INTRODUCTION

Development of industrial complexes has caused substantial damage to the environment and to human health (^{Janmaimool and Watanabe 2014}). Several studies have demonstrated that dwellers of residential areas bordered by industrial installations report foul odors, are at greater risk for symptoms such as wheezing and upper respiratory disorders (^{Tanyanont and Vichit-Vadakan 2012}) and exhibit higher levels of heavy metals such as cadmium, than their control group counterparts (^{Interdonato et al. 2014}).

Other studies have indicated that exposed communities are at greater risk for death from cancer (^{Bulka et al. 2013}, ^{García-Pérez et al. 2015}, ^{Fernández-Navarro et al. 2017}), and display increased anxiety and psychological stress with a wide range of effects on their health and wellbeing (^{Cutchin et al. 2008}).

The Quetzalcóatl industrial corridor in the state of Puebla is a vivid example of these conditions. This industrial corridor has 16 different industries, including textiles, metallurgy, automotive, medical pharmaceutical products and food dehydration. One of these companies incinerates biological-infectious wastes (BIW). These industries also coexist with agricultural activities that use insecticides and chemical fertilizers.

All of these economic activities have caused serious environmental pollution in the region, which has created a complex situation related to chemical exposure in the communities (^{Montero 2004}). In addition, it is important to note that this environmental deterioration is part of a much broader and complex problem in the Atoyac River basin, one of the largest basins in the state of Puebla. It has resulted from the establishment in the region of the maquila textile factories as well as three large industrial complexes-the Independence Petrochemical Complex belonging to Petróleos Mexicanos (built in 1968) and the Quetzalcóatl and Huejotzingo corridors-(^{Morales et al. 2005}).

Recent studies indicate that water from the river poses a public health risk involving toxicological and mutagenic parameters (^{Montero et al. 2006}, ^{Sandoval et al. 2009}). The population potentially exposed to pollutants in the Atoyac River sub-basin is estimated to be 2 005 444 inhabitants-1 875 808 in Puebla and 124 508 in Tlaxcala (neighboring state) (^{Montero et al. 2006}).

One approach to understanding environmental risk is to analyze the perceptions of community members. Factoring people’s perceptions into risk analysis can contribute to identifying effective mitigation measures (^{Ramos-Ribeiro 2014}). It has been documented that lack of knowledge on how communities perceive risk hinders the implementation of potentially successful educational, communication and public policy strategies. Disregarding local perceptions often results in interventional contents that do not address the wide variety of social views and necessities attached to environmental problems (^{Ávila 2015}).

To characterize the environmental risks attributed to the Quetzalcóatl industrial corridor and the social response to perceived pollution, as well as to compare records of increased mortality of inhabitants with the social perception of deaths in the localities, a sequential mixed method study was conducted in two communities which used to be entirely rural but are now located in an industrial zone.

The two communities are located at different distances from the Quetzalcóatl industrial corridor in the municipality of Huejotzingo, state of Puebla, Mexico. One is the Santa Ana Xalmimilulco community, which is 1 km from the corridor and the other is Huejotzingo, which is further away (7 km) (Fig. 1).

Fig. 1 Study localities 

The study was designed based on the concern expressed by the inhabitants about the health effects from industrial pollution and their perception of an increase in cancer cases in the communities.

The residents of Santa Ana Xalmimilulco have been protesting since 2006 because of the foul odors and health effects which they associate with the pollution from the BIW incinerator, particularly an increase in cancer cases among women in the community. In May 2007 the residents of Santa Ana Xalmimilulco continued their protests to demand the definitive shutdown of the company by closing schools and attempting to block the Mexico-Puebla highway. The protest was repressed by public authorities (^{Mota 2007}).

As a result of these social protests, the Procuraduría Federal de Protección al Ambiente (the Mexican environmental protection agency) temporarily closed the incinerator until the company installed a continuous monitoring system and a gas cleaning system to decrease the foul odors (personal communication, former public official, Santa Ana Xalmimilulco).

To analyze the social concern about the health effects of industrial pollution in the study area, our investigation raised three questions: What are the perceived risks of the women residing in the polluted industrial zone? What mortality trends can be observed among newborns and people with cancer in the area? Is there any association between social perceptions and the mortality data registered?

CONCEPTUAL FRAMEWORK

Risk perception refers not only to the processing of physical signals and/or information regarding potentially harmful events and activities; it also involves the formation of judgments regarding the severity, probability of harm and acceptance of those events and activities (^{Renn 2008}). Thus far, two broad approaches-one psychological, the other sociocultural-have been adopted to study how risks are perceived (^{Pidgeon 1998}).

The psychological method explores the manner in which individuals process and select information about risks; the mental and heuristic strategies they use to make sense of the world (common sense); the semantic images they elaborate regarding risks, and psychometric assessment. The latter, known as the psychometric paradigm, is the one most commonly applied in this field of study (^{Renn 2008}).

Under the psychometric paradigm, risk perception is an individual and inherently subjective psychological process. Also considered quantifiable and predictable, however, psychometric techniques are regarded as the best means of identifying similarities and differences in people’s perceptions and attitudes towards risk.

Psychometric studies have shown that the judgments of common people, that is, of those who are not risk experts, include hazard-related properties such as voluntariness, dread, catastrophic consequences, knowledge, familiarity, controllability, prior experience, perceived benefits, equity, and implications for future generations (^{Slovic 2007}).

By contrast, the sociocultural paradigm holds that risk perception and acceptance are rooted in social, institutional and cultural processes. They respond to determinants such as culture, values, influence from the media, socioeconomic status and trust in institutions that regulate and manage risks (^{Renn 2008}). The cultural theory of risk is the most representative model of the sociocultural approach.

According to this theory, risk perception is a collective and cultural construction, not an individual phenomenon (^{Douglas 1996}). Among other exponents of this proposal, ^{Douglas and Wildavsky (1982)} argue that people act as parts of social groups; they minimize certain hazards and emphasize others to maintain and control their groups. Consequently, they perceive that which is understood to be polluted and therefore dangerous as a threat to their social order, and to the integrity of both their physical body and the symbolic body of the community or society to which they belong.

Risk perception is thus a multidimensional process that takes place both individually and as part of society. It is influenced by psycho-cognitive and sociocultural factors that vary according to the type of hazard at hand. Specifically with regard to the environment, understanding the differences in the judgments and opinions of risk held by people in industrial communities offers valuable information for designing effective communication and risk management strategies (^{Janmaimool and Watanabe 2014}).

METHODS

Design

We selected these localities for two reasons: Xalmimilulco was located near the Quetzalcóatl industrial corridor and had engaged in social protest against the effects of the neighboring industrial installations; Huejotzingo allowed us to compare the risk perceptions of two communities located at different distances from an industrial corridor. Our study used a sequential mixed design (^{Creswell and Plano 2011}) with a small quantitative component embedded in a predominantly qualitative schema (QUAL-quan) (^{Bronstein and Kovacs 2013}) (Fig. 2).

Fig 2 QUAL-quan mixed method design. Adapted from ^{Morse and Niehaus (2009)}  

RESULTS

DISCUSSION

What were the risks perceived by women in the polluted industrial zone? According to our qualitative findings, women in Xalmimilulco perceived that industrial activity had caused substantial environmental pollution and serious damage to their health. The environmental problems they identified were ranked by visual and olfactory perception. With respect to health risks, participants were most concerned about the growing incidence of respiratory disorders, cancer cases, and problems during pregnancy.

The women from Xalmimilulco were aware that industrial development provided more work for locals and migrants, but also involuntarily exposed them and their families to environmental pollutants and health risks. They were reminded daily of these risks by the foul odors from the incinerator. These results match those of a study conducted by ^{Luria et al. (2009)} in northeast England. According to these authors, unfamiliar odors can trigger intense risk perception because of their association with asthma and cancer. The mere presence of a foul odor can give rise to adverse sensory responses and complaints. Additionally, individuals who are hypersensitive to odor-causing chemicals can experience physical and mental distress culminating in risk perception.

Other studies have demonstrated that, even when unrelated to any known danger to human health, environmental odors can affect the psychophysical wellbeing of those exposed (^{Capelli et al. 2007}). The connection of foul odors to health risk perception has also been evidenced by ^{Axelsson et al. (2013)} in Sweden, where the opening of a petrochemical facility caused irritation and concern about health risks among the neighboring residents.

Sensory experience renders the proximity of an industrial zone a key variable in how people perceive the air quality and health status of their communities. They regard pollution as a serious health hazard for themselves, their families and their communities (^{Guo et al. 2016}). As reported by ^{Omanga et al. (2014)}, “health status and distance from a polluting facility have been shown to be a strong predictor of risk perception… Obviously, people who live closest bear the burden of pollution (odor, smoke, dust, etc.) and naturally believe it is affecting their health”.

Other studies on risk perception related specifically to waste incinerators indicated that people residing near an incinerator plant exhibited higher levels of stress, anxiety and depression than those residing far from the plant. According to the study, these psychological conditions emerged from the interaction between ailments attributed to incinerator wastes and the perception of risk in the community (^{Lima 2004}).

The quantitative data showed an increase in the deaths of women with cancer in the study area. Integrated analysis of the qualitative and quantitative information demonstrated that the social perception of risk from industrial pollution corresponded to the mortality rates registered in the study area during specific periods.

The perception of risk from industrial pollution is considered to be substantiated by two events that were identified by the quantitative study and by what the women undoubtedly observed in their community: increased deaths of women with cancer during productive ages (15-64) between the years 2005 and 2007, and increased cases of neonatal mortality rates, which surpassed the national rates, in 1998 and 2000.

Our observations regarding differences in the risk perceptions of the Santa Ana Xalmimilulco and Huejotzingo communities, as well as an association between the social perception of risk and local mortality records, are consistent with the findings of ^{Interdonato et al. (2014)} in Sicily and with those of ^{Janmaimool and Watanabe (2014)} in Thailand. The former relate that, unlike individuals residing in environmentally unexposed areas, those subjected to industrial pollution present a high risk perception. Janmaimool and Watanabe (2014) indicate that residents of high-risk areas (determined by levels of dibenzene, 1,3-butadiene, SO2 and NO2) display greater risk perception regarding their psychological and physical health than their low-risk counterparts. These people specify respiratory disorders, discomfort due to pollution, and adverse effects on their everyday lives. The authors conclude that the risk perception of common people, that is, of those who are not risk experts, can reflect the real nature of risks.

In line with the findings of ^{Wachinger et al. (2013)}, the results of our study support the notion that experiencing pollution on a daily basis is a decisive factor in the formation of risk perception.

Studies on environmental risk and hazard have generally argued that particular hazards are understood within the context of a community as a whole (^{Day 2007}). For example, in a study conducted in an Australian community, ^{Whittaker (1998)} found that social concerns regarding the risk of cancer due to the contamination derived from a residual disposal site and from the radioactivity from mining activities incorporated a range of social meanings. The perception of threat among residents pertained not only to the sites themselves, but also to certain characteristics of the community’s social space and underlying values related to the country’s political culture and economic policy. Additionally, risk perception among community members included a sense of loss and disempowerment regarding the ability to control their own future as a community.

In our case, the women from Xalmimilulco perceived a larger amount of environmental problems than those in Huejotzingo, both in their micro-environment due to pollution from the maquila factories and the lack of public services, as well as in their macro-environment from the large-scale industrial activities surrounding them, including the BIW incinerator.

Since most participants were 20-30-years-old, this group of women represented a generation of peasant girls who became the daughters and wives of the laborers or became workers themselves. They experienced the transformation of agricultural fields into industrialized zones as a result of the industrial development model followed in the region in the 1990s.

The macro-environment changed from a landscape dominated by nature to one controlled by the industries that slowly established themselves in the industrial corridor. Meanwhile, the textile industry led to the change in the micro-environment from an entirely rural community to one of a maquila factory community.

Furthermore, industries that began to take over very important local resources-for example water-began to create environmental pollution problems such that this entire industrial transformation resulted in great exposure to various pollutants and large health risks, as documented by previous studies (^{Montero 2004}, ^{Morales et al. 2005}).

These changes undoubtedly also impact the perception and representation of risk in the community. As a result, women cannot discuss pollution, environmental risk or public health independently from the social, political, economic and cultural context in which these events evolve and are experienced.

At the time of the study, the social response in the face of environmental contamination was directed to a greater extent toward carrying out actions to mitigate the impacts of exposure. Although this aids in decreasing environmental exposure, it also reduces the children’s physical activity, which raises some concern given the prevalence of overweight and obesity at the national level in Mexico, which is 34.4 % for children aged 5-11 years old and 35 % for adolescents between the ages of 12 and 19 years (^{INSP 2012}).

While women allude to a lack of leadership in responding to the environmental problems in the region, they also live with the memory of confrontation and the state government’s repression of the community in 2007, when residents attempted to close the Mexico-Puebla highway to protest pollution and the perceived health damages (^{SERAPAZ 2008}).

CONCLUSIONS

The perception of risk and the social response of communities to perceived risk emanate, on one hand, from the interaction of the psychological and cognitive factors behind the daily-life experiences of people in specific environments, and on the other, from the distinctive characteristics of each community as regards its economic, social and cultural dimensions. This approach enriched our quantitative data with the voices, experiences and perceptions of the people. It allowed us to demonstrate that the social perception of risk from industrial pollution matched the mortality records in the study area during specific periods. We therefore wish to highlight the contributions of mixed designs to integral risk assessment.

The findings of this case study can contribute not only to the formulation of targeted environmental and social policies, but also, and most importantly, to an effective involvement of women in strategies aimed at communicating and managing risks.

The descriptions that women in our study provided of their deteriorated environment and associated health risks as well as the results of our mortality analysis point out the need for conducting studies that broadly characterize the sources of pollution, polluting emissions, health risks and local vulnerability. It is important to focus future lines of research on the relationship between industrial complexes and local epidemiologic profiles.

A more in-depth analysis is required on the environmental and social determinants of the collective health-disease process in the region, specifically with regard to the manner in which cancer is historically and socially produced in these communities. There is evidence that exposure to dioxins and furans causes a high prevalence of genetic damage and a predisposition to cancer (^{Baltazar et al. 2013}).

Because there is no single cancer register in Mexico, we were unable to analyze the incidence of different forms of cancer at the national, regional and local levels. Our study therefore considered only deaths registered under local death certificates, and identified diagnosed cases of cancer at the time of death. We relied on death certificates as the most reliable official source of information on cancer, but they provided information on expired community members only; we did not have the opportunity to identify the incidence of cancer among those still living.

It is imperative to review the public policies for the disposal of hazardous biological-infectious wastes in Mexico that permit BIW incinerators to operate near communities. Accordingly, it is important to strengthen institutional capacities for establishing an environmental monitoring system linked to an epidemiological surveillance system and to an integral communication program on environmental pollution risks in the area. Setting up participative communication mechanisms would not only contribute to restoring social trust, but would also favor effective risk management. Additionally, land use policies must be reformed with a view to enhancing wellbeing and development in communities.

Finally, it is of primary importance to assess the industrial development model that is creating a social system focused on the profits and productivity of large companies without enforcing the rights of communities to health and wellbeing.

It is noteworthy that accountability mechanisms for environmental issues in Mexico are limited by legal loopholes that prevent populations from protecting themselves against the hazard exposure they often endure. For instance, a diagnosis of the Atoyac River demonstrated that, in Puebla, relationships between enterprises and local and state-level government institutions, as well as between institutions and local residents, were characterized by a lack of transparency and accountability (^{SDUS 2017}).

The two elements that favored our study case were the presence of social mobilization and linkages with the academic community, both of which gave social visibility to the local crisis. Instruments and mechanisms for creating a culture of environmental protection are sorely needed, e.g., external audits, social comptrollers, and provisions for ensuring access to environmental information.

Notwithstanding the above, it should be noted that some progress has been made in the establishment of institutional spaces intended to address social demands for environmental protection. In 2016, the municipality of Puebla established a committee comprised of the three government levels, the business sector and the academic community, with the purpose of conducting a comprehensive diagnosis of the Atoyac River situation. The results served as a basis for the formulation of a master plan to rescue the Atoyac River (^{SDUS 2017}). Although the plan was confined to the municipality of Puebla, it marked a milestone in the management of environmental problems by contemplating transparency and accountability as components of a major strategy.

Finally, this year (2017), the Comisión Nacional de Derechos Humanos (National Commission for Human Rights, CNDH) issued recommendation 10/2017 to the effect that civil servants at all government levels shall be responsible for acts and omissions in the exercise of their functions regarding prevention of pollution. Specific reference is made to the responsibility of the federal, state (Puebla and Tlaxcala) and municipal (Huejotzingo and the other municipalities involved in the public demand) governments. The CNDH recommendation also establishes that the victims of environmental damage shall be provided reparation by implementing the principles of restitution, rehabilitation, satisfaction and guarantees of non-repetition.

In addition, the recommendation mandates identifying, prosecuting, trying and sanctioning the parties liable for the damage with a view to restoring and promoting the conditions required for the enjoyment and exercise of the human right to a healthy environment, clean water, and accessible information (^{CNDH 2017}).

The fact that the study does not report the perceptions and positions of other social actors can be considered a limitation. Nonetheless, our work offers a twofold contribution to risk analysis: 1) it is the first-ever attempt to study environmental risk in Mexico under a methodological approach based on both perception and measurement, and 2) evidence from our integrated analysis of qualitative and quantitative components clearly demonstrates that perception is an accurate indicator for actual and objective risk experienced by environmentally exposed populations.