The role of arsenic in obesity and diabetes

Abstract

As many individuals worlwide are exposed to arsenic, it is necessary to unravel the role of arsenic in the risk of obesity and diabetes. Therefore, the present study reviewed the effects of arsenic exposure on the risk and potential etiologic mechanisms of obesity and diabetes. It has been suggested that inflammation, oxidative stress, and apoptosis contribute to the pathogenesis of arsenic‐induced diabetes and obesity. Though arsenic is known to cause diabetes through different mechanisms, the role of adipose tissue in diabetes is still unclear. This review exhibited the effects of arsenic on the metabolism and signaling pathways within adipose tissue (such as sirtuin 3 [SIRT3]‐ forkhead box O3 [FOXO3a], mitogen‐activated protein kinase [MAPK], phosphoinositide‐dependant kinase‐1 [PDK‐1], unfolded protein response, and C/EBP homologous protein [CHOP10]). Different types of adipokines involved in arsenic‐induced diabetes are yet to be elucidated. Arsenic exerts negative effects on the white adipose tissue by decreasing adipogenesis and enhancing lipolysis. Some epidemiological studies have shown that arsenic can promote obesity. Nevertheless, few studies have indicated that arsenic may induce lipodystrophy. Arsenic multifactorial effects include accelerating birth and postnatal weight gains, elevated body fat content, glucose intolerance, insulin resistance, and increased serum lipid profile. Arsenic also elevated cord blood and placental, as well as postnatal serum leptin levels. The data from human studies indicate an association between inorganic arsenic exposure and the risk of diabetes and obesity. However, the currently available evidence is insufficient to conclude that low‐moderate dose arsenic is associated with diabetes or obesity development. Therefore, more investigations are needed to determine biological mechanisms linking arsenic exposure to obesity and diabetes.

Low‐moderate dose arsenic is associated with diabetes or obesity development.

1. INTRODUCTION

Diabetes mellitus (DM) is a metabolic disorder that is characterized by defects in insulin secretion, insulin function, or both (Goodarzi, Mehrpour, & Eizadi‐Mood, 2011). Besides β‐cell failure, the major pathophysiological event contributing to the development of Type 2 DM (T2DM) is insulin resistance, which is usually associated with abnormal insulin secretion (Ginsberg, 2000). The pathogenesis of T2DM involves abnormalities in both insulin function and secretion (Saltiel, 2001; Samarghandian, Azimi‐Nezhad, & Farkhondeh, 2016). Insulin is released in response to β‐cell stimulation and mediates the uptake of glucose, amino acids, and fatty acids into insulin‐sensitive tissues. In turn, these tissues feedback the information to islet regarding their need for insulin (Samarghandian et al., 2016). The mediator of this communication process has not yet been identified, but is likely to involve an integration between the brain and humoral systems (Hruby & Hu, 2015). In insulin resistance, as seen most commonly in obesity, β‐cell increases the insulin output to maintain normal glucose tolerance. However, the shortage of β‐cell function to accomplish this task results in an elevation of plasma glucose levels. Insulin resistance has already been well established in the pathogenesis of impaired glucose tolerance. Therefore, elevated glucose levels, above the normal range, are attributed to a continuous decline in β‐cell function. Furthermore, progressive deterioration of β‐cell function is responsible for the progression of the condition from impaired glucose tolerance to T2DM (Hruby & Hu, 2015).

Free fatty acids (FFAs) are known to play a key role in the progressing loss of insulin sensitivity in T2DM; however, the underlying mechanisms are still unclear. It has been postulated that an increase in the intracellular concentration of fatty acid metabolites activates the serine kinase cascade, leading to defects in insulin signaling originating from insulin receptors (Hajer, van Haeften, & Visseren, 2008). In addition, a complex network of adipokines released from adipose tissue modulates the response of tissues to insulin. Insulin receptor substrate‐2, protein kinase B, and forkhead transcription factor Foxo 1 are molecules involved in the intracellular interactions triggered by insulin. Recent data have shown strong evidence suggesting that dysfunction of these proteins results in insulin resistance in‐vivo (Hajer et al., 2008). Studies have recently revealed that phosphoinositide‐dependent kinase 1‐independent phosphorylation of protein kinase Cε causes a reduction in insulin receptor gene expression. In addition, it has been suggested that mitochondrial dysfunction triggers the activation of several serine kinases and weakens insulin signal transduction (Fantuzzi, 2005). Obesity is characterized with an increased adipose tissue mass and adipocyte dysfunction, leading to the overproduction of proinflammatory cytokines, oxidative stress, endoplasmic reticulum (ER) stress, and insulin resistance (Furukawa, Fujita, & Shimabukuro, 2004). There are many proposed pathophysiological mechanisms involved in the development and progress of obesity. Research on obesity entered a new area after leptin gene was discovered by Zhang et al. (1994). Whereas leptin and ghrelin are produced in peripheral tissues, however, they can control appetite through their actions on central nervous system. Leptin and ghrelin, along with the other appetite‐related hormones, act on hypothalamus, a region of the brain regulating food intake and energy expenditure (Falagas & Kompoti, 2006).

Genetic susceptibility and inappropriate lifestyle such as consuming fat‐rich diet, stress, and low exercise have been recognized as the main risk factors of obesity and diabetes (Özcan et al., 2004). However, several studies have indicated the effects of environmental pollutants on the risk of obesity and diabetes (Maury & Brichard, 2010). Heavy metals may also be involved in the pathogenesis of metabolic syndrome (Maury & Brichard, 2010). A potential relationship has been indicated between exposure to arsenic (Samarghandian, Azimi‐Nezhad, Samini, & Farkhondeh, 2015; Samarghandian, Azimi‐Nezhad, Shabestari et al., 2015; Samarghandian, Borji, Afshari, Delkhosh, & Gholami, 2013), cadmium, mercury (Samarghandian et al., 2016; S.L. Wang et al., 2007; S.L. Wang, Liou, Wang, Li, & Chang, 2010), and lead (Tinkov et al., 2015) and development of diabetes and obesity. Arsenic is one of the major heavy metals which has been recognized as an endocrine disruptor agent. Arsenic exposure has also been noted as a risk factor for cancers, heart disease, diabetes, as well as reproductive, and developmental problems in humans (S.J. Park, Yeum, Choi, Kim, & Joo, 2016). The underlying mechanisms of arsenic toxicity involves modulating cell signaling, controlling cell cycle, including oxidative and ER stresses, promoting inflammatory response, and disturbing DNA repair (S.K. Park et al., 2006). However, there is an association between arsenic exposure and development of diabetes and obesity. Thus, the current review focused on the clinical and experimental scientific literature concerning the effects of arsenic exposure on the pathogenesis of obesity and diabetes. We also have discussed the possible underlying mechanisms of arsenic cytotoxicity.

2. MOLECULAR MECHANISMS OF INSULIN RESISTANCE

Oxidative stress, inflammation, glucotoxicity, and lipotoxicity are the main mechanisms involved in the pathogenesis of insulin resistance (Lasram, Dhouib, Annabi, El Fazaa, & Gharbi, 2014). Oxidative stress plays a main role in the pathogenesis of insulin resistance during hyperglycemia. Oxidative stress induces the activation of certain signaling pathways and insulin signal transduction (Lasram et al., 2014). The insulin signaling pathway activation is caused by promoting oxidative autophosphorylation of the insulin receptor. In addition, the NADPH oxidase activation in response to insulin rapidly causes a moderate generation of hydrogen peroxide (H₂O₂), which is a second messenger of insulin (Lasram et al., 2014). Chronic exposure to free radicals disturbs redox status and can block the insulin signaling pathway, resulting in insulin resistance (Lasram et al., 2014). The activation of proinflammatory pathways can also induce insulin resistance. Increased levels of proinflammatory cytokines, such as tumor necrosis factor‐alpha (TNF‐α), interleukin‐1β (IL‐1β), and IL‐6, have been seen in insulin resistant and type 2 diabetes (Lasram et al., 2014). The proinflammatory cytokines are induced by the activation of the c‐Jun N‐terminal kinase (JNK) and the nhibitory kappa kinase (IKK)/ nuclear factor‐kappa B (NF‐nB) signaling pathways. It was indicated that signaling pathways of JNK (Hirosumi et al., 2002) and IKKþ (Cai et al., 2005) are induced and upregulated in skeletal muscle. These two kinases activate the phosphorylation of the the transcription factors activator protein 1 and NF‐nB, resulting in the increase of the inflammatory response and decrease in insulin sensitivity (Lasram et al., 2014). Glucotoxicity or chronic hyperglycemia impairs insulin secretion and worsens insulin resistance via downregulation of the glucose transporter system. Lipotoxicity is another pathological mechanism involved in diabetes (Lasram et al., 2014). It is similar to glucotoxicity, and acts as a link between obesity and insulin resistance. Increased FFAs block the glucose oxidation and uptake in muscle cells. High levels of serum FFAs disturb the cascade of insulin signaling (Lasram et al., 2014).

3. ARSENIC AS AN ENVIRONMENTAL POLLUTANT

Two types of arsenic, including organic and inorganic, can be naturally found in water, food, air, and soil (Holcomb et al., 2017). It has been confirmed that the toxicity of organic arsenic is less than the inorganic form. Thus, contamination of food with organic arsenic may not be a serious public health threat (Oyagbemi et al., 2017). Arsenic is released into the air, water, and soil as a result of natural and anthropogenic activities (Sarker, Song, & Jhung, 2017). Arsenic is used in agricultural chemicals, herbicides pharmaceuticals, wood preservatives, and also in glass‐making, mining, metallurgical, and semiconductor industries (Sarker et al., 2017). The primary contaminating sources of arsenic for humans are drinking water and agricultural products (Petrick, Ayala‐Fierro, Cullen, Carter, & Vasken, 2000). A high concentration of arsenic is found in seafood, leafy vegetables, rice, apple, and grape juice (Petrick et al., 2000). Arsenic has been detected as an occupational and environmental pollutant, causing multiorgan dysfunctions (Petrick et al., 2000). The minimal risk levels for acute and chronic oral exposure to arsenic are 0.005 and 0.0003 mg/kg/day, respectively (Petrick et al., 2000). Research has shown that the main targets of arsenic intoxication are kidney, liver, bone, respiratory, and reproductive systems (Petrick et al., 2000). Skin, peripheral nervous system, and bone marrow can also be invaded by arsenic (Petrick et al., 2000). Arsenic is transformed to trivalent (III) or pentavalent forms. Among pentavalent arsenic‐containing metabolites are methylarsenite (MAsIII), dimethylarsenite (DMAsIII), methylarsenate (MAsV), and dimethylarsenate (DMAsV; Kain, Prabhu, & Halade, 2014). Strong evidence suggests that trivalent methylated species, including MAsIII, and DMAsIII are more toxic than the unmethylated arsenic in laboratory models. Although methylated arsenic species are excreted in urine, the methylated species of arsenic contribute to adverse health effects in humans (Kain et al., 2014). The biological half‐life of arsenic is about 10 hr in humans. Although approximately 70% of arsenic is mainly excreted in urine (Kain et al., 2014), high arsenic exposure causes adverse kidney complications. It has also been indicated that arsenic can affect fetal brain (Kain et al., 2014). It was proposed that the toxic effects of arsenic may be related to the induction of inflammation, oxidative stress, and apoptosis in tissues (Kain et al., 2014). Figure 1 indicates the mechanism of arsenic toxicity involved in obesity and diabetes.

Figure 1.

The mechanism of As toxicity involved in diabetes and obesity. As: arsenic; ROS: reactive oxygen species

4. ARSENIC, OXIDATIVE STRESS, APOPTOSIS, AND AUTOPHAGY

Reactive oxygen species (ROS) are produced by living organisms as a result of normal cellular metabolism. The ROS exposure is also warranted by environmental factors, such as air pollutants (Alizadeh, Hassanian‐Moghaddam, Shadnia, Zamani, & Mehrpour, 2014; Karrari, Mehrpour, Afshari, & Keyler, 2013). Low to moderate concentrations of ROS contribute in physiological cell processes; however, higher concentrations of ROS produce adverse effects toward cellular components such as lipids, proteins, and DNA (Alinejad, Aaseth, Abdollahi, Hassanian‐Moghaddam, & Mehrpour, 2018). The shift in the oxidant/antioxidant balance in the favor of oxidants is termed “oxidative stress.” Regulating the redox potential within cells is critical for cell viability, activation, proliferation, and finally organ function (Alinejad et al., 2018). Oxidative stress contributes to many pathological conditions, including cancer, neurological disorders, atherosclerosis, hypertension, ischemia/perfusion, diabetes, acute respiratory distress syndrome, idiopathic pulmonary fibrosis, chronic obstructive pulmonary disease, and asthma. ROS‐mediated oxidative damage is a common denominator in arsenic‐induced pathogenesis (Ghaderi et al., 2015). In addition, arsenic induces morphological changes in mitochondrial integrity. Initiation of cascades leads to free radical formation in combination with attenuated cellular reducing potential (i.e., glutathione [GSH]) can increase cellular sensitivity to arsenic toxicity. Research has indicated an increased formation of ROS/reactive nitrogen species (RNS), including peroxyl radicals (ROO•), superoxide radical, singlet oxygen, hydroxyl radical (OH•), H₂O₂, dimethyl‐arsenic radical, dimethyl‐arsenic peroxyl radical, and/or oxidant‐induced DNA damage when both humans and animals are exposed to arsenic (Lee & Ho, 1994). Oxidation of lipids by arsenic generates several bioactive molecules such as ROS, peroxides, and isoprostanes of which aldehydes (malondialdehyde [MDA] and 4‐hydroxy‐nonenal [HNE]) are the major end products (Srivastava et al., 2009). Arsenic‐induced oxidative stress is either directly caused by oxidizing –SH groups, or indirectly via production of ROS. Because of its high affinity for sulfhydryl groups, arsenic can deplete cellular antioxidants such as GSH, superoxide dismutase (SOD), catalase, glutathione‐s‐transferase, and glutathione peroxidase in various tissues (Moon, Navas‐Acien et al., 2017; Moon, Oberoi et al., 2017). Experimental studies indicated that arsenic increased the lipidperoxidation (LPO) content in tissues through inducing ROS production (Li, Guo et al., 2017; Li, Wang et al., 2017; Li, Yang, Dong, & Wang, 2017). Several studies have illustrated that arsenic‐exposed animals expressed increased MDA levels in tissues and plasma compared with unexposed animals (Duan et al., 2017; Haque, Chaudhary, & Sadaf, 2017). Arsenic also triggered the production of several free radicals such as H₂O₂, superoxide anion (O2−), hydroxyl radical species (HO−), nitric oxide (NO−), dimethyl arsenic peroxyl radical [(CH3)2 AsOO−], and dimethyl arsenic radical [(CH3) 2As−] (Mehrpour, Keyler, & Shadnia, 2009). It was described that hydroxyl radicals initiate LPO through iron‐catalyzed Fenton reaction in cell membranes (Wei, Guo, Rui, Wang, & Feng, 2017). Research has also shown that the low concentration of arsenic triggered apoptosis, whereas higher concentrations caused necrosis (Li, Guo et al., 2017; Li, Wang et al., 2017; Li, Yang et al., 2017). Apoptosis plays a critical role in acute and chronic arsenic intoxication (Li, Guo et al., 2017; Li, Wang et al., 2017; Li, Yang et al., 2017). Apoptosis is a vital component of various processes, including normal cell turnover, hormone‐dependent atrophy, immune system function, chemical‐induced cell death, and proper embryonic development (X.M. Wang et al., 2014). Dysregulated apoptosis pathways are involved in many diseases, including neurodegenerative diseases, autoimmune disorders, and many types of cancers. The two best‐understood apoptotic mechanisms are intrinsic (also called the mitochondrial pathway) and extrinsic pathways (X.M. Wang et al., 2014). Caspases, central mainstay enzymes in apoptotic signaling network, are activated in almost all apoptotic pathways. Initiator caspases including Procaspases‐2, ‐8, ‐9, and ‐10 are involved in the extrinsic apoptotic pathway. This pathway is promoted by death effector domains and death‐inducing signaling complex (DISC), a membrane receptor complex forming following to incorporation with a member of tumor necrosis factor receptor family (X.M. Wang et al., 2014). It is assumed that several Procaspase‐8 molecules are in close proximity to each other and activate each other by autoproteolysis when bound to DISC. The initiator caspases cleave and thereby activates effector caspases (Procaspases‐3, ‐6, and ‐7), inducing apoptosis. Intrinsic apoptotic pathway involves Procaspase‐9, which initiates mitochondrial proapoptotic events (X.M. Wang et al., 2014). After that, a cytosolic death signaling protein complex called apoptosome formed upon the release of cytochrome C from mitochondria. Research has suggested that mitochondrial ROS plays a critical role in arsenic‐induced apoptosis (X.M. Wang et al., 2014). Deletion of ROS‐scavenging enzyme CuZnSOD enhanced the arsenic‐induced apoptotic effects indicating a central role for ROS in this process. Moreover, a reduction in mitochondrial membrane potential is accompanied by increasing ROS production in arsenic treated cells. This study highlighted functional mitochondrial alterations during arsenic‐induced apoptosis. Therefore, oxidative stress and increased permeability of mitochondrial membranes may contribute to the pathogenesis of arsenic‐induced apoptotic cell injury (Urrialde et al., 2017). Upregulation of autophagy is also considered a possible mechanism of arsenic‐induced diabetes. Autophagy promotes clearing the cells from damaged proteins or other intracellular structures; its dysregulation causes aging and several diseases. The association between autophagy and various types of metabolic disorders has been also reported (Gao et al., 2018). There are several types of autophagy in the cells, including macroautophagy, microautophagy, and chaperone‐mediated autophagy. It was indicated that the arsenic changes autophagy pathway in liver cells via a ROS‐dependent pathway. Arsenic also induced diabetes by autophagy upregulation (Gao et al., 2018).

5. ARSENIC AND INFLAMMATION

Inflammation is a pervasive phenomenon operated in condition with severe perturbations of homeostasis, such as infection, injury, and exposure to contaminants (Kain et al., 2014). It is triggered by innate immune receptors that recognize pathogens and damaged cells (Kain et al., 2014). Among vertebrates, the inflammatory cascade is a complex network of immunological, physiological, and behavioral processes that are coordinated by cytokines and immune signaling molecules (Kain et al., 2014). The initial inflammation consists of three subphases: acute, subacute, and chronic. The acute phase typically lasts 1–3 days and is characterized by five classic clinical signs: heat, redness, swelling, pain, and loss of function. The subacute phase may last from 3 to 4 days to about a month, corresponding to a cleaning phase required before the repair phase (Hall et al., 1994). If the subacute phase is not resolved within about one month, inflammation is said to have become chronic and can last for several months (Hall et al., 1994). Tissues can be degenerated and particularly in the locomotor system, chronic inflammation may lead to tearing and rupture of tissues. Alternatively, after subacute inflammatory phase, tissues can repair themselves and be regenerated during the remodeling phase (Hall et al., 1994). From a mechanistic point of view, the acute response to tissue damage is triggered in microcirculation at the site of injury. Initially, there is a transient constriction of arterioles; however, within several minutes, chemical mediators are released at the injury site, contributing to arteriolar smooth muscle relaxation, leading to vasodilation and thereby increased capillary permeability (Iantorno et al., 2014). A protein‐rich fluid then is exuded from capillaries into interstitial space. This fluid contains plasma components such as albumin, fibrinogen, kinins, complement proteins, and immunoglobulins, which mediate inflammatory response (Iantorno et al., 2014). The subacute phase is characterized by moving of phagocytic cells to the injury site (Iantorno et al., 2014). In response to inflammation, adhesion molecules are expressed on activated endothelial cells, leukocytes, platelets, and erythrocytes in injured vessels (Iantorno et al., 2014). Polymorphonuclear leukocytes, such as neutrophils, are the first cells to infiltrate the injury site. Basophils and eosinophils are more pronounced in allergic reactions or parasitic infections (C.H. Wang et al., 2002). As inflammation continues, macrophages actively remove damaged cells. If the cause of injury is eliminated, the subacute phase of inflammation may be followed by a period of tissue repair. Blood clots are removed by fibrinolysis, and damaged tissues are regenerated or replaced with fibroblasts, collagen, or endothelial cells. During the remodeling phase, new collagen is laid down during the repair phase (mainly Type III collagen), which is progressively replaced by Type I collagen to be assimilated into the original tissue (C.H. Wang et al., 2002). However, if inflammation becomes chronic, further tissue destruction and/or fibrosis may occur (C.H. Wang et al., 2002).

Arsenic exposure is one of the main risk factors for public health leading to multiple outcomes such as atherosclerosis, cardiovascular disease, renal disease, and cancer (Moon, Navas‐Acien et al., 2017; Moon, Oberoi et al., 2017). Arsenic exposure leads to the activation of specific cell types (e.g., Kupffer cells in the liver), as well as enhanced release of proinflammatory, and also anti‐inflammatory cytokines (Peters et al., 2015). When tissues are exposed to arsenic, nuclear transcription factor‐κB (NF‐κB) is induced and the levels of ROS are increased. NF‐κB is attached to regulatory proteins named inhibitors of κB (IκB) and remains an inactive molecule in cytosol (Agrawal, Bhatnagar, & Flora, 2015). IκBα and IκBβ are the main proteins involved in NF‐κB primary and sustained inhibitions, respectively. NF‐κB is activated under inflammation. After moving to nucleus, it recognizes the promoter region and regulates the transcription of proinflammatory genes, including inducible NO synthase (iNOS) and cyclooxygenase‐2 (COX‐2). By inducing NO production, both iNOS and COX‐2 promote vital roles in the pathophysiology of inflammation. Although low concentrations of NO are sufficient to maintain physiological functions, elevated NO levels exert genotoxic harms to the host (Agrawal et al., 2015). In this context, the effects of arsenic have been studied on atherosclerotic lesion formation and lesion composition in ApoE^−/− mice. The results have indicated that exposure to arsenic at early postnatal and adulthood increases the atherosclerotic lesion in the aortic valve and arch. Macrophage recruitment increased expression of proinflammatory mediators, including IL‐6, MCP‐1, HNE, and MDA, have been observed in the lesions exposed to arsenic in mice. Studies have indicated that arsenic exposure induced atherosclerotic lesion formation through increasing inflammation (Kain et al., 2014) In American‐Indian adults (45–75 years old), an association has been reported between the low–moderate arsenic levels in urine and plasma indices of thrombosis and inflammation. The results indicated that arsenic levels in urine were positively associated only with baseline fibrinogen level in diabetes patients. However, urine arsenic levels were not associated with the plasma concentration of plasminogen activator inhibitor‐1. Nevertheless, these observations could not confirm an absolute role for thrombosis and inflammation in the pathogenesis of arsenic‐induced lesions (Kain et al., 2014).

Arsenic activates proinflammatory factors and upregulates the expression of IL‐6, IL‐8, TNF‐α, NF‐κB, and other chemokines in various cell types (Agrawal et al., 2015).The low level of arsenic also promotes proinflammatory cytokines involved in the activation of CD14 expressing monocytes and prolonged persistence of pulmonary and systemic inflammation. In turn, these alternations may promote oxidative DNA damage in human (Agrawal et al., 2015).

6. ARSENIC AND OBESITY

6.1. Experimental studies

In‐vivo and in‐vitro studies have shown the effects of arsenic exposure on adipose tissue and the pathogenesis of obesity. It has recently been demonstrated that arsenic significantly decreased the size of adipocytes or white adipose tissue (WAT). Garciafigueroa and his colleague have illustrated a significant decrease in the expression of adiponectin and leptin in adipocytes derived from primary human mesenchymal stem cells (hMSC) exposed to arsenic over 72 hr. In addition, adiponectin messenger RNA (mRNA) levels were reduced in the hMSC exposed to low doses (0.1–2.5 mM) of arsenite for 12 days at the beginning of the differentiation process (Garciafigueroa, Klei, Ambrosio, & Barchowsky, 2013). Mice exposed to low levels of sodium arsenite (10 ppb) in utero showed a significant reduction in adiponectin during the first postnatal month (Kozul‐Horvath, Zandbergen, Jackson, Enelow, & Hamilton, 2012). Low birth weight is one of the main risk factors for obesity and other metabolic disorders in adulthood. Therefore, the effect of arsenic function may be associated with induction of obesity (Kozul‐Horvath et al., 2012). Recent research has shown that arsenic contaminated drinking water increased leptin levels in serum, placental tissue (Saggese, Fanos, & Simi, 2013), and cord blood (Rodriguez et al., 2016) in offspring (Renu et al., 2017) of pregnant rodents and women. It has been postulated that the increased level of leptin in umbilical cord blood was related to high birth weight and also to the risk of obesity in future (Ahmed et al., 2011). Further works have also proved that exposure to low dose arsenite (up to 2 mM) for 8 weeks decreased glucose transporter type 4 (GLUT‐4) expression upon insulin stimulation in 3T3‐L1 adipocytes primed to differentiate (Gossai et al., 2015). An in‐vitro study indicated that incubation of adipocytes with arsenic disrupts multiple metabolic pathways. In particular, it has been shown that arsenic exposure caused a significant arsenic accumulation in adipose tissue and a decrease in depot, size, and perilipin (PLIN1, a lipid droplet coat protein) expression of epididymal adipose tissue as well as an increase in perivascular ectopic fat deposition in skeletal muscle in mice (Karoutsou & Polymeris, 2012). Another in‐vitro study indicated that exposure to arsenic caused a dose‐dependent reduction in lipid accumulation in 3T3‐L1, 3T3‐F442A, and C3H 10T1/2 cells at the stage of preadipocyte differentiation (Xue et al., 2011). It has also been observed that arsenic exposure significantly modified the expression of adipogenesis activators, CCAAT/enhancer‐binding protein alpha (C/EBPα), and peroxisome proliferator‐activator receptor gamma (PPARγ; Ambrosio et al., 2014). The prominent effects of arsenic have been investigated on carbohydrate metabolism in adipocytes. The exposure to sodium arsenite (10 mM), trivalent monomethylarsonic acid (MMA; 5 mM), and sodium arsenate (pentavalent arsenic; 1 mM) for 4 hr decreased glucose uptake in 3 T3‐L1 adipocytes. Lower doses of arsenic (20e50 mM sodium arsenite or 1e2 mM trivalent MMA) significantly decreased insulin‐stimulated glucose uptake (ISGU; Andersen & Pi, 2013). The inhibition of the PDK‐1/Akt insulin signaling pathway by impairing Akt phosphorylation at Ser473 and Thr308 may cause a reduction in ISGU, and in turn diminish the translocation of GLUT4 from perinuclear compartment to plasma membrane in response to insulin (Andersen & Pi, 2013). The above mechanism alleviates the effects of oxidative stress through a significant decrease in mitochondrial deacetylase Sirt3 expression and also reduces binding activity of its related transcription factor forkhead box O3 (FOXO3a) to the promoters of PPARγ coactivator (PGC) −1a gene (Wauson, Langan, & Vorce, 2002). Therefore, these findings supported that arsenic can modulate adipose tissue metabolism and provoke an obesity‐related metabolic syndrome. Arsenic exposure has also been shown to inhibit adipogenesis, enhance lipolysis, as well as downregulate the levels of serum adiponectin and leptin. Arsenic‐induced modulations in adipose tissue may be executed through the regulation of genes involved in adipogenesis (Paul, Harmon, Devesa, Thomas, & Styblo, 2007; Paul, Hernández‐Zavala et al., 2007) such as Sirt3 and FOXO3, which modulate oxidative stress and decrease the mitochondrial respiration rate in adipose tissue. Arsenic can enhance CHOP10 expression, which in turn encourages an effective role in the induction of diabetes by deregulation of adipogenesis. Arsenic also inhibits adipogenesis by reducing lipid droplets and enhancing lipolysis (Padmaja Divya et al., 2015). The body weights of mice exposed to 100 μg/L sodium arsenite in utero until 13 weeks after birth, increased versus controls starting at 5 weeks of age (Ditzel, Nguyen, Parker, & Camenisch, 2015). The inhibition of adipogenesis by arsenic may lead to lipodystrophy and insulin resistance. However, molecular mechanisms are yet to be elucidated. On the other hand, arsenic‐induced obesity can further accelerate insulin resistance development. In both arsenic‐induced lipodystrophy and obesity conditions, fat ectopic accumulation is observed in peripheral adipose issues. Table 1 summarizes the experimental studies on association between arsenic exposure and risk of obesity.

Table 1.

Experimental studies on the association between arsenic exposure and risk of obesity

Author, year	Experimental model	Types of arsenic	Dose and duration of exposure	Findings
Garciafigueroa et al. (2013)	Mice	Arsenic	100 μg/L 5 weeks	Decreased the adipose tissue expression of perilipin
Kozul‐Horvath et al. (2012)	Mice	Arsenite	10 ppb exposure during embryonic	Decreased adiponectin during the first postnatal month
Xue et al. (2011)	3T3‐L1 adipocytes	Arsenite	Up to 2 mM for 8 weeks	Decreased GLUT4 expression
Ambrosio et al. (2014)	Mice	Arsenite	100 μg/L for 5 weeks	Increased perivascular ectopic fat deposition in skeletal muscle
Wauson et al. (2002)	C3H 10T1/2 cell	Arsenite	6 µM for 8 weeks	Decreased adipocytic differentiation
Paul, Harmon et al., (2007) and Paul, Hernández‐Zavala et al. (2007)	3 T3‐L1 adipocytes	Arsenite (iAs(III))	5 and 10 mM for 4 hr	Inhibited the PDK‐1/PKB/Akt‐mediated transduction step is the key mechanism for the inhibition of ISGU in adipocytes exposed to iAs(III) or MAs(III)
		Methylarsonous acid (MAs(III))	1 mM for 4 hr
Ditzel et al. (2015)	Mice	Arsenite	100 μg/L in utero until 13 weeks after birth	Increased body weight

Note. GLUT4: glucose transporter type 4; ISGU: insulin‐stimulated glucose uptake; PDK‐1: phosphoinositide‐dependant kinase‐1; PKB: protein kinase B.

6.2. Epidemiological studies

Epidemiological studies on the association between arsenic exposure and obesity are rather inconsistent. In this context, it has been indicated that urinary arsenic levels were associated with an increase in body mass index (BMI) in adult women exposed to arsenic in drinking water (Gomez‐Rubio et al., 2011). However, a prospective study in a Taiwanese population exposed to arsenic‐contaminated drinking water indicated no association between BMI and arsenic exposure. Nevertheless, the urinary arsenic levels were associated with BMI in patients with T2DM suggesting a relationship between arsenic‐exposure and BMI (Tseng, Tai, & Chong, 2000). Controversially, several epidemiological studies have illustrated a negative interaction between arsenic exposure and obesity indices. A recent survey indicated no correlation between urinary arsenic levels and fat mass in voluntary childbearing‐age women (Ronco et al., 2010). Chronic consumption of low dose arsenic contaminated water had no significant effects on the BMI of women and men aged >30 years (Islam et al., 2012). Likewise, no significant relationship was seen between BMI and urinary arsenic levels in Korean adults older than 20 years old (Bae, Ryu, Choi, & Park, 2013). Another study on Taiwanese adolescents indicated that the total urinary arsenic levels were significantly lower in subjects with higher BMI (Grashow et al., 2014). Similarly, there was a significant inverse association between arsenic levels and BMI in arsenic‐exposed welders (Grashow et al., 2014). This effect was independent of smoking status or alcohol consumption (Gruber et al., 2012). Data from the National Health and Nutrition Examination Survey indicated that arsenic exposure is not associated with obesity in pregnant or breastfeeding women (Bulka, Mabila, Lash, Turyk, & Argos, 2017).

Thus, these findings indicated that the effects of arsenic exposure on body weight may be variable, based on multiple epidemiological factors. The inconsistencies between the above results may be partly related to the differences in arsenic dose and route of exposure. It was indicated that maternal total arsenic levels are associated with an elevated risk of gestational diabetes (GDM). Table 2 summarizes the epidemiological studies on the association between arsenic exposure and risk of obesity.

Table 2.

Epidemiological studies on the association between arsenic exposure and risk of obesity

Author, year	Study design	Study population	Obesity diagnosis	Sample	As level	Outcome
Gomez‐Rubio et al. (2011)	Cross‐sectional	624 adult women from state of Arizona in the United States, and the state of Sonora in Mexico	BMI	Urine (µg/L)	53.6	The strong association between high BM and urinary As
Tseng et al. (2000)	Cohort	446 nondiabetic residents in arseniasis‐hyperendemic villages in Taiwan	BMI	Urine (µg/L)	49	The association between high BM and urinary As in new diagnostic diabetic patient
Ronco et al. (2010)	Cross‐sectional	107 childbearing‐age women residents in area of Santiago city with no evidence of As contamination sources	BMI	Urine (µg/L)	50	No association between high BM and urinary As
Islam et al. (2012)	Cross‐sectional	1,004 consenting women and men residents in Bangladesh exposed to As‐contaminated drinking water	BMI	Urine (µg/L)	>50	No association between high BM and urinary As
Bae et al. (2013)	Cross‐sectional	580 adults resident in two districts of urban, rural and costal area in Korea that not been exposed to As occupationally	BMI	Urine (µg/L)	0.36–36.7	No association between high BM and urinary As
Grashow et al. (2014)	Cross‐sectional	74 welders with known As exposure	BMI	Toenail (µg/g)	0.18	The inverse association between BMI and toenail As
Bulka et al. (2017)	Cohort	3,097 females who self‐reported being pregnant or breastfeeding, or who had a positive urine pregnancy test	BMI	Urine (µg/L)	1.25	No association between high BM and urinary As

Note. As: arsenic; BMI: body mass index.

7. ARSENIC AND T2DM

7.1. Experimental data

The side effects of arsenic exposure on the regulation of carbohydrate metabolism have been investigated in experimental models for years. It was indicated that oral exposure of mice to arsenic (50 ppm) impaired glucose tolerance and increased the concentration of arsenic and its metabolites within the liver and other organs inflicted by T2DM including pancreas, skeletal muscle, and adipose tissue (Paul, Harmon et al., 2007; Paul, Hernández‐Zavala et al., 2007). The effects of low‐dose (3 ppm) subchronic arsenic exposure on glucose homeostasis have been shown in rats from prenatal development stages through adult life. Arsenic exposure caused an impaired glucose tolerance test, β‐cell damage, hyperglycemia, and also increase the levels of homeostatic model assessment of insulin resistance (HOMA‐IR), glycosylated hemoglobin, cholesterol, and pancreatic insulin in arsenic‐exposed rats. The research also confirmed an association between arsenic exposure and diabetes throughout prenatal developmental stages until adult life (Kawaguchi, 1981).

Previous research has indicated that arsenic can induce hyperglycemia through influencing central nervous and adrenal systems in rats (Kawaguchi, 1981).

The oral administration of arsenic (1.5 and 5 mg/kg) triggers an increase in blood glucose through enhancing oxidative stress in hepatic and pancreatic tissues in experimental animals. Arsenic exposure results in metal accumulation in hepatic and pancreatic tissues compared with controls. The research has also indicated that arsenic induced diabetes through stimulating pancreatic and hepatic oxidative stress in an experimental model (Rezaei, Khodayar, Seydi, Soheila, & Parsi, 2017). Another study showed that acute arsenic exposure can atteneute normal glucose tolerance due to overproduction of ROS and oxidative stress in rats (Bonaventura et al., 2017). The administration of arsenic (5 and 50 ppm) for 8 weeks caused glucose intolerance via altering β‐cell function in pregnant rats and their female offspring (Bonaventura et al., 2017). Palasioc et al. (2012) have shown that low level of arsenic (30 ppb) could induce insulin resistance in male rats. In particular, oral glucose tolerance test and insulin resistance were significantly increased in female rats exposed to arsenic. Lu et al. (2011) indicated that arsenic exposure may exploit various mechanisms in the development of insulin resistance. The authors have shown that arsenic exposure caused a significant increase in ROS and MDA formation in pancreatic β‐cell‐derived RIN‐m5F cells. In parallel, research has suggested that arsenic exposition induces the activation of MAPKs, mitochondria dysfunction, upregulation of p53, downregulation of Bcl‐2, and Mdm‐2 as well as poly (ADP‐ribose) polymerase (PARP), and caspase cascade activation. In addition, arsenic exposure of RIN‐m5F cells led to the activation of ER stress‐related molecules (such as GRP78, GRP94, CHOP, and XBP1), Procaspase‐12 cleavage, and calpain activation. Arsenic exposure of mice through drinking water for 6 weeks significantly decreased plasma insulin, glucose intolerance, LPO, and induced islet cell apoptosis. Studies have indicated that arsenic‐related oxidative stress induced pancreatic β‐cells apoptosis through activating mitochondria‐dependent and ER stress‐triggered signaling pathways (Díaz‐Villaseñor et al., 2008).

Arsenic has decreased insulin secretion in RINm5F cells through a calcium–calpain pathway‐dependent insulin exocytosis. Arsenic can further impair insulin secretion by reducing oscillations of free [Ca(2+)]i and decreasing calcium‐dependent Calpain‐10 induced partial proteolysis of SNAP‐25 (Díaz‐Villaseñor et al., 2008). A high‐fat diet in combination with arsenic 25 or 50 ppm in drinking water for 20 weeks deteriorated hyperglycemia, insulin resistance, and lipid profile by inducing mitochondrial oxidative stress in rats (Ahangarpour et al., 2018). Low‐level inorganic arsenic‐induced insulin resistance by activating PPARγ–mTOR Complex 2 (mTORC2) signaling and inhibiting hepatic autophagy (Gao et al., 2018). Table 3 shows experimental research on the associations between arsenic exposure and risk of diabetes.

Table 3.

Experimental studies on the association between arsenic exposure and risk of diabetes

Author, year	Experimental model	Types of arsenic	Dose and duration of exposure	Findings
Paul, Harmon et al. (2007) and Paul, Hernández‐Zavala et al. (2007)	Mice	Arsenite	25 or 50 ppm for 8 weeks	Impaired glucose tolerance
Kawaguchi (1981)	Rat	Arsenic trioxide	3 ppm subchronic arsenic exposure	Impaired glucose tolerance test, β‐cell damage, hyperglycemia, and also increased the levels of HOMA‐IR, glycosylated hemoglobin, cholesterol, and pancreatic insulin
Rezaei et al. (2017)	Rat	Arsenic	1.5 and 5 mg/kg	Induced diabetes through stimulating pancreatic and hepatic oxidative stress
			Acute exposure
Bonaventura et al. (2017)	Pregnant rats and their female offspring	Arsenic	5 and 50 mg/L for 8 weeks	Impaired normal glucose tolerance due to overproduction of ROS
Palasioc et al. (2012)	Rat	Arsenic	Arsenic (30 ppb)	Induce insulin resistance
Lu et al. (2011)	Pancreatic β‐cell‐derived RIN‐m5F cells	Inorganic arsenic	5 mM for 4 hr	Increased in ROS and MDA formation in pancreatic β‐cell‐derived RIN‐m5F cells
				Induced the activation of MAPKs, mitochondria dysfunction, upregulation of p53, downregulation of Bcl‐2, and Mdm‐2 as well as PARP, and caspase cascade
Díaz‐Villaseñor et al. (2008)	RINm5F cells	Arsenic	0.5–2 µM	Decreased insulin secretion through calcium–calpain pathway dependent insulin exocytosis
Ahangarpour et al. (2018)	High fat diet rat	Arsenic	25 or 50 ppm for 20 weeks	Impaired hyperglycemia, insulin resistance, and lipid profile by inducing mitochondrial oxidative stress
Gao et al. (2018)	HepG2	Inorganic arsenic	4 µM for 2 hr	Induced insulin resistance by activating PPARγ–mTORC2 signaling and inhibiting hepatic autophagy

Note. HOMA‐IR: homeostatic model assessment of insulin resistance; MAPKs: mitogen‐activated protein kinas; MDA: malondialdehyde; mTORC2: mTOR complex 2; PARP: poly (ADP‐ribose) polymerase; PPARγ: proliferator‐activator receptor gamma; ROS: reactive oxygen species.

7.2. Epidemiological studies

In contrast with obesity, the relation between arsenic exposure and T2DM has been more extensively investigated. In particular, multiple cross‐sectional studies have shown a significant relationship between diabetes and exposure to arsenic.

Data analysis of NHANES III (2003 and 2004) involving 1,279 participants (160 diabetic patients) aged 20 years or older indicated that urinary arsenic levels were associated with diabetes in a dose‐dependent manner (Navas‐Acien, Silbergeld, Pastor‐Barriuso, & Guallar, 2009). Another cross‐sectional study in Bangladesh involving 1,004 adults demonstrated a significant direct association between arsenic levels in drinking water and T2DM in a model adjusted for age, sex, education, BMI, and family history of diabetes (Islam et al., 2012). In a longitudinal study on 11,319 subjects in Araihazar and Bangladesh, a significant association was observed between urinary arsenic and glycated hemoglobin (HbA1c) levels in both men and women. Whereas no association was found between urinary arsenic levels and glucose levels in a recent study. Moreover, no relationship was observed between urinary arsenic levels and diabetes incidence (Chen et al., 2010).

A significant association has been detected between hair and blood arsenic levels and impaired glucose tolerance. In particular, the odds of impaired glucose tolerance test was 2.8 higher in pregnant arsenic‐exposed women than nonexposed females. After adjusting for various factors, including age, pregnancy BMI, race, medication usage, and marital status, the association remained significant in blood highest quintile (Ettinger et al., 2009). It is also notable that findings of a cohort study indicated a significant association between higher water arsenic levels and the incidence of GDM and glucose intolerance after adjusting for obesity (Farzan et al., 2016).

The urinary arsenic levels have been found to be increased in Korean patients with T2DM. The levels significantly related to fasting plasma glucose (Rhee et al., 2013). It has also been indicated that urinary arsenic levels were comparable between diabetic patients regardless of their smoking status (Huang, Cheng, Sung, Guo, & Sthiannopkao, 2014). In a research conducted in Cambodia on 43‐male and 99‐females exposed to arsenic‐contaminated drinking water showed that arsenic doses above the medium levels (907.25 μg/L) were associated with a nearly twice increase in the risk of T2DM. However, the research further stipulated no significant difference in T2DM prevalence in participants with urine arsenic levels above or below the median. Another study also represented higher urinary arsenic levels in Iranian diabetic patients (Mahram, Shahsavari, Oveisi, & Jalilolghadr, 2013). The association between urinary arsenic levels and HOMA of β‐cell function was investigated in 369 Korean diabetic patients. The research indicated an association between urinary arsenic levels and pancreatic β‐cell dysfunction in healthy subjects, especially in males (Baek, Lee, & Chung, 2017). In an investigation in r4ww individuals who were occupationally or environmentally exposed to high arsenic levels, a significant association was found between arsenic exposure and T2DM. The findings of a large prospective cohort study in an arsenic polluted area of Denmark indicated a significant increase in diabetes prevalence in individuals who were exposed to arsenic‐contaminated drinking water for a long‐term period (Bräuner et al., 2014). A 3‐year follow‐up study in polluted areas in Arizona, Oklahoma, North and South Dakota also indicated a significant increase in blood glucose levels in the exposed group. Urine arsenic levels were associated with diabetes in a rural population in the United States with a high burden of diabetes (Gribble et al., 2012). Controversially, no association was described between exposition to low‐level arsenic‐containing water and diabetes incidences in the residents of three villages in Ron Phibun subdistrict, Nakhon Si Thammarat Province (Sripaoraya, Siriwong, Pavittranon, & Chapman, 2017). A recent systematic review and an epidemiological study showed that diabetes development may be related to lower and higher activities of MMA and dimethylarsinic acid (DMA), respectively (Kuo, Moon, Wang, Silbergeld, & Navas‐Acien, 2017). Conclusively, more epidemiological and experimental studies are needed to determine the association between arsenic exposure and risk of diabetes. A case–control study on diabetic patients in Northern Chile also indicated the association between high exposures to arsenic in drinking water and increased risks of T2D (Castriota et al., 2018). A cohort study indicated that the Canadian women with urine levels of DMA 3.52 μg As/L have an increased risk of GDM (a odds ratio = 3.86; 95% confidence interval: 1.18, 12.57; Ashley‐Martin et al., 2018). Table 4 shows the epidemiological research on the associations between arsenic exposure and risk of diabetes.

Table 4.

Epidemiological studies on the association between arsenic exposure and risk of diabetes

Author, year	Study design	Study population	Diabetes diagnosis	Sample	As mean level	Outcome
Navas‐Acien et al. (2009)	Cross‐sectional	1,279 participants (160 with diabetes) aged 20 years or older exposed to As in drinking water	OGTT	Urine (µg/L)	30.4	Significant association between urine As and diabetes
Islam et al. (2012)	Cross‐sectional	1,004 adults living in Bangladesh who exposed to As in drinking water	FPG or self reported disease			Significant association between As exposure and diabetes
Chen et al. (2010)	Cohort	11,319 subjects who participated in the Health Effects of As Longitudinal Study in Araihazar, Bangladesh	Glucosuria test and glycosylated hemoglobin	Urine (µg/L)	45.9	No significant association between urine As and diabetes
Ettinger et al. (2009)	Cross‐sectional	532 pregnant women living proximate to the Tar Creek Superfund Site	Blood glucose and GGT	Blood (μg/L)	1.7	Significant association between urine As and gestation diabetes
				Hair (ng/g)	27.4
Farzan et al. (2016)	Cohort	1,151 pregnant women who participated in in the Hampshire Birth Cohort Study.	GGT	Toenail (µg/g)	0.1	Significant association between toenail As and gestation diabetes
				Urine (μg/L)	5.9
Rhee et al. (2013)	Cross‐sectional	3,602 subjects who participated in the Korea National Health and Nutrition Examination Survey	FPG	Urine (µg/g creatinine)	117.7	Significant association between urine As and diabetes
			glycosylated hemoglobin
Huang et al. (2014)	Cross‐sectional	43‐male and 99‐female living in Cambodia who exposed to As in drinking water	FPG	Urine (μg/L)	68.88	No significant association between urine As and diabetes
			glycosylated hemoglobin
Mahram et al. (2013)	Cohort	6,769 male and 8,300 female living in As contaminated and noncontaminated area in Iran	FBG			Significant association between As exposure (20–30 µg/L in water) and diabetes
Baek et al. (2017)	Cross‐sectional	369 subjects who participated in the fourth Korea National Health and Nutrition Examination Survey	FBG	Urine (µg/g Cr)	136.89	Significant association between urine As and diabetes in men
			glycosylated hemoglobin
Bräuner et al. (2014)	Cohort	57,053 subjects living in Denmark who were registered in the Danish health care system	FBG			Significant association between As exposure (long‐term exposure to low‐level As in drinking water) and diabetes
Gribble et al. (2012)	Cross‐sectional	3,925 subjects living in American Indian adults from Arizona, Oklahoma, and North and South Dakota	FBG	Urine (μg/L)	14.1	Significant association between urine As and diabetes in rural communities in the United States
			glycosylated hemoglobin
Sripaoraya et al. (2017)	Case–control	385 diabetic and nondiabetic residents of three villages of Ron Phibun subdistrict, Nakhon Si Thammarat Province, Thailand	Self‐reported			No significant association between As exposure in drinking water and diabetes
Castriota et al. (2018)	Case–control	1,053 over the age 25 living in Northern Chile	Self‐reported			Significant association between As exposure and diabetes
Ashley‐Martin et al. (2018)	Cohort study	1,243 Canadian women over the age of 18, with singleton, live births, no pre‐existing diabetes, and complete urinary arsenic and glucose testing data	FBG	Urine (μgAs/L)	3.52 μg	Significant association between urine As metabolites and gestation diabetes

Note. As: arsenic; FBG: fasting blood glucose; FPG: fasting plasma glucose; GGT: gamma‐glutamyl transferase; OGTT: oral glucose tolerance test.

8. CONCLUSION

As many individuals are exposed to arsenic contaminated water worldwide, it is important to precisely unravel the role of arsenic on the risk of obesity and diabetes. There are several evidence on the effect of arsenic on glucose metabolism disruption and insulin resistance. Arsenic activates the metabolic pathways in skeletal muscles and liver in favor of elevated glucose generation. Hyperglycemia may occur due to arsenic exposure as a result of glycogenolysis and gluconeogenesis stimulation. In the early stages, glucose is used to generate energy to meet the stress situation and serve as the major source of energy. In addition, arsenic induced lipid diseases and elevated FFAs synthesis. Dyslipidemia has a main role during the early stages of insulin resistance. Arsenic induces proinflammatory cytokines which activate JNK and IKKþ pathways. Cytokines‐induced IKKþ activation cause NF‐nB translocation and the elevated expression of proinflammatory cytokines that can modulate the insulin signaling. Arsenic causes oxidative stress resulting in disturbance in the metabolism of carbohydrates. In addition, disruption in the mitochondrial function induces anaerobic metabolism and ROS production. An increased ROS triggers the activation of serine kinase that involved in impairment of insulin signaling. Accordingly, the obesogen and diabetogen hypothesis are strongly supported by the publishing literature. Weight gain impairs the functions of many organs, particularly adipose tissue itself. To date, the majority of in‐vitro and in‐vivo effects of arsenic on adipose tissue suggest that arsenic can negatively affect adipocytes/WAT metabolism. Whereas arsenic diminishes adipogenesis in preadipocytes, it can concomitantly increase adipocyte size or WAT weight. Arsenic increases basal lipolysis in‐vitro as well as lipogenesis during fasting. Arsenic also reduces basal glucose uptake and ISGU, and downregulates adiponectin mRNA expression. Although there are evidence suggesting that arsenic regulates adipose tissue metabolism, the exact mechanisms, as for many endocrine‐disrupting chemicals, are still not fully understood. However, one suggested underlying mechanism is the primary promotion of weight gain by adipocyte hypertrophy. There are also evidence indicating arsenic interactions with other environmental factors such as high‐fat diets or folate‐related nutrients. Among other factors regulating arsenic effects are obesogenic or disease‐related factors positively or negatively affecting the arsenic modulating impacts. In addition, arsenic turn over itself also plays an important role in its modulating effects. Arsenic metabolism is influenced both by genetic polymorphisms in genes involved in arsenic metabolism such as arsenic (3 oxidation state) methyl‐transferase (known as AS3MT), as well as environmental factors such as BMI. However, there is still controversy regarding the effects of BMI. Overall, the current review indicated a close association between arsenic exposure and obesity and diabetes. However, available evidence are currently insufficient to conclude that low–moderate doses arsenic are also associated with diabetes development necessitating more studies in populations with low–moderate exposure to arsenic. Some gaps in this field include the lack of prospective studies, the use of death certificates, self‐reported diagnosis for diabetes, and using ecological methods for arsenic exposure assessment. Because of these limitations, there are a few evidence regarding the effects of high‐dose arsenic exposure on the risk of obesity and diabetes. Therefore, it is recommended to conduct cohort, case–control, and follow‐up cross‐sectional studies (such as NHANES) to explore the role of different arsenic doses on diabetes. The HEALS study in Bangladesh (Chen et al., 2010) exploring the impacts of moderate‐to‐high exposure to arsenic showed conflicting results compared with other similar studies indicating a multifactorial phenomenon in this condition. Research on interactions between arsenic exposure and factors such as BMI, diet, physical activity levels, coexposures to other metals juxtaposed to arsenic, including metals that occur with arsenic, exposure duration, and timing of exposure (i.e., the importance of early life or prenatal exposures) may help address these issues. In addition, future research should include genetic studies, as well as environmental interactions regarding arsenic metabolism and diabetes susceptibility.

CONFLICTS OF INTEREST

The authors declare that there are no conflicts of interest regarding the publication of this study.