Genetic Epidemiology of Type 2 Diabetes in Mexican Mestizos

There are currently about 415 million people with diabetes worldwide, a figure likely to increase to 642 million by 2040. In 2015, Mexico was the second Latin American country and sixth in the world in prevalence of this disorder with nearly 11.5 million of patients. Type 2 diabetes (T2D) is the main kind of diabetes and its etiology is complex with environmental and genetic factors involved. Indeed, polymorphisms in several genes have been associated with this disease worldwide. To estimate the genetic epidemiology of T2D in Mexican mestizos a systematic bibliographic search of published articles through PubMed, Scopus, Google Scholar, and Web of Science was conducted. Just case-control studies of candidate genes about T2D in Mexican mestizo inhabitants were included. Nineteen studies that met the inclusion criteria were found. In total, 68 polymorphisms of 41 genes were assessed; 26 of them were associated with T2D risk, which were located in ABCA1, ADRB3, CAPN10, CDC123/CAMK1D, CDKAL1, CDKN2A/2B, CRP, ELMO1, FTO, HHEX, IGF2BP2, IRS1, JAZF1, KCNQ1, LOC387761, LTA, NXPH1, SIRT1, SLC30A8, TCF7L2, and TNF-α genes. Overall, 21 of the 41 analyzed genes were associated with T2D in Mexican mestizos. Such a genetic heterogeneity compares with findings in other ethnic groups.


Introduction
Type 2 diabetes (T2D) is a metabolic disorder characterized by impaired glucose uptake in muscle and fat, altered glucose-induced insulin secretion, and increased hepatic glucose production, which lead to hyperglycemia. It is the most common type of diabetes and generally occurs in adults [1]. According to the International Diabetes Federation there are currently around 415 million people with diabetes worldwide, a figure likely to increase to 642 million by 2040 [2]. This disorder accounts for high morbidity and mortality due to complications like renal failure, blindness, amputations, cardiovascular disease, and cerebrovascular events [1]. In 2015 there were approximately 5.0 million deaths by diabetes worldwide [2]. With about 7.3 million patients in 2010 [3], our country was second in Latin America and tenth in the world in prevalence of this disorder [4]. Five years later, the number of diabetic patients was estimated to be 11.5 million and our country ranked sixth in the world [2]. In 2011 most frequent morbidities by T2D were renal failure (24.2%) and peripheral circulatory complications (17.3%), and the mortality rate was 70/100,000 inhabitants (http://fmdiabetes.org/wp-content/uploads/2014/11/diabe-tes2013INEGI.pdf). The complex etiology of T2D includes factors that influence the risk and evolution of the disease, such as ethnicity, poor alimentation, sedentary lifestyle, obesity, dyslipidemia, and family history [1,5]. Regarding genetics, worldwide researches have shown association of this disease with numerous allelic variants of nearly 80 candidate genes [6]. The aim of this study is to carry out a literature review about genetic researches conducted in Mexican mestizos for a better understanding of the genetic epidemiology of T2D in our population.

Methods
A systematic search was done through PubMed, Scopus, Google Scholar, and Web of Science for genetic studies conducted in Mexican mestizo inhabitants with T2D. Key words derived from the phrase "Genetic polymorphisms associated with Diabetes Mellitus type 2 in Mexico, Mexican patients and/or Mexican mestizo" were used. Related terms such as "variants", "alleles", and "SNP associated with diabetes, T2DM, or T2D" complemented our search. Just case-control studies of candidate genes performed in Mexican mestizos resident in the country were included. Researches conducted in Mexican native populations were excluded, as well as those done in patients with metabolic syndrome. In surveys that included both patients with metabolic syndrome and patients with T2D, only cases with T2D were registered. Although in the selected studies different models of genotype analyses were used (recessive, dominant, or codominant), solely comparisons between allele frequencies were considered in our review. In studies without described odds ratio (OR), unadjusted OR were estimated from the reported allele or genotype frequencies. Allele comparisons were performed by 2 × 2 contingency tables [Yates' correction chisquare test (http://vassarstats.net/odds2x2.html)] and genotypes were contrasted by chi-square test (https://ihg.gsf.de/ cgi-bin/hw/hwa2.pl). In both comparisons, OR were estimated using 95% confidence intervals (95% CI). A ≤ 0.05 defined a significant association.
Whenever a polymorphism was analyzed in different studies, data were combined and unadjusted OR for alleles were calculated using a 2 × 2 contingency table (Yates' correction chi-square test). However, studies with suspicion of overlapping of patients were not included in this analysis.

Discussion
This review about genetics of T2D in Mexican mestizo subjects shows that 26 polymorphisms distributed in 21 genes are associated with this disease, so T2D has a high heterogeneity in our population, the same as that in other ethnic groups. Therefore, in some individuals alleles of certain genes are involved, while in others subjects are implicated variants of different genes. A previous conclusion that T2D in Mexican mestizos is genetically homogeneous was based on an analysis of three genetic markers [26] and here appears untenable. Though the Mexican mestizo population has a European genetic ancestry near 30% [27], not all the alleles conferring diabetes risk in Europeans are associated with T2D in our population [8]. These variations could be related to genetic background, differences in clinical classifications, sample size, selection and analysis criteria, and environmental factors such as obesity, lifestyle, and diet. On the other hand, researches in several ethnic groups have shown association of T2D with genes not yet analyzed in Mexican population [5,6,[28][29][30][31][32]. It would be important to carry out the analysis of such genes to determine whether these variants are also associated with T2D in Mexican patients and increase the knowledge about the genetic epidemiology of this disorder in our country.
Regarding Mexican studies, an increased risk was detected when analysis was performed adjusting covariates. For instance, Cruz et al. observed an additive effect in the T2D risk when they considered variables such as age, education, sex, body mass index, and ancestry [10]. Gamboa-Meléndez et al. reported association with T2D for the polymorphisms rs7923837 (HHEX), rs4402960 (IGF2BP2), and rs2237892 (KCNQ1) only when ancestry was adjusted [8]. For the polymorphisms rs864745 (JAZF1) and rs757705 (NXPH1), the analysis stratified by ancestry did not show significant differences, whereas an association was observed in the comparison without such an adjustment. In addition, they found association for rs7903146 (TCF7L2) and rs7754840 = 0.009], respectively. Another study found a lower OR when the analysis was adjusted by sex, body mass index, and family history of T2D for three polymorphisms of IRS1 in a dominant model [19].
The reported association of rs3842570 (CAPN10) [14], rs909253 (LTA) [20], and rs1800629 (TNF-) [24] with T2D should be interpreted with caution given the small sample sizes and poor statistical power. With respect to the rs1345365 polymorphism (ELMO1), the authors reported a protector effect for the A allele [OR = 0.65 (0.55-0.78), p < 0.001] [16]. But in our analysis we took as reference the A allele, as it is the most common; thus, the G allele showed association with T2D [OR = 1.37 (1.02 to 1.84), = 0.035].
Since T2D is a complex disorder and several genes are implicated in its etiology and evolution, the identification  . [11]        Only Genotypes of T2D patients and healthy controls were used in our analysis. i Assessment derived from the sum of T2D patients (obese and nonobese). j The authors reported a protector effect for the A allele (OR = 0.65; p < 0.001), but in our estimation we took as reference the A allele, since it is the most common. k Significant analysis with ancestry correction was taken. l Association was only found in early-onset T2D (OR = 1.39; = 0.024). m Just the population of Guerrero was recorded due to possible overlapping of the patients from the Mexico City with [10]. n The G allele was assessed as risk by the authors; but in our analysis we took the A allele, the same as that in previous studies. of risk alleles could be useful, because if the involved genes and their function are known, it is more probable to achieve prevention, treatment, prognosis, and/or cure of the disease. Complications could also be prevented or treated better [29,33]. However, published studies demonstrate that genetic screening for the prediction of T2D in high risk subjects is currently of little value in clinical practice. Actually, genetic risks are difficult to calculate because several alleles could contribute to an additive effect to T2D susceptibility [34], not to mention the diverse environmental factors involved. Although some of these genes are implicated in the glucose and fat metabolism, -cell function, and sensitivity and secretion of insulin [29,35], how some of their variants increase the T2D risk remains to be elucidated [29]. Anyway, it is fundamental to analyze the genetic epidemiology of this disease in each population because of the underlying differences in genetic background and lifestyle among ethnic groups. So, it is possible that polymorphisms associated with T2D in some races do not show association in others. Genome-wide association studies will ultimately precise the genetic landscape.

Conflicts of Interest
The authors declare no conflicts of interest concerning the publication of this paper.

Authors' Contributions
Eiralí Guadalupe García-Chapa and Juan Pablo Meza-Espinoza contributed equally to this work. All the authors participated in drafting and review of this paper.