Metabolic-related diseases and chronic conditions, such as obesity, diabetes, dyslipidemia, hyperuricemia and hyperhomocysteinemia, can occur in people of all ages and confer a very high risk of developing life-threatening cardiovascular and cerebrovascular diseases (1). Metabolic-related diseases are a group of multifactorial diseases with a number of risk factors including genetic susceptibility, sedentary lifestyle, high-fat diet, old age, cigarette smoking, and alcohol drinking. Genetic susceptibility may be the most important risk factor for metabolic-related diseases (2, 3). Researchers around the world have paid a great effort to investigate the relations of genetic variants with metabolic-related diseases in the past few decades, but the findings were inconsistent and inconclusive due to various reasons such as small sample size and differences in race, dietary habit, health condition, and social and natural environments across the study populations (4–7).

Fat mass and obesity-associated gene (FTO) encodes a N6-methyladenosine (m6A) demethylase which belongs to the superfamily of 2-oxoglutarate- and Fe(II)-dependent dioxygenases. The main function of FTO is to remove m6A from specific RNA species, such as mRNAs, thereby affecting the stability of RNAs, translation efficiency of mRNAs or binding activity of m6A readers (8, 9). FTO mostly targets the mRNAs that are closely related to glucolipid metabolism and energy homoeostasis, such as sterol regulatory element binding protein 1c (SREBP1c) (10–12), carbohydrate responsive element binding protein (ChREBP) (12), peroxisome proliferator-activated receptor alpha (PPARα) (13), PPARγ (14), PPARγ coactivator 1α (PGC1α) (15), apolipoprotein E (APOE) (16), uncoupling protein 1 (UCP1) (17) and ghrelin (18). Based on the functions of the target mRNAs of FTO, it is logical to speculate that the genetic variations in FTO are involved in the occurrence and development of metabolic-related diseases. Human FTO is located on chromosome 16q12.2 and consists of 9 exons and 8 introns. The rs9939609 (g.53786615T>A) and rs17817449 (g.53779455T>G) variants are localized within intron 1 of FTO, and rs17817449 is positioned approximately 7.2 kb upstream of the rs9939609 polymorphism. The rs9939609 polymorphism is formed by a transversion from thymine (T) to adenine (A), while the rs17817449 polymorphism is formed from thymine (T) to guanine (G). These two loci have been reported to be associated to metabolic-related diseases such as obesity (19–23), dyslipidemia (22, 24), hypertension (22, 25–27), type 2 diabetes mellitus (T2DM) (28, 29) and cardiovascular disease (20, 27, 30). However, some other investigators have come to completely different and even conflicting conclusions, demonstrating that the minor alleles of these genetic variants were either not associated or negatively associated with metabolic-related diseases (31–33).

PPARδ (also known as PPARβ) is a transcription factor and an important member of the PPAR family of nuclear receptors, with the other two members PPARα and PPARγ. As a nuclear transcription factor, PPARδ mainly initiates transcription of the genes related to lipid metabolism (34–36), adaptive thermogenesis (37) and adipocyte formation (38). Human PPARδ gene (PPARD) is located on chromosome 6p21.3 and contains 8 exons and 7 introns. The rs2016520 and rs2267668 polymorphisms are positioned in the 5’ untranslated region (5’UTR) and intron 2 of PPARD, respectively. The rs2016520 variant is also known as +294T/C, -87T/C or +15T/C, with cytosine (C) as its minor allele. The rs2267668 polymorphism is localized 856 bp upstream of the rs2016520 variant. The two polymorphic loci have been proved to be related to metabolic-related diseases such as obesity (39) and dyslipidemia (40). However, several other studies did not support these findings (41, 42).

The relationships of FTO and PPARD variants with metabolic-related diseases deserve to be further explored since there were so many inconsistencies in the scientific publications. Moreover, the associations of the four polymorphisms with hyperuricemia and hyperhomocysteinemia have not yet been investigated to date. Here, a hospital-based study with the clinical diagnosis of metabolic-related diseases or chronic conditions including obesity, dyslipidemia, hypertension, hyperhomocysteinemia, hyperuricemia, T2DM and coronary artery disease (CAD) was conducted to systematically explore the relations of FTO rs9939609 and rs17817449, and PPARD rs2016520 and rs2267668 polymorphisms with metabolic-related diseases. The results from this study can provide an opportunity to unveil the interrelationships among variants in FTO and PPARD, metabolic disorders and CAD.

The present study demonstrates that FTO rs9939609 and rs17817449 polymorphisms are significantly and independently associated with the susceptibility to dyslipidemia and T2DM in a Chinese Han population. The A allele of rs9939609 and G allele of rs17817449 polymorphisms confer a higher risk of dyslipidemia and T2DM. In addition, carriers of the A allele of the rs9939609 polymorphism or G allele of the rs17817449 polymorphism have a higher risk of developing dyslipidemia in obese patients than those in non-obese individuals. To our knowledge, this is the first time to demonstrate that the relations between the rs9939609 and rs17817449 polymorphisms and dyslipidemia are modulated by obesity.

In line with our findings, a series of studies have reported significant correlations of FTO rs9939609 and rs17817449 polymorphisms with the risk of T2DM and dyslipidemia (24, 28, 29, 33, 46–51). Younus et al. conducted a case-control study to investigate the impact of the rs9939609 and rs17817449 polymorphisms on the development of T2DM in a sample of obese Iraqis, and found that the minor alleles of both variants are associated with an increased risk of T2DM (33). In addition, the significant relation between the rs9939609 and rs17817449 polymorphisms and T2DM risk was reported in Vietnamese (28), Indians (29), Egyptians (47) and Czechs (48) and Pakistanis (51). Sierra-Ruelas and teammates explored the association of the rs9939609 variant with serum lipids among Mexican adults, and found that the A allele carriers had significantly higher levels of TC and LDL-C (24). Jalili et al. observed that the A allele carriers of the rs9939609 polymorphism had a lower average level of HDL-C than noncarriers in a cohort of Iranian women (49). In a population of Romanian Children, researchers demonstrated that the G allele carriers of the rs17817449 variant had higher levels of TC and TG (50).

As the rs9939609 and rs17817449 polymorphic sites are localized within FTO, a gene with the strongest association with obesity discovered so far, a large number studies have reported a significant association between the rs9939609 and rs17817449 polymorphisms and obesity among various populations including East and South Asians (52–58), Latin Americans (23, 59–61), Europeans (62), Africans (63) and Middle Easterners (64). The A allele of the rs9939609 polymorphism was also found to be associated with an increased risk of obesity in several cohorts of children respectively from Romania (50), China (65), Portugal (66) and Chile (67). These relations were further strengthened with a genome-wide association study (68) and a couple of meta-analyses (69, 70) which consistently demonstrated that the A allele of rs9939609 and G allele of rs17817449 polymorphisms confer a higher risk of obesity in children and adults. Somehow, no significant associations between the rs9939609 and rs17817449 polymorphisms and obesity were found in our study population. One of the reasons could be that the present study population was not made up of general or healthy individuals, but those people who came to hospital to see doctors with chest pain symptoms and suspected CAD. In fact, more than half of these subjects had CAD or hypertension, and close to half of them had dyslipidemia. In addition, quite a few subjects in this study had T2DM, obesity, hyperhomocysteinemia, hyperuricemia, dyslipidemia and other diseases. Similarly, Sylwia et al. also reported that the rs17817449 polymorphism is not related to obesity in a population of patients with CAD, diabetes mellitus, hypertension and dyslipidemia (27).

The results of this study indicated that PPARD rs2016520 and rs2267668 polymorphisms are not associated with any of the metabolic-related diseases including obesity, dyslipidemia, hyperhomocysteinemia, hyperuricemia, hypertension, T2DM and CAD. In agreement with our findings, several previous studies also demonstrated no associations between the genotypes of PPARD rs2016520 or rs2267668 polymorphism and the risk of obesity (41), dyslipidemia (41, 71, 72), T2DM (41, 73) or CAD (42). However, Jguirim-Souissi et al. (72) concluded that the rs2016520 polymorphism is an independent risk factor of CAD, although it was not found to be associated with plasma lipid levels either in CAD patients or in control subjects. Chehaibi and colleagues (71) obtained a similar finding that the C allele frequency of PPARD rs2016520 polymorphism was higher in stroke patients than in controls, but plasma lipid levels did not differ significantly between the genotypes. Two studies demonstrated the rs2016520 polymorphism is associated with both CAD and blood lipids (74, 75). Maculewicz et al. (39) found that the rs2267668 polymorphism is significantly associated with BMI. As mentioned above, there were numerous inconsistent reports in the scientific community regarding these two PPARD polymorphic sites and metabolic-related diseases. The main reason for these inconsistencies may be that PPARδ as well as its gene polymorphisms just have a weak impact on the metabolic-related diseases, so their relations are easily affected by environmental and other factors. Among the three members of PPAR family, PPARγ is most closely related to metabolic-related diseases, PPARα being the next, and both of them have been served as drug targets for clinical use (3).

The mechanisms underlying the correlations between FTO rs9939609 and rs17817449 polymorphisms and T2DM as well as dyslipidemia have not been fully elucidated. The first potential explanation could be that the rs9939609 and rs17817449 polymorphisms have led to aberrant expression of FTO, which in turn disturbs the methylation status of FTO-targeted mRNAs, resulting in abnormalities in glucose and lipid metabolism. Indeed, Villalobos-Comparán et al. observed that the A allele carriers of the rs9939609 variant expressed significantly more FTO mRNA than those with the TT genotype in biopsies of subcutaneous fat tissue of Mexican women (61, 76). Karra et al. (18) and Berulava et al. (77) had similar finding that the A allele carriers of the rs9939609 polymorphism had significantly more abundant FTO transcripts in skin biopsies and/or blood cells than those with the TT genotype. The target mRNAs of FTO, such as SREBP1c (10–12), ChREBP (12), PPARα (13), PPARγ (14) and PGC1α (15), are all well-known to be extensively involved in glucose and lipid metabolism. The second explanation may involve energy intake. Ghrelin is a stomach hormone that has been proved to stimulate appetite and increase energy intake (78). Karra et al. found that subjects with the AA genotype of the rs9939609 polymorphism had elevated FTO transcripts, decreased ghrelin mRNA m6A methylation, and increased ghrelin levels as compared to the TT homozygotes (18). Several other studies have also demonstrated that FTO polymorphisms are highly associated with appetite, satiety, and energy intake in children and adolescents (79–81). Therefore, the association between the rs9939609 polymorphism and T2DM or dyslipidemia may be mediated by excessive energy intake. The third explanation may be that the rs9939609 and rs17817449 polymorphisms work in concert with other genetic variants to affect the susceptibility of obesity, diabetes and hypertension (82–84). For instance, Sirtuin 1 (SIRT1), a histone deacetylase and an anti-aging gene, is involved in the regulation of PPAR with relevance to lipid metabolism in the adipose tissue and liver, and is associated with metabolic-related diseases, such as metabolic syndrome and cardiovascular disease (85, 86). FTO polymorphisms may interact with the genetic variants in SIRT1 and other genes, and jointly regulate glucose and lipid metabolism.

There are several limitations to this study. First, the research population was not from general subjects, but a special group of people who had symptoms of chest pain came to the hospital for treatment and underwent coronary angiography examination. This may lead to sample-selection bias. Second, the research subjects are mainly made up of Chinese Han people, and hence the findings of this study may not apply to other ethnic origins. Third, the research adopted a cross-sectional design. It is necessary to carry out a follow-up investigation to determine the exact roles of the polymorphisms in FTO and PPARD in metabolic-related diseases.

Minor alleles of FTO rs9939609 and rs17817449 polymorphisms confer a higher risk of T2DM and dyslipidemia, and the risk is further increased among obese individuals. Relations between FTO polymorphisms and T2DM as well as dyslipidemia are possibly mediated by abnormal glucolipid metabolism and increased energy intake.

The datasets presented in this article are not readily available because of ethical and privacy restrictions. Requests to access the datasets should be directed to the corresponding author/s.

The study protocol was reviewed and approved by the Ethics Committee of Chengdu University. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants or the participants’ legal guardians/next of kin.

YS, YZ and LC conceived of the study, participated in the design, analyzed the data, and drafted the manuscript. LC, JZ and CH made the diagnosis of metabolic-related diseases. YZ, HL, LX and YW carried out DNA extraction, genotyping and collection of clinical data. SY revised the manuscript critically for important intellectual content. All authors reviewed and approved the final version of the manuscript for submission.