Article Text

Download PDFPDF

Aggravating effect of INSIG2 and FTO on overweight reduction in a one-year lifestyle intervention
  1. T Reinehr1,
  2. A Hinney2,
  3. A M Toschke3,
  4. J Hebebrand2
  1. 1
    Vestische Hospital for Children and Adolescents, University of Witten/Herdecke, Datteln, Germany
  2. 2
    Department of Child and Adolescent Psychiatry; University of Duisburg, Essen, Germany
  3. 3
    King’s College London, Division of Health and Social Care Research, London, UK
  1. Correspondence to Dr P D T Reinehr, Vestische Hospital for Children and Adolescents, University of Witten/Herdecke, Dr F Steiner Strasse 5, 45711 Datteln, Germany; T.Reinehr{at}


Objective: Obesity is considered a polygenic and multifactorial disorder and different single nucleotide polymorphisms (SNP) are involved. Studies concerning their impact on weight loss in lifestyle intervention are scarce.

Methods: The effect of two different SNP (INSIG2: rs7566605, FTO: rs9939609) was analysed on the change of weight status in a one-year lifestyle intervention among 280 overweight children (mean age 10.8 years, mean body mass index (BMI) 28.1 kg/m2).

Results: The children reduced their mean SDS-BMI by −0.28 (95% CI −0.32 to −0.23). Modelling the impact of different genotypes and their statistical interactions on SDS-BMI change adjusting for age, gender and baseline BMI or SDS-BMI, respectively, revealed that the combination of the CC genotype in INSIG2 and the AA genotype in FTO was significantly associated with the lowest degree of overweight reduction, but even with an increase in overweight (SDS-BMI change +0.51; 95% CI 0.22 to 0.79).

Conclusions: These findings provide some evidence that the effects of different genotypes aggravate each other concerning weight change.

View Full Text

Statistics from

Several polymorphisms are associated with weight status, whereas their effect on change of weight status in a lifestyle intervention and especially their potential additive effects are less clear.1 One previous study in adults showed that double homozygosity in polymorphisms in the “fat mass and obesity associated” gene (FTO) and insulin-induced gene 2 (INSIG2) as well as the combination of FTO/INSIG2 homozygosity/heterozygosity leads to an increased body mass index (BMI).2 Another study showed no interaction between INSIG2 and FTO.3 As studies in children and analyses in weight loss interventions are missing, we analysed the potential additive effect of the single nucleotide polymorphism (SNP) located approximately 10 kb upstream of INSIG2 (rs7566605) and the polymorphism located in intron 1 of FTO (rs9939609) on the weight change of obese children in a lifestyle intervention.

Patients and methods

The local ethics committees of the Universities of Witten/Herdecke and Duisburg-Essen approved this study. Written informed consent was obtained from all subjects and their parents.

A total of 280 unrelated overweight children (mean age 10.8 years (minimum 4.5, maximum 16.5), 45% male; mean BMI 28.1 kg/m2; 95% CI 27.6 to 28.7 kg/m2) participating consecutively in the one-year outpatient lifestyle intervention “Obeldicks” was screened concerning polymorphisms in FTO and INSIG2. Overweight was defined by a BMI greater than the 90th percentile for German children. We used the last mean square method to calculate SDS-BMI as a measure for the degree of overweight. The intervention “Obeldicks” is based on physical activity, nutrition education and behaviour therapy, and has been described in detail elsewhere.4

The SNP in INSIG2 were screened as described previously.5 FTO SNP rs9939609 was genotyped using amplification refractory mutation system PCR. For validity of the genotypes, allele determinations were rated independently by at least two experienced individuals. Discrepancies were resolved unambiguously either by reaching consensus or by re-typing. The genotype distribution did not depart from Hardy–Weinberg equilibrium for any analysed SNP.

Statistical analysis was performed using the statistical software package SAS version 9.1. We analysed the impact of these two genetic polymorphisms and the interaction between them in separate multivariable regression models with BMI and SDS-BMI as response variables and adjusting for age, sex and baseline BMI or SDS-BMI, respectively. We additionally calculated regression models considering the variables and also both genes as well as an interaction term between the two allele frequencies of INSIG2 and FTO in order to assess a synergistically acting of INSIG2 and FTO.


The mean change of BMI of the children was −0.71 kg/m2 (95% CI −1.11 to −0.47 kg/m2) at the end of the intervention, the mean change of SDS-BMI was −0.28 (95% CI −0.32 to −0.23). A total of 240 overweight children completed the lifestyle intervention, 191 (68%) reduced their SDS-BMI, 49 (18%) children increased their SDS-BMI, whereas 40 (14%) children dropped out of the intervention. These children had the same mean SDS-BMI at the last visit as compared with baseline.

Children homozygous for the C allele in INSIG2 increased their BMI and SDS-BMI in contrast to the children heterozygous or homozygous for the wild-type allele (table 1). A linear effect of the INSIG2 gene was significant for BMI change and SDS-BMI change in separate multivariable regression analyses with response variables BMI and SDS-BMI considering age, gender, the two gene polymorphisms and baseline BMI or SDS-BMI, respectively (table 2).

Table 1

Distribution of polymorphisms in relation to change of BMI and SDS-BMI among 240 overweight children completing the one-year lifestyle intervention

Table 2

Multivariable regression analysis with the dependent variables change of BMI (r2  =  0.22) or change of SDS-BMI (r2  =  0.16) and the dependent variables INSIG2, FTO and the combination of INSIG2 and FTO in 240 overweight children completing a one-year lifestyle intervention

Children homozygous for the A allele in FTO lost weight in terms of BMI in the intervention but to a lower extent when compared with children heterozygous or homozygous for the wild-type allele, whereas there was no significant difference with respect to SDS-BMI (table 1). After adjusting for age, gender, baseline BMI and the two gene polymorphisms, this association between additional A alleles in FTO and BMI change was not significant (table 2).

The children with the combination of the INSIG2 CC genotype and the FTO AA genotype showed the highest increase in BMI compared with the children with any other combination of the INSIG2 and FTO genotypes in the multivariable regression analyses adjusted for the other gene polymorphisms, age, gender and baseline BMI or SDS-BMI; the model was significant (table 2).

Additional analyses considered age, gender and baseline BMI or SDS-BMI, respectively, as well as INSIG2 allele frequency, FTO allele frequency and the interaction between both allele frequencies with the response variables BMI or SDS-BMI, respectively. The interaction term between the allele frequencies was significant (p = 0.02 for BMI as response and p = 0.003 for SDS as response variable).

The 40 drop-outs did not differ in respect of age, gender, or SDS-BMI from the children who completed the intervention. Frequencies of CC carriers in INSIG2 (3%) and AA carriers in FTO (34%) were similar in the drop-outs compared with the children completing the intervention (comparison of completers and drop-outs INSIG2 p = 0.233; FTO p = 0.161).


This is the first study examining the impact of two polymorphisms on weight loss in a lifestyle intervention programme for overweight children in one study sample. The achieved reduction of overweight and the success rate in our intervention programme were comparable with previous reports of interventions for obese children.4 We did not find an increased risk of dropping out of lifestyle interventions for the polymorphisms in INSIG2 and FTO. We were able to demonstrate that the INSIG2 CC genotype was associated with a significantly lower degree of overweight reduction in the lifestyle intervention according to our previous smaller study,5 whereas a trend towards lower weight loss was observed for AA carriers in FTO. Most importantly, the combination of the INSIG2 CC genotype and the FTO AA genotype was significantly associated with the lowest degree of overweight reduction, suggesting that the effects of INSIG2 and FTO aggravate each other. Therefore, this finding provides a hint towards gene–gene interactions for weight loss during a lifestyle intervention lasting one year.

Whereas the function of INSIG2 in weight regulation is unclear, studies of wild-type mice indicate that FTO messenger RNA is most abundant in the brain, particularly in hypothalamic nuclei governing energy balance, and that FTO messenger RNA levels in the arcuate nucleus are regulated by feeding and fasting.1

The strengths of this study are its longitudinal design and the analysis of two variants in one study sample. However, some potential limitations have to be kept in mind. Due to the relatively small sample size particularly of CC carriers in INSIG2, the findings have to be interpreted very cautiously. However, longitudinal studies in overweight children undergoing a lifestyle intervention are difficult to perform and data are currently scarce. Furthermore, our study only considered one polymorphism by gene; but especially for the FTO the a priori evidence is so compelling that the analysis of a single SNP can be viewed as sufficient for an initial analysis.1 Finally, it is also possible that an interaction between the SNP analysed and environmental factors such as diet may influence the change of weight status. Obesity occurs as a result of gene–gene and gene–environment interactions that vary across individuals. However, analysing dietary records in obese children is difficult to interpret due to underreporting.

In summary, children with the combination of the INSIG2 CC genotype and the FTO AA genotype showed the worst outcome in a lifestyle intervention, suggesting that the effects of INSIG2 and FTO aggravate each other. With regard to the validity of our findings and the moderate study sample, subsequent replication studies are necessary.


The authors would like to thank all probands and their families for their participation. The skilful technical assistance of J Andrä, Essen, was highly appreciated.


View Abstract


  • This study is registered at (NCT00435734).

  • Funding This work was supported by grants from the German Ministry of Education and Research (Bundesministerium für Bildung und Forschung; National Genome Research Network, NGFN1, 2 and NGFNplus), the European Union (FP6 LSHMCT-2003-503041) and the Deutsche Forschungsgemeinschaft (DFG; HE 1446/4-1).

  • Competing interests None.

  • Ethics approval The local ethics committees of the Universities of Witten/Herdecke and Duisburg-Essen approved this study.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.