The (FTO) gene polymorphism is associated with metabolic syndrome risk in Egyptian females: a case- control study

Metabolic syndrome (MetS) is considered a major risk factor for cardiovascular diseases and stroke [1] in both developing and developed countries [2, 3] . The central obesity and type 2 diabetes mellitus (T2DM) are considered the major endophenotypes of MetS which are inheritable and genetically based [4‐6]. Interestingly, Egypt had the highest mean female BMI in developing countries and even exceeded USA [7]. In Egypt, 22% of men while 48% of women were obese according to 2010 WHO estimate. The Middle East and North Africa Region are considered to have the highest prevalence of diabetes at 10.9% [8]. The prevalence of MetS in Egyptian females has been reached about 45.8% [9]. The FTO (fat mass- and obesity-associated) gene is a good candidate gene for T2DM and MetS, since its possible link to obesity with genome wide association study based evidence [10]. In addition, human FTO gene (16q12.2) shows ubiquitous expression in adult and fetal tissues with the most significant high expression in hypothalamus [10] linking the FTO protein function with control of energy state [11]. FTO protein is one of the non-heme dioxygenase superfamily, which demethylate nucleic acids in a 2- oxoglutarate-dependent manner [11]. Despite several studies showed association of FTO gene polymorphism with obesity in different populations mainly of European descent [12‐16], this came in contrast with other studies in which the FTO polymorphisms do not associate with BMI or obesity risk in consistent manner as in many populations with different ethnicity [16‐19]. Moreover, The association of FTO with T2DM is still inconclusive [20]. The first intronic FTO variant rs9939609 was selected specifically in this study as it depicted the highest genotyping success rate among other obesity risk SNPs and has strong genome wide association evidence with obesity [10]. Furthermore, it is highly possible that this variant, even it is not the true functional variant, is in linkage disequilibrium with true functional ones either within FTO gene or nearby genes [16].

As shown in ‘Table 1’, there was significant difference in anthropometric and biochemical measurements between the study groups. The BMI, waist circumference, plasma glucose, HOMA-IR, TAG, TC and LDL-C were significantly higher, while HDL-C level was lower in the group with MetS vs control group.

In the whole population the distribution of genotype for rs 9,939,609 variant were 34% TT, 52.8% TA and 13.2% AA. In the MetS group, the genotype distribution was 26.1% TT, 54.4% TA and 19.5% AA vs 41% TT, 51.4% TA and 7.6% AA in the control group, as shown in ‘Table 2’. This came in agreement with Hardy–Weinberg equilibrium (P > 0.05) in the whole population and in the individual study groups. The overall minor allelic frequency (MAF) of the risk allele A of FTO gene rs9939609 was (0.39).

As depicted in ‘Table 2’, the dominant genetic model showed a 1.97-fold increase in the odds of MetS (95% C.I = 1.07–3.6, p = 0.029). Under recessive genetic mode, the AA genotype showed a 2.95-fold increase in the odds of MetS (95% C.I = 1.22–7.22, p = 0.017).

There was no significant difference in both anthropometric and biochemical measurements of the whole population among the different genotypes at all models (additive, dominant, and recessive) except for ALT and HDL-C levels. Regarding ALT, it was significantly higher in AA carriers (P = 0.02), as shown in ‘Table 3’. The association remained significant after correction of BMI and serum TAG. However, this association was abolished after applying the conservative Bonferroni adjustment for multiple testing. Regarding HDL-C levels, the AA genotype showed significantly lower levels (P = 0.009) and interestingly it remained significant after conservative Bonferroni adjustment.

This is the first study to investigate the allele frequency and genotype distribution of FTO gene rs9939609 and its association with MetS including its endophenotypes of serum lipids in Egyptian female population. The overall MAF (A) in our study was (0.39) which is very similar to some European population as that of adult population from Spain which was (0.39) [12] and slightly lower than that of Italian population which was (0.48) [13]. Regarding Asian ancestry, a remarkably different MAF appeared which was (0.2) in Japanese population [27], (0.19) in Pakistani females [28] and (0.12) in Chinese population [29]. Our allele frequency results were also higher than that of black African population which was (0.35) [30] and slightly lower than the Brazilian one which was (0.435) [31] .

Despite MAF similarity in the current study with some European population, the genotypes distribution in our study groups was unique. In addition, our study revealed a positive association between FTO variant and MetS for the first time in Egyptian Arab female population. Al-Attar et al. (2008) reported in non-Caucasian population an association between rs9939609 FTO SNP and MetS risk that was dependent on obesity and HDL-C components. Moreover, the association was more prominent in men [32]. Our study declared association with MetS in cohort of Arab Egyptian Caucasian females that may be dependent partially on the association with lowered HDL-C levels. In agreement with our study, another positive association was observed in Han Chinese between the FTO SNP rs8047395 and MetS but with no significant associations with any endophenotypes of MetS [33]. Another important strength point of our study is that the cases and controls were carefully matched in other non-genetic risk factors of the MetS like the dietary habits and exercise levels which sheds more light on the impact of genetic factors. Regarding the association of FTO variation with lipid profile levels, our study showed non-significant association except for HDL-C levels. This came in contrast with other studies which showed non-significant association with all lipid profile [29, 34] or showed association with elevated triglycerides levels [35, 36]. However, it came in agreement with other studies depicted reduced HDL-c levels [36‐38]. Moreover, the associations in other studies were more prominent in men [32, 39]. Subjects with AA genotype have higher BMI and WC compared to TA and TT carriers, however this difference didn’t reach statistical significance, may be due to relatively small sample size and different effect size in case of BMI. This is in agreement with three previous studies that didn’t detect association of rs9939609 FTO polymorphism with BMI and/or WC in a similar sample size [19, 35, 40]. In addition, there was association of FTO polymorphism with MetS in our study under both the dominant and recessive models unlike other studies like the one carried on Italian male and female population in which only the TA genotype showed association [41] and the one carried among Tunisian population [42]. Xu et al. (2014) showed significant association of FTO variations with some metabolic parameters under recessive model and even in higher sample size of our study [43]. We hypothesized that the association under recessive model was apparent here in our study may be due to association of (AA) subjects with lowered HDL-C levels compared to (TT) subjects. Those differences may be due to different ethnic group, different sample size, different study design or different population age.

The FTO association with MetS may be related to association with obesity [10], T2DM [44] and/or lipid abnormalities [32, 35]. The exact mechanism by which FTO protein is linked to obesity and T2DM is still completely unknown. Being highly expressed in the hypothalamus, FTO plays a key role in energy homeostasis [11]. Genetic variants of FTO were associated with increased energy intake [40, 45, 46]. Based on bioinformatics and in vitro studies FTO encodes for demethylase enzyme for DNA and RNA nucleotides especially RNA single strands [11, 47]. Changes in the DNA methylation state are significantly associated with many pathological diseases and abnormalities including obesity [48, 49]. For example, it is interesting that Indian children obesity was affected by vitamin B12 and folate maternal levels [50] indicating that methylation state plays a key role in risk contribution for obesity and related disorders [51]. This highlights that FTO link with obesity could be mediated by epigenetic changes in DNA methylation [52]. Moreover, the FTO plays a key role in the regulation of food intake through its function of sensing the levels of amino acids in the hypothalamus [53] which is known to affect orexigenic and anorexigenic pathways [54]. Very recently, it was shown that genetic variants within FTO gene are linked functionally to another distant gene called IRX3 which may link the association of FTO SNPs with both obesity and T2DM [55, 56]. Moreover, our study suggests for the first time an association of FTO polymorphism with high ALT levels that seems to be BMI independent, though the association was no longer statistically significant when corrected for multiple comparisons. Wu et al., 2010 hypothesized that the intronic genetic variants like rs9939609 affected the expression of FTO protein which may modulate adipogenesis in adipose tissue [57]. Interestingly nonalcoholic fatty liver disease (NAFLD) rat model in which the FTO overexpressed in liver showed increased oxidative stress and lipogenesis and was associated also with increased ALT levels [58]. This suggests an association of FTO genetic variation with ALT might be mediated through NAFLD, though participants with overt liver disease were excluded from our study. Another possibility is that FTO contributes directly to the regulation of hepatic metabolism [59]. Very recently, Guan et al. (2014) showed a significant association between ALT level and variants of some obesity-susceptible genes, but not FTO, either dependent or independent of BMI [60]. As a result this association with ALT needs to be validated in other populations with large sample size. We acknowledge the relative small sample size, however we are taking an advantage of the high frequency of the variant understudy in the Egyptian population, as reported to be 39% with unique distribution in the control and cases group which improves the power of study. Hopefully, our reported findings act as base for further future studies with larger sample size.

this study highlighted for the first time a possible link between the FTO homozygous variant (AA) and the risk for MetS in Arab Egyptian female population that may be mediated through HDL-C lowered levels. Moreover, our study suggests a novel BMI and TAG-independent association of FTO genetic variation with elevated ALT levels that needs to be validated in other large sample sizes. We believe that this study hopefully helps in clarifying the pathophysiology and etiology of metabolic syndrome. Moreover, it will aid in early predicting those with high risk of developing metabolic syndrome which will result in better personalized strategies of treatment, nutrition or prevention. In addition our study, by linking FTO polymorphism with metabolic syndrome highlights the possible link between the FTO polymorphism with cardiovascular disorders.