The world has witnessed a massive increase in the prevalence of diabetes and its complications during the period 1980–2014 (1). The increase in diabetes rate during this period has not been similar in all nations (1, 2). The epidemic of diabetes prevalence is of several orders of magnitude especially in middle- and low-income countries, which are predominantly occupied by rural communities (2). It has been generally thought that urbanization, westernization, and affluence have significant roles in the explosion of diabetes, but we believe that these factors alone could not explain the diabetes epidemic in middle- and low-income countries, especially in rural areas. Intuitively, one would expect a low prevalence of diabetes in a rural population, where adherence to traditional lifestyles includes more physical activity and access to a more nutrient-rich (less processed foods) diet. In recent years, studies from different regions of the world indicated the increasing prevalence of diabetes and other metabolic diseases in rural communities (3–8).

India can be considered as an ideal reference for rural community, as 67% of its population still live in villages. In addition, India houses 17% of the world population and 15.3% of the global diabetes population (1, 2). India witnessed a large increase in diabetes population during 1980–2014 (4.6% increase), but during the same period, no similar significant increase in the level of common risk factors like obesity, hypertension, hypercholesterolemia, and smoking was observed in India, (2). Concurrently during the period of escalation of diabetes incidence, the world has witnessed the massive production and release of toxic chemicals into the environment (2). At present, there is a growing body of evidence that suggests an important role for endocrine-disrupting chemicals (EDCs) in the etiology of diabetes and cardiovascular disorders (2, 9, 10). The major source of EDCs in rural community is exposure to agrochemicals, which include different pesticides (organochlorines, organophosphates, carbamates, pyrethroids, neonicotinoids, etc.) and synthetic fertilizers, which are rich in toxic heavy metals like arsenic and cadmium (8). Epidemiological and experimental studies proved the role of agrochemicals in the development of insulin resistance, glucose intolerance, and pancreatic beta-cell disruption (2–10). The aim of this study is to assess the prevalence of self-reported diabetes and cardiovascular risk factors in a large rural community in South India and in parallel to unravel the association of diabetes and its complications with occupational exposure to agrochemicals.

A total of 106,111 adult individuals from 61 villages in Tamil Nadu (Figure 1) were surveyed in the REDSI study using a mobile app. The adult population of the villages ranged from 728 to 3,485, and the average population per village was 1,751 people. The characteristics of the study population are described in Table 1. We had equal representation of both sexes, and 32% of the population were below 35 years and 10% of the population were aged more than 65 years (Table 1). Of the total population, 51.7% are involved in farming activities, and they are sub-categorized into pesticide applicators, agrochemical handlers, and field workers based on their level of exposure to agrochemicals. In overall, 11.9% of self-reported diabetes prevalence was observed. Among the diabetes complications, we observed 1.9, 1, 0.3, and 0.2% of known CVDs, kidney failure, cancer, and tuberculosis in the rural population. Although there is no significant difference in the prevalence of diabetes complications between farming and non-farming communities, the increase in kidney failure was the highest among the farming community (Table 1).

On further analysis based on occupation, the prevalence of diabetes was 15.0% among the farming community (occupationally exposed to agrochemicals) and 8.7% among the non-farming community (P < 0.0001, Figure 2A). A direct correlation between age and self-reported diabetes was observed, and in all the age groups, the prevalence of self-reported diabetes was much higher in the farming population (Supplementary Figure 1). Among the farming community, people directly handling agrochemicals, including pesticide applicators and fertilizer/pesticide mixers, had self-reported diabetes prevalence of 15.2 and 14.2%, respectively, while the self-reported diabetes prevalence among the field workers was 12.2% (Figure 2B). The family history of diabetes among the self-reported diabetes population was 24.6%, and no significant difference (p = 0.452) was noticed between the farming and non-farming communities. No significant difference between farming and non-farming populations was observed for sex (p = 0.776), daily alcohol intake (p = 0.786), smoking (p = 0.236), oral tobacco use (p = 0.845), and obesity (p = 0.521).

On binary logistic regression analysis, among the traditional risk factors for diabetes, only age (p < 0.001) and hypertension (p < 0.001) showed significant association with high odds ratio for self-reported diabetes in the farming population (Figure 2C). The lesser than expected association of traditional risk factors made us to analyze the association of occupational exposure to agrochemicals with self-reported diabetes prevalence among the farming community. On binary logistic regression analysis, a strong association of exposure to agrochemicals with self-reported diabetes prevalence was observed in all the different exposure categories (pesticide applicators, agrochemical handlers, and field workers) even after adjustment for all confounding variables (Figure 2D). Among the diabetes complications reported in survey, only cardiovascular diseases, kidney failure, cancer, and tuberculosis were validated on cross-reference to medical records in the sub-set of population. A significant association of self-reported diabetes was observed only with CVDs in both farming and non-farming populations (Supplementary Figures 2, 3).

We noticed equal participation of both sexes in the farming activities in the villages of Tamil Nadu, India. Among the age groups, the participation of youth (<35 years) in farming occupation is relatively low, indicating the transition from farming occupation to non-farming occupation among the new generation. Our study shows a nearly two-fold higher prevalence of diabetes in rural farming vs. non-farming communities, seriously raising a research question on the role of agrochemicals in diabetes. As per our previous report (4), there exists around 5–8% of undiagnosed diabetes among rural communities, and hence, the self-reported diabetes survey would have only underestimated the diabetes burden in the rural population. A follow-up study on this rural population will be performed after a period of 3–5 years. A detailed investigation on understanding the molecular mechanism behind the diabetes associated with traditional risk factors like obesity and diabetes associated with EDCs is the need of the hour so that new strategies can be framed to prevent and manage diabetes in the rural world. Among the complications of diabetes, an increased prevalence of kidney failure is associated with farming occupation that is attributed to the increased use of arsenic-rich phosphate fertilizers and glyphosate use. The Sri Lankan Agricultural Nephropathy (SAN) study (11) revealed the association of co-exposure to toxic metals and glyphosate with chronic kidney disease of unknown etiology (CKD-u) among the farming population11. We noticed increased prevalence of cancer among the non-farming communities especially among a few villages in this study, and further investigations revealed the association of coir dyeing and textile industries with cancer incidence in these villages. This study indicates the presence of environmental pollutants of different nature in farming and non-farming villages.

The study has its own limitations. Although the villages for the study were selected based on criteria listed in section “Methods,” the final selection process depended on convenience, access, and contact with the village authorities and willingness of the villagers. The other limitation is that data on dietary habits, physical activity, and education were not collected, but largely, there is no huge variation in these factors within a rural community. The study communities in South India have more intake of carbohydrate diet, which will impact the diabetes prevalence, but this behavior is common to both farming and non-farming communities. In addition, high carbohydrate intake, reduced physical activity due to introduction of machineries in agriculture, non-adherence to diet practices, and medications may also add to the etiology of diabetes in farming community. The self-reported disease status is cross-verified only in 12.8% of the participants, and exposure to agrochemicals is not quantified by measuring the levels of these chemicals in their body fluids.

Altogether, this REDSI study performed among ~0.1 million rural population in Tamil Nadu, South India, indicates the huge burden of diabetes in the rural world. The strong association of occupational exposure to agrochemicals with diabetes in rural communities highlights the need for change in diabetes clinical practice by focusing on occupational safety measures, urging governmental regulatory agencies on safe agricultural policies, and development of new therapeutic strategies targeted at not only understanding the molecular mechanisms but also detoxifying the diabetogenic effect of agrochemicals.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by the Institutional Ethics Committee of KMCH and Madurai Kamaraj University. The study participants underwent validation provided informed consent to participate in the study.

Conceived and designed the experiments: GV, SM, SR, and KS. Acquisition and analysis of data: GV, SM, JC, SKe, BR, SS, AK, SB, AA, PA, TR, MG, JV, VD, SVa, PP, SC, JP, VSS, KA, SVe, GA, SKa, and VG. Statistical analysis of data and interpretation: GV, SM, and KS. Wrote the manuscript: GV. Revised the manuscript: SM, SR, and KS. All authors read and approved the final version of the manuscript.

This study was funded by KMCH Research Foundation, India via grant no. KMCHRF-RDS, 2017.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.