The strongest evidence for the ability for environmental exposures to contribute to the development of diabetes comes from longitudinal studies. These are studies that take place over a period of time, where the exposure is measured before the disease develops.
A study of pesticide applicators in the U.S. found that diabetes incidence increased with the use (both cumulative lifetime days of use and ever use) of some organophosphate pesticides: dichlorvos, trichlorfon, alachlor, cyanazine, and the organochlorine pesticides aldrin, chlordane, and heptachlor. Those who had been diagnosed more than one year prior to the study were excluded, and the participants were followed over time, ensuring that exposures were reported prior to diagnosis. While these people were exposed occupationally, many of these pesticides are available to the general public. This study was based on data from the Agricultural Health Study, which includes over 33,000 participants from Iowa and North Carolina (Montgomery et al. 2008).
Another longitudinal study, also using data from the Agricultural Health Study, looked at exposure data from farmers' wives. It found that diabetes incidence was associated with exposure to five pesticides: three organophosphate pesticides: fonofos, phorate, and parathion; as well as the organochlorine pesticide dieldrin, and the herbicide 2,4,5-T (Starling et al. 2014).
A study of farmers in China found that various pesticides were associated with changes in glucose levels, as well as kidney, liver, and nerve damage (Huang et al. 2016).
Children exposed to modern-day pesticides in the womb (because their mothers worked in greenhouses) had a lower birth weight, and then an increased body fat accumulation from birth to school age than children who were not exposed to pesticides in the womb (the effects were increased if the mothers smoked as well) (Wohlfahrt-Veje et al. 2011). The same authors also found that the metabolic and cardiovascular effects associated with these pesticide exposures depended on genetic background (Andersen et al. 2012; Jørgensen et al. 2016).
Prenatal exposure to the antibacterial agent triclosan was not associated with weight or height in newborn boys, although slightly associated with smaller head circumference. Other phenols were associated with higher birth weight, and the fungicide methylparaben continued to be associated with weight at age 3 (Philippat et al. 2014). Prenatal triclosan was not associated with fat mass during childhood either (age 4-9) (Buckley et al. 2016).
Organophosphorous pesticide levels in the mother were associated with lower birth size in Black U.S. infants, but not in other ethnic groups (Harley et al. 2015).
In a study from France, fetal exposure to organophosphate pesticides was associated with insulin levels as early as birth (Debost-Legrand et al. 2016).
Cross-sectional studies are studies that measure exposure and disease at one point in time. These provide weaker evidence than longitudinal studies, since the disease may potentially affect the exposure, and not vice versa.
A survey of farmers from Saskatchewan, Canada, found that men who worked with insecticides had an increased risk of diabetes as compared to farmers who did not work with insecticides. On the other hand, overall, living on a farm was associated with a decreased risk of diabetes (as compared to other rural residences), probably due to the outdoor lifestyle (Dyck et al. 2013).
During the 1980s and 1990s in the northern U.S. Midwest, death rates from type 2 diabetes were higher in counties that had a higher level of spring wheat farming than in counties with lower levels of this crop. The herbicide 2,4-D is commonly used on this crop. A study compared people who have had a previous exposure to 2,4-D to those who had non-detectable levels of exposure, and found that exposure to 2,4-D was associated with adverse changes in glucose metabolism, a possible predisposing factor for diabetes. The effects were only seen in people with low levels of HDL, the "good" cholesterol (Schreinemachers 2010).
A study of the staff of an Australian insecticide application program found higher mortality rates for diabetes, as compared with the general Australian population, especially people reporting occupational use of herbicides (Beard et al. 2003).
Urinary levels of a dichlorophenol pesticide, 2,5-DCP, has been associated with obesity in U.S. children (Twum and Wei 2011), as well as in U.S. adults (Wei et al. 2014). In both of these studies, the risk of obesity increased as exposure increased, in a dose-dependent manner. These studies suggest that exposure to the fumigant insecticide paradichlorobenzene may increase the risk of obesity.
The dichlorophenol pesticides, 2,4-DCP and 2,5-DCP, were associated with body weight measures (BMI, waist circumference, and obesity) in U.S. adolescents. Triclosan was not associated with any of these measurements in this study (Buser et al. 2014), nor was triclosan associated with obesity in children from India (Xue et al. 2014).
Exposure to high levels of pesticides is common in developing countries, especially organophosphate pesticides. A study of these farmers showed that they had higher blood glucose levels (both fasting and after a glucose tolerance test), as well as neurological symptoms such as depression, as compared to a comparison group who were not exposed (Malekirad et al. 2013).
A study of Egyptian farmers (without diabetes) found that those with higher levels of malathion in their blood had higher insulin resistance, waist circumference, and body mass index (BMI). Not surprisingly, the farmers, who had been working with pesticides for 15-20 years, had higher levels of malathion in their blood than the comparison group who were not farmers (Raafat et al. 2012).
Pesticides may contribute to the growing rates of diabetes in sub-Saharan Africa. People in these countries may be more susceptible to the effects of pesticides due to a variety of factors, such as undernutrition, lack of access to health care, genetic predisposition, high exposure levels, and exposure during developmental periods, such as in the womb and during childhood (Azandjeme et al. 2013).
Comparing pesticide sprayers with unexposed controls in Bolivia, exposure to the pyrethroid pesticides was associated with pre-diabetes and higher blood glucose levels (HbA1c) in the sprayers (Hansen et al. 2014).
There are case studies documented in the scientific literature of people who developed high blood sugar and what was thought to be diabetic ketoacidosis immediately after consuming pesticides (e.g., in a suicide attempt). For example:
A larger study, of 184 people without diabetes who had been poisoned by organophosphate pesticides, found that 121 of them had high blood sugar levels when hospitalized. Those with the highest glucose levels had the highest risk of death. However, the fatality risk varied based on which type of pesticide it was (Moon et al. 2016). These cases present a challenge for endocrinologists (Shahid et al. 2014).
However, acute, large-dose organophosphate pesticide poisonings, while increasing blood glucose and ketone levels in the short run, seem to resolve relatively quickly, and do not seem to trigger diabetes in the first month or so (although longer-term follow-up studies are needed). Nor does prior diabetes increase the risk of mortality from these poisonings, at least in one study from Taiwan (Liu et al. 2014).
A number of organophosphate pesticides have been found to disrupt beta cell function, including malthion (Hectors et al. 2011). Animals exposed to malathion develop high blood sugar levels, and their carbohydrate metabolism is affected in ways that could promote insulin resistance (Rezg et al. 2007; Rezg et al. 2010). In fact, more recent studies shows that malathion does indeed produce insulin resistance in adult rats, in addition to high blood sugar levels (Lasram et al. 2014a), high insulin levels, and high HbA1c, whereas an anti-oxidant protects against these effects (Lasran et al. 2014d). Acute exposure to malathion causes transitory high blood glucose in rats accompanied by glucagon depletion, implying that the malathion caused the liver to release its sugar stores, raising blood sugar levels. Triglycerides and LDL (the "bad" cholesterol) were also increased (Lasram et al. 2009). Both acute and chronic exposures to malathion increased blood glucose and insulin secretion in rats. The higher insulin levels were not enough to overcome the high blood sugar levels (Panahi et al. 2006). In another study, exposure to malathion increased blood glucose and insulin levels in rats, as well as decreased glycogen levels in the liver. Malathion also increased insulin resistance in these rats, which continued one month after exposure ended (Lasram et al. 2014b).
Exposure to low doses of chlorpyrifos for 2-4 weeks resulted in high blood glucose levels in rats (Lukaszewicz-Hussain, 2013). Short-term acute exposure-- a single dose-- raised blood glucose, LDL and triglyceride levels in rats as well (Acker and Noqueria 2012). Long-term exposure of rats to monocrotophos led to glucose intolerance, insulin resistance, and high blood sugar (Nagaraju et al. 2014). Chronic exposure to chlorpyrifos also increased body weight in mice (Peris-Sampedro et al. 2015a); the same authors found that certain genes seem to increase susceptibility to the effects of chloripyrifos (including increasing insulin resistance, blood sugar levels, food ingestion, and cholesterol levels) (Peris-Sampedro et al. 2015b).
Animals exposed to diazinon, another organophosphate pesticide, were found to have impaired glucose tolerance and lower insulin levels (Pakzad et al. 2013). Diazinon has also been found to cause the liver to release glucose into the blood in rats, supporting the idea that diazinon exposure may predispose people to diabetes (Teimouri et al. 2006). In fish, glucose levels rise after exposure to diazinon (Ghasemzadeh et al. 2015). A review summarizes the many potential molecular mechanisms involved in how organophosphorous pesticides can contribute to insulin resistance and type 2 diabetes (Lasram et al. 2014c).
Dichlorvos, another organophosporous pesticide, causes changes in energy metabolism genes in zebrafish (an animal used to test for toxic effects) (Bui-Nguyen et al. 2015).
Long term, low dose exposure to the herbicide atrazine resulted in increased body weight and increased insulin resistance in rats. Those rats that were exposed and also ate a high-fat diet showed exacerbated weight gain and insulin resistance (Lim et al. 2009). Fish exposed to atrazine developed high blood glucose levels-- the higher the dose, the higher the glucose. Seven days after the exposure was removed, blood glucose levels went back to normal. (Other studies have found this same pattern in fish exposed to other pesticides as well) (Blahova et al. 2014).
A fungicide, tolylfluanid, used in paint and on fruit crops, commonly detected in Europe, has been shown to promote the formation of fat cells as well as induce insulin resistance in fat cells. These findings raise a concern that this chemical, an endocrine disruptor, could disrupt metabolism and contribute to the development of diabetes (Sargis et al. 2012). Further investigation shows that tolylfluanid alters fat cell function by activating the glucocorticoid receptor, which plays an important role in controlling metabolism. This mechanism may be a new way that chemicals could promote metabolic diseases such as diabetes and obesity (Neel et al. 2013). Mice exposed to tolylfluanid had more weight gain, higher total fat mass, glucose intolerance, and increased insulin resistance (Reginer et al. 2014).
When researchers exposed fat cells to imidacloprid, a neonicotinoid insecticide (now restricted in Europe due to bee colony collapse disorder), they found that there was increased fat accumulation in these cells (Park et al. 2013). When they exposed fat, liver, and muscle cells to this insecticide, they found that there was increased insulin resistance. Essentially, the exposed cells did not take up as much glucose as unexposed cells did (Kim et al. 2013).
When rats were exposed to omethoate, a commonly used insecticide in most developing countries, the effects suggested that omethoate has the potential to cause insulin resistance (Zhang et al. 2014).
Permethrin, an insecticide used as an insect repellent, impairs glucose homeostasis and alters fat cell development in laboratory studies (Kim et al. 2014). Also, early life exposure to permethrin leads to low vitamin D levels in adult offspring (Fedeli et al. 2013); vitamin D deficiency is linked to diabetes.
One pesticide, TFM, is used to kill sea lampreys (an invasive species) in the Great Lakes. One of the ways is works is by affecting glycogen levels in the lampreys, disturbing energy metabolism (curiously, adults are more susceptible to these effects) (Henry et al. 2015).
Early life exposure to organophosphate pesticides causes metabolic dysfunction resembling pre-diabetes in animals, especially when adults eat a high-fat diet (Slotkin 2011). Male rats exposed to the organophosphate pesticide chlorpyrifos just after birth, showed high insulin levels when not fasting as adults that resembles the metabolic pattern seen in type 2 diabetes in humans (Slotkin et al. 2005).
Male rats exposed to low doses of parathion just after birth showed high blood glucose levels and increased weight gain later in life (Lassiter et al. 2008). These authors point out that animals exposed to organophosphates as adults show increased weight gain and other diabetes-like changes. Exposures in early development may be even more significant. A further study by the same authors found that unlike chlorpyrifos and malathion, the effects of early life parathion exposure in rats lessened by adolescence, although other changes occur later that affect glucose utilization. The effects of parathion were not worsened by a high fat diet, but the effects of this diet and parathion were similar to each other (Adigun et al. 2010).
When pregnant mice were exposed to very low levels (400-times below the EPA's "no observed adverse effect level") of triflumizole, a fungicide used on food and ornamental crops, their offspring had excess fatty tissue, as compared to unexposed controls. Triflumizole also caused stem cells and pre-fat cells to develop into fat cells (Li et al. 2012).
A study found that women who mixed or applied pesticides to crops or repaired pesticide application equipment during the first trimester of pregnancy had a higher risk of developing gestational diabetes. In the women who reported agricultural exposure during pregnancy, the risk of gestational diabetes was associated with the use of four herbicides (2,4,5-T; 2,4,5-TP; atrazine; butylate) and three insecticides (diazinon; phorate carbofuran) (Saldana et al. 2007).
A Canadian study found that first-trimester levels of pesticides generally were not associated with gestational diabetes or impaired glucose tolerance during pregnancy. However women with higher levels of the organophosphate pesticides dimethylphosphate (DMP) and dimethylthiophosphate (DMTP) had a lower risk of gestational diabetes. The authors propose that this could be because of increased fruit/vegetable consumption (which is also associated with higher pesticide levels) (Shapiro et al. 2016).
Only one study has looked at pesticide levels in people with type 1 diabetes. It found that Egyptian children with newly diagnosed type 1 diabetes had higher levels of malathion in their blood than healthy controls (as well as numerous organochlorine pesticides, discussed on the persistent organic pollutants page). However, these children also had lower levels of the organophosphate pesticides chlorpyrifos and profenofos than the control children (El-Morsi et al. 2012).
Pesticides are a food contaminant, as a result of their use in agriculture. Daily ingestion of low doses of diquat, an extensively used herbicide, induces intestinal inflammation in rats. The authors of this study suggest that repeated ingestion of small amounts of pesticides, as could be found in food, may have consequences for human health and may be involved in the development of gastrointestinal disorders (Anton et al. 2000). Another pesticide, carbendazim, reduces the richness and diversity of gut microbiota in mice (Jin et al. 2015). Chlorpyrifos exposure during development affects the gut microbiota, impairs the intestinal lining, and stimulates the immune system of pup rats (Joly Condette et al. 2015). Another study also shows that chlorpyrifos affects gut microbiota, resulting in intestinal inflammation and abnormal intestinal permeability (Zhao et al. 2016). Intestinal disorders like these are common in people with type 1 diabetes (see the diet and the gut page for more studies on this topic).
Organophosphate pesticides have been found to be toxic to the immune system in animals and sometimes humans (Galloway and Handy 2003). Humans chronically exposed to chlorpyrifos have also been found to have increased levels of autoantibodies (Thrasher et al. 2002). In animal studies, chlorpyrifos affects the immune system-- both stimulating and suppressing it (Noworyta-Głowacka et al. 2014). Another pesticide carbaryl, causes inflammation and is toxic to the immune system, causing an unbalanced immune response that promotes autoimmunity in animals (Jorsaraei et al. 2014). Other pesticides are also known to affect the immune system, especially during development (e.g., Jiang et al. 2014).
A review on pesticides and immunotoxicity finds that there is some human and animal evidence indicates that some pesticides can affect the immune system. This evidence, however, is too sparse to be conclusive (Corsini et al. 2013). More research is ongoing (e.g., Lee et al. 2015).
Cross-sectional data from a large U.S. study (NHANES) found that of all toxins measured, the association between autoimmune antibodies (not type 1 specific) and triclosan was statistically significant (Dinse et al. 2015).
Some authors argue that glyphosate, the pesticide used on wheat, is to blame for the autoimmune disease celiac disease (Samsel and Seneff 2013).
One chemical known to cause type 1 diabetes in humans is the now-banned rat poison Vacor. In the late 1970s, a few people tried to kill themselves by eating Vacor, and ended up with type 1 diabetes instead. Vacor destroys beta cells directly, but has also been found to be linked to type 1-related autoimmunity (Karam et al 1980).
Does pesticide exposure affect blood glucose control or the risk of complications in people with diabetes? We do not know yet.
One diabetes complication, non-alcoholic fatty liver disease (NAFLD), is associated with environmental chemical exposure. A search of the toxicological literature found that pesticides were among the most frequently identified chemicals associated with fatty liver in rodent studies. Of the 123 chemicals associated with fatty liver, 44% were pesticides-- especially fungicides and herbicides (Al-Eryani et al. 2014). Chlorpyrifos, for example, promotes fat accumulation in the liver (Howell III et al. 2016).
Some researchers have looked to see whether the pesticide diazinon affects rats with diabetes differently than rats without diabetes. An animal model of type 1 diabetes shows that diabetes increases the toxicity of diazinon (Ueyama et al. 2007). An animal model of type 2 diabetes shows that diazinon worsens glucose tolerance in these rats, leading to higher blood glucose levels (Ueyama et al. 2008).
The organophosphorous pesticide monocrotophos has been shown to exacerbate the complications in rats with chemically-induced type 1 diabetes, including raising glucose levels (Begum and Rajini, 2011), and causing intestinal dysfunction (Vismaya and Rajini, 2014).
To download or see a list of all the references cited on this page, as well as additional article on pesticides related to diabetes, see the collection Pesticides and diabetes/obesity in PubMed.
One additional reference not on PubMed is:El-Morsi DA, Rahman RHA, Abou-Arab AAK. Pesticides Residues in Egyptian Diabetic Children: A Preliminary Study. J Clinic Toxicol. 2012;2:138. Full text