Impact of Variation in the FTO Gene on Whole Body Fat Distribution, Ectopic Fat, and Weight Loss
The first two authors contibuted equally to this work.
Polymorphisms in the fat mass- and obesity-associated (FTO) gene have been identified to be associated with obesity and diabetes in large genome-wide association studies. We hypothesized that variation in the FTO gene has an impact on whole body fat distribution and insulin sensitivity, and influences weight change during lifestyle intervention. To test this hypothesis, we genotyped 1,466 German subjects, with increased risk for type 2 diabetes, for single-nucleotide polymorphism rs8050136 in the FTO gene and estimated glucose tolerance and insulin sensitivity from an oral glucose tolerance test (OGTT). Distribution of fat depots was quantified using whole body magnetic resonance (MR) imaging and spectroscopy in 298 subjects. Two-hundred and four subjects participated in a lifestyle intervention program and were examined after a follow-up of 9 months. In the cross-sectional analysis, the A allele of rs8050136 in FTO was associated with a higher BMI, body fat, and lean body mass (all P < 0.001). There was a significant effect of variation in the FTO gene on subcutaneous fat (P ≤ 0.05) and a trend for liver fat content, nonvisceral adipose tissue, and visceral fat (all P ≤ 0.1). However, the single-nucleotide polymorphism was not associated with insulin sensitivity or secretion independent of BMI (all P > 0.05). During lifestyle intervention, there was also no influence of the FTO polymorphism on changes in body weight or fat distribution. In conclusion, despite an association with BMI and whole body fat distribution, variation in the FTO locus has no effect on the success of a lifestyle intervention program.
Recent genome-wide association studies aiming to identify genes responsible for diabetes mellitus described a strong link between variations in the FTO (fat mass- and obesity-associated) gene and body weight (1,2). The association with obesity is replicable in most populations (3,4,5), but the mechanism how variants in FTO lead to obesity is still not completely understood.
FTO was originally described in a mouse model and is involved in cell death programming (6). In heterozygote mice, the mutation results in fused toes. Second, the thymus is enlarged whereas other organs or body weight are similar to the wild type (7). With detection of the association of this gene locus with human obesity, the Gene Nomenclature Committee renamed the gene FTO. Recently, it has been shown that the FTO gene encodes a Fe(II)- and 2-oxoglutarate-dependent oxygenase putatively involved in DNA demethylation (8,9).
In humans, the FTO gene is located on the chromosome 16q12.2. Duplication of this region results in mental retardation, obesity, dysmorphic facies, and digital anomalies (10). It is known that FTO is expressed in the hypothalamic region (11) and may play a role in regulation of central body weight (12,13). In addition, carriers of the FTO allele which is associated with increased BMI exhibit a reduced basal fat cell lipolysis (14), suggesting that the FTO gene may also play a role in the adipose tissue metabolism.
In the initial studies, the higher diabetes incidence of the risk allele carriers depended only on the strong association with obesity (1). The influence of the FTO gene on body composition and different fat compartments is still largely unknown. We used a whole body imaging approach enabling quantification of body fat stores to test whether the FTO polymorphism influences body composition or ectopic lipid storage in the liver. Furthermore, variation in the FTO locus may also affect weight loss during lifestyle intervention (15,16). We, therefore, studied the effect of variations in FTO on weight loss and body composition in a lifestyle intervention program.
Subjects and Methods
We studied 1,466 nondiabetic subjects from the southern part of Germany who participated in the ongoing Tübingen Family Study for type 2 diabetes mellitus, which currently includes ∼2,000 individuals (17). Recruitment of the subjects was based on inclusion of subjects for whom DNA samples and complete data sets (glucose, insulin, and C-peptide measurements) were available. All participants underwent a physical examination, medical history, routine blood tests, and oral glucose tolerance test (OGTT). All subjects were genotyped for rs8050136 in the FTO gene. A subgroup (n = 298) participated in a magnetic resonance (MR) imaging and spectroscopy with determination of total body fat, visceral fat, nonvisceral fat, and subcutaneous fat.
We examined 204 subjects who participated in the Tübingen Lifestyle Intervention Program (TULIP). For inclusion into the study, the subjects had to match one of the following criteria—a family history of type 2 diabetes, or a BMI >27 kg/m2, or a previous diagnosis of impaired glucose tolerance, or gestational diabetes. The participants underwent an OGTT and MR imaging and spectroscopy at baseline and after 9 and 24 months of intervention. Because this is an ongoing project and most data are available for the baseline and 9-month visit, we analyzed these data in this study.
The TULIP consists of an exercise and dietary intervention for 2 years which was adopted from the intervention used in the Diabetes Prevention study (18,19). The participants aimed at a weight loss of at least 5%, a reduction of caloric intake from fat of <30% and an increase of fiber intake to at least 15 g/1,000 kcal. Furthermore, they were asked to reduce the intake of saturated fat to <10%. Individuals were asked to perform at least 3 h of moderate exercise per week, which was monitored using a heart rate monitor (Polar, Büttelborn, Germany).
All individuals had up to 10 sessions with a dietitian during the 9-month period. During each visit, participants presented a 3-day food diary and discussed the results with the dietitians. The study was approved by the local ethics committee. All subjects gave their written informed consent.
After an overnight fast, all subjects underwent a 75-g OGTT. Venous blood samples were obtained at 0, 30, 60, 90, and 120 min for determining plasma glucose, plasma insulin, and plasma C-peptide. Insulin sensitivity during OGTT was calculated as proposed by Matsuda and DeFronzo (20). Insulin secretion was measured using the ratio of area under the curve C-peptide/area under the curve glucose. The area under the curve of plasma glucose levels during OGTT was calculated as 0.5·(0.5·Glc0 + Glc30 + Glc60 + Glc90 + 0.5·Glc120). The area under the curve of plasma C-peptide levels during OGTT was calculated analogously.
Total body fat and lean body mass were measured by bioelectrical impedance (BIA-101; RJL Systems, Detroit, MI). An exact quantification of fat distribution was performed by MR imaging and MR spectroscopy. Subcutaneous fat, visceral fat, nonvisceral fat, and total body fat were expressed as kilogram of body weight. In addition, ectopic fat stores such as liver fat and intramyocellular lipids (musculus tibialis anterior) were measured in percent of the organ previously described (21,22).
Serum insulin was determined with a microparticle enzyme immunoassay (Abbott, Wiesbaden, Germany). Venous plasma glucose was measured using a bedside glucose analyzer (glucose oxidase method; Yellow Springs Instruments, Yellow Springs, CO). DNA was isolated from whole blood using a commercial DNA isolation kit (NucleoSpin; Macherey-Nagel, Düren, Germany). Genotyping of rs8050136 in the FTO gene was achieved using the TaqMan assay (Applied Biosystems, Foster City, CA) as previously described (12). This single-nucleotide polymorphism was manually picked as a representative for the linkage disequilibrium block encompassing also the previously reported single-nucleotide polymorphism rs9939609 (ref. 1).
All data are given as unadjusted mean ± s.e.m. Non-normally distributed parameters were log transformed to approximate normal curve of distribution before statistical analysis. Differences in anthropometrics between genotypes were tested using multivariate linear regression model. Differences between measurements at baseline and at follow-up were tested with a multivariate ANOVA. Metabolic data were adjusted for gender and age unless otherwise stated. Results with values of P ≤ 0.05 were considered statistically significant. The JMP (SAS Institute, Cary, NC) statistical software package was used.
In 1,466 subjects, the A allele of rs8050136 in the FTO gene was associated with a higher BMI (CC: 27.2 ± 0.3 kg/m2; CA: 29.0 ± 0.3 kg/m2; AA: 29.6 ± 0.5 kg/m2) (P < 0.0001) and a higher prevalence (24.3 ± 2% vs. 35.3 ± 2%, P < 0.0001) of obesity (BMI > 30 kg/m2). In the unadjusted data, there was a significant effect of variation in the FTO locus on percent body fat, lean body mass, fasting insulin, 120-min insulin, insulin sensitivity, and insulin secretion (all P ≤ 0.01). The effects of the polymorphism on insulin levels, insulin sensitivity, and insulin secretion lost significance after adjustment for gender, age, and BMI (all P > 0.05, Table 1).
Table 1. Associations of FTO rs8050136 with anthropometric and metabolic parameters (N = 1,466)
In the subgroup of subjects with the MR imaging (n = 298), we measured body composition including total body fat, nonvisceral adipose tissue, visceral adipose tissue, and ectopic fat stores (liver fat and intramyocellular lipids). In the unadjusted data, there was a significant effect of variation in the FTO gene on total fat mass and subcutaneous fat (all P ≤ 0.05) and a trend for liver fat content, nonvisceral adipose tissue, and visceral fat (all P ≤ 0.1). There was no association of the variation in the FTO locus with intramyocellular lipids (P = 0.57). Adjustment for gender and age altered the significance of the effect of the polymorphism only in fat compartments, which are highly influenced by gender (subcutaneous and visceral fat, Table 2).
Table 2. Associations of FTO rs8050136 with body fat composition (n = 298)
During the intervention, body weight, total fat, subcutaneous fat, visceral and nonvisceral fat, and liver fat were reduced (all P ≤ 0.0004). The genetic variation in the FTO gene did not affect the reduction in body weight (P = 0.95 for intervention vs. genotype effect (multivariate ANOVA)) or specific fat compartments or ectopic fat stores (waist-to-hip ratio, total body fat, nonvisceral fat, visceral fat, liver fat, all P > 0.2) (Table 3). After lifestyle intervention, the A-allele carriers still had a higher BMI (28.4 ± 0.4 kg/m2) than the homozygous CC-allele carriers (27.5 ± 0.5 kg/m2, P = 0.047). In addition, the variation in the FTO locus had no effect on improvement of insulin sensitivity during 9 months of lifestyle intervention.
Table 3. Associations of FTO rs8050136 with change in anthropometric and metabolic parameters
In the cross-sectional data of this study in a German population, the genetic variation in the FTO gene was associated with increased body weight, increased insulin secretion, and reduced insulin sensitivity. The effects on glucose metabolism lost statistical significance after adjusting for the independent effects of obesity. This is in accordance with the findings of the initial study by Frayling et al. (1). Therefore, our data may support the idea that the effect of genetic variation in the FTO gene on glucose metabolism reflects purely the effect on BMI. However, the possibility that FTO primarily influences glucose metabolism cannot be excluded in a cross-sectional study.
Previous studies showed a strong association between variation in the FTO gene and body weight (1,2). Our results confirm the high association of the risk allele with increased body weight and with higher BMI, total fat mass, and lean body mass. In addition, we used a whole body MR imaging protocol to study the effects of FTO on body composition and ectopic fat storage. Our data suggest that the A-allele carriers have increased subcutaneous, visceral, and nonvisceral fat mass as well as increased ectopic fat stores in the liver. Only intramyocellular lipids were unaffected. Therefore, most fat compartments are increased in the A-allele carriers, and the effect of the FTO gene on specific fat depots is not accentuated.
Because the A allele has a strong influence on body weight, we also analyzed the effect of genetic variation in FTO on weight change during lifestyle intervention. We expected less weight loss in carriers of the A allele. However, we could not observe any effect of the FTO polymorphism on body weight change or change in fat distribution in the longitudinal setting. In addition, the increase in insulin sensitivity during weight loss was not affected by the variation in the FTO locus. One explanation for this finding might be that the mechanism by which FTO increases body weight, for example by decreasing insulin sensitivity of the brain (12), needs a longer period to be effective. A relatively extensive lifestyle intervention designed to lose body weight might attenuate the risk conferred by the genetic background.
In conclusion, genetic variation in the FTO locus is associated with body fat mass, whole body fat distribution, and ectopic fat storage in the liver. Despite the strong association with body mass, the variation in the FTO locus has no influence on the changes in body weight after lifestyle intervention. There are no specific effects of FTO polymorphism on the glucose metabolism beyond obesity.
We thank all the research volunteers for their participation. We gratefully acknowledge the superb technical assistance of Anna Bury, Alke Guirguis, Heike Luz, Melanie Weisser, Roman Werner, Ellen Kollmar, and Barbara Horrer. The studies were supported by a grant from the Deutsche Forschungsgemeinschaft (KFO 114).
The authors declared no conflict of interest.