 90%. 15,16 However, evidence on these diseases in the workplace is still scarce for the formal sector of economy, 17 and is even scarcer for informal workers, which turn street and sidewalks of the cities into their workplace. It was observed that most participants in Table 5. Sociodemographic, labor, and environmental conditions, and pollutant sources that contribute to explain eye and skin irritation in informal street vendors working in downtown Medellin and participating in the study, Medellin, Colombia, 2015-2019  20 unlike what was observed in the present working population, of which only 4.7% reported to be itinerant vendors. With regard to the type of products, nearly 60% sold goods and pots, a result similar to that found in other studies 18 .
With regard to environmental conditions and polluted areas or sites (pollutant sources) identified by the workers, it was observed that 50.2% of them perceived unpleasant odors, and 89.4% perceived that the air was polluted in the workplace and its surroundings. These findings are somewhat related to those reported by a study conducted in a market place in Bogotá, which investigated the conditions that informal workers perceived as dangerous for the development of their activities and observed that 47.3% reported unpleasant odors, 39.5% lack of order or cleanliness, and 30.4% the presence or rats 21 . The present study also observed that 86.6% of workers felt excessive hot or cold in their workplace. There are variable results in the literature, which reported a prevalence of 96.0% for workers involved in the production of fique in the municipality of Chachagui, located in department of Nariño. 22 There was a significant association between ESI, job stability, and working license, with workers who considered to have fair, poor, or very poor stability showing a lower prevalence (19.0%) of ESI than those who considered to have good or very good stability. This finding is different from that reported in a study on the Indian informal sector, 23 which found that job tenure is related with risk of developing diseases in informal workers, although not specifying the type of occupational diseases these workers developed. As for working license, it was found that those who did not have such license showed the lowest prevalence of ESI (19%). However, when comparing this finding with evidence reported in the literature, there was no studies reporting this type of data, a situation that leave open the possibility of further investigating these characteristics in future studies.
Conversely, waste accumulation in containers provided by EEVV was associated with a higher prevalence of ESI, a situation that, similar to the one mentioned above, has not been reported yet in the specialized literature. There was also an association between ESI and air pollution, in which, paradoxically, being in environments where vendors considered that air was polluted reduced the prevalence of eye and skin diseases by 26%, contrary to findings reported in the literature, which stated that passive smoking and polluted air was risk factors for developing eye diseases. 24 Furthermore, a higher prevalence of ESI was observed among workers who reported to feel exaggerated heat or cold in the workplace, among those who reported to be exposed to chemicals, among those who perceived that their workplace was affected by other pollutants, such as smoke; and among vendors whose workplace was located near restaurants. These findings are similar to those reported in the literature, which showed that high temperatures, exposure to irritating chemical compounds 24 and windy and dusty areas 25 have been related to eye diseases.
With regard to personal protection mechanisms that are supposed to be essential factors in occupational health, they are not broadly used by the population of informal workers, which exacerbates their exposure to risk factors and predisposes to work accidents. 23 The present study indicated that the use of personal protection measures reduces the risk of developing cutaneous and ocular manifestations by 2.0%. Similar results were reported by Ballesteros et al. in the city Medellin in 2005 with a population of recyclers in downtown Medellin, of which less than 52.0% used the required protection measures, thus increasing the likelihood of developing respiratory and gastrointestinal diseases, among others. 19 A meta-analysis on informal work found that waste collectors do not often use any type of personal protection equipment, and they frequently use their hands to collect the recyclable material; one of the studies revealed that half of the collectors (56%) reported never having worn gloves during work. 4 Conversely, with regard to labor conditions, a clean, tidy workplace favors a lower prevalence of eye diseases. 25 . Similar data were observed in the present study, in which an 11.0% lower prevalence of conjunctivitis was observed among worker who reported to have a clean, tidy workplace.
The facts of considering job tenure as fair, poor, or very poor and not having a working license explained a lower prevalence of eye and skin irritation. However, no reports have been identified in the literature that allow to assert which factors explain a lower or a higher prevalence of this disease for informal workers. The following variables were also able to explain a higher prevalence of ESI: contact with polluted water, length of time in the professional greater than 30 years, male sex, street vending as only profession, working from 4 to 8 hours a day, and observing waste in containers of the Medellin public service company named EEVV. These are some of the characteristics reported for the general population, as shown by a study of pediatric populations in India and in low and medium-income countries, which observed that the impact of water, sanitation and hygiene interventions reduced the risk of conjunctivitis by 51% (risk reduction = 0.49; 95%CI = 0.45-0.55). 26 Consistent with this, the present study showing that being near polluted water increases the prevalence of ESI by 175% (PR adj = 2.75). This study also found that working for more than 30 years in the profession is able to explain a greater IOP, a result comparable with that reported in an article on environmental health and preventive medicine which observed that working for more than10 years in the profession was a risk factor for occupational dermatitis ocupacional. 15 Additionally, another paper found that there is a high prevalence of tear secretion dysfunction (risk factor for conjunctivitis) among female workers who perform lights tests; moreover, this prevalence increases as length of time in the professional increases. 24 Male sex was able to explain higher prevalence of ESI, in agreement with results observed in the literature showing that male sex was associated with a higher prevalence of occupational dermatitis, since men accounted for 64.0% of individuals with this disease. 16 The same prevalence was also found in a study assessing the clinical and sociodemographic profile of patients with occupational dermatitis in a dermatology service. 27 However, other studies with formal workers reported different findings, with a 70.0% higher prevalence of dry eye among women. 24 With regard to age, a study reported that there is a higher prevalence of dry eye (up to 4.34%) among men older than 50 years, 24 differently from the results of some studies which found a higher prevalence of occupational dermatitis among individual younger than 50 years of age. 28,29 liMiTATiOnS There were difficulties in comparing some environmental and labor characteristics and conditions associated with ESI, given the scarcity of scientific evidence related to informal workers or to other types of jobs, and although the type of design only allows us to speak of factors associated and that explain prevalences these results. Moreover, study design limited statistical analyses. Both situations strengthen lines of research with this working population. cOnclusiOn It was found that the population of informal street vendors in downtown Medellin are directly exposed to certain environmental factors and labor conditions that are able to shape its condition of socio-labor and environmental vulnerability. This study highlights the importance, in terms of public health, of conducting relevant interventions in occupational health that have a positive impact on workers' life and health.

acKnOwledgment
To CES university for its support and cooperation, both in the design and in the execution of the doctoral thesis from which this study is derived.