DOI 10.1515/jpem-2013-0179 J Pediatr Endocr Met 2013; 26(11-12): 1209 1213 Short communication Anke Hinney, Barbara Wolters, Carolin Pütter, Harald Grallert, Thomas Illig, Johannes Hebebrand and Thomas Reinehr* No impact of obesity susceptibility loci on weight regain after a lifestyle intervention in overweight children Abstract Objective: An obesity risk allele at the NEGR1 locus was shown to be associated with weight regain after a lifestyle intervention in obese adults. Independent confirmation and studies in children are lacking. Therefore, we analyzed the impact of this and 11 additional obesity susceptibility loci on weight regain after a lifestyle intervention in overweight children. Design and Methods: We longitudinally analyzed the changes in weight status as body mass index standard deviation score (BMI-SDS) in 282 overweight children (10.6 ± 2.5 years, 47% male, BMI 27.1 ± 3.9 kg/m 2 ) both at the end of a 1-year lifestyle intervention and at 1 year after the end of intervention. We genotyped obesity risk single nucleotide polymorphisms (SNPs) derived from genome-wide association studies in or in proximity to the following genes: NEGR1, TNKS, SDCCAG8, FTO, MC4R, TMEM18, PTER, MTCH2, SH2B1, MAF, NPC1, and KCTD15. Results: The children reduced their BMI-SDS ( 0.28 ± 0.35; p < 0.001) during intervention and increased their BMI-SDS between the end of intervention and 1 year later (+ ± 0.36; p = 0.027). None of the SNPs including NEGR1 was related significantly to weight regain. Conclusions: We found no evidence for effects of any of the GWAS-based obesity marker alleles on weight regain in the course of 1 year after an intervention. Keywords: children; lifestyle intervention; obesity; SNPs; weight regain. *Corresponding author: Prof. Dr. Thomas Reinehr, Department of Pediatric Endocrinology, Diabetes and Nutrition Medicine, Vestische Hospital for Children and Children Datteln, University of Witten/Herdecke, Dr. F. Steiner Str. 5, D-45711 Datteln, Germany, Phone: +49 2363975229, Fax: +49 2363975218, E-mail: T.Reinehr@kinderklinik-datteln.de Anke Hinney and Johannes Hebebrand: Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy, Universitätsklinikum Essen, University of Duisburg-Essen, Essen, Germany Barbara Wolters: Department of Pediatric Endocrinology, Diabetes and Nutrition Medicine, Vestische Hospital for Children and Children Datteln, University of Witten/Herdecke, Witten, Germany Carolin Pütter: Institute for Medical Informatics, Biometry and Epidemiology (IMIBE), University of Duisburg-Essen, Essen, Germany Harald Grallert and Thomas Illig: Research Unit of Molecular Epidemiology, Helmholtz Zentrum München, German Research Center for Environmental Health, Neuherberg, Germany Thomas Illig: Hannover Unified Biobank, Hannover Medical School, Hannover, Germany Introduction Obesity is highly heritable, but its underlying genetic factors remain largely elusive. The largest meta-analysis of genome-wide association studies (GWAS) revealed 32 loci associated with increased body mass index (BMI) (1). Previous studies suggested that some obesity susceptibility loci are related to weight loss during lifestyle intervention: poly morphisms near the fat mass and obesity-associated gene (FTO) or the serologically defined colon cancer antigen 8 gene (SDCCAG8) were associated with weight loss in lifestyle interventions for obese children or adults (2 7). Studies concerning the impact of obesity susceptibility loci on weight maintenance after the end of lifestyle interventions are scarce. We have previously shown that mutations in the melanocortin 4 receptor gene (MC4R) were associated with weight regain after lifestyle intervention (8). Delahanty et al. (7) reported that the obesity risk allele at a single nucleotide polymorphism (SNP) in the obesity gene neuronal growth regulator 1 (NEGR1) was associated with weight regain after lifestyle intervention in adults (7). However, confirmatory studies as well as
1210 Hinney et al.: Genes and weight regain analyses in children are lacking. Therefore, we analyzed the associations between 12 obesity risk alleles (including NEGR1, SDCCAG8, and FTO) and weight regain after a lifestyle intervention for overweight children. Examining children seems preferable since genetic causes of obesity are likely manifested early in life (9). A further advantage of studying this age group is that there is no potential confusion with other diseases or with medications. The primary aim of this study was to analyze whether the obesity risk alleles predispose to weight regain 1 year after a lifestyle intervention. Methods and procedures 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 282 unrelated children (mean age 10.6 ± 2.5 years, 47% male, mean BMI 27.1 ± 3.9 kg/m 2, mean BMI-SDS 2.35 ± 0.46, 14% overweight, 86% obese) participated in the 1-year outpatient lifestyle intervention Obeldicks and were re-evaluated 1 year after the end of the intervention. The study population is a subgroup of a recent study demonstrating an impact of the SNP related to SDCCAG8 on weight loss during a lifestyle intervention (3). Measurements Weight status was determined as BMI at baseline, at the end of the 1-year intervention, and again 1 year later (i.e., 2 years after baseline). Height was measured to the nearest centimeter. Weight was measured in underwear to the nearest 0.1 kg using a calibrated balance scale. The degree of overweight was quantified using Cole s least mean square method, which normalized the BMI skewed distribution and expressed BMI as a standard deviation score (BMI-SDS) (10). Overweight was defined by a BMI > 90th percentile for German children and obesity by a BMI > 97th percentile (11). All children were also overweight according to the definition of the International Task Force of childhood obesity (12). SNP genotyping for rs10789336 (NEGR1), rs17150703 (TNKS), rs10926984 (SDCCAG8), rs15589902 (FTO), rs17700144 (MC4R), rs11127485 (TMEM18), rs10508503 (PTER), rs10838738 (MTCH2), rs7498665 (SH2B1), rs1424233 (MAF), rs1805081 (NPC1), and rs11084753 (KCTD15) was performed by the matrix-assisted laser desorption/ionization-time of flight mass spectrometry-based iplex Gold assay (Sequenom, San Diego, CA, USA) at the Helmholtz Zentrum München. There was no evidence for a strong departure from Hardy-Weinberg equilibrium for any of the analyzed SNPs (smallest exact two-sided p = ). Intervention The 1-year lifestyle intervention Obeldicks has been described in detail elsewhere (13, 14): Briefly, this intervention is based on physical exercise, nutrition education, and behavior therapy including the individual psychological care of the child and its family. The exercise therapy consists of sports, instructions on physical exercise as part of all-day life, and on reduction of the amount of time spent watching television. The nutritional course is based on the prevention concept of the optimized mixed diet which is both fat and sugar reduced. Statistics Statistical analyses were performed using R (version 2.15.1) and plink (http://cran.r-project.org/). We used standard methods for descriptive statistics (such as means ± SD). Changes in BMI-SDS were compared by the Wilcoxon test. We used an additive genetic model for the impact of SNPs on BMI-SDS change. Multiple linear regression analyses with BMI-SDS changes as dependent variables and age, sex, baseline BMI-SDS, and the respective SNPs as independent variables were calculated. All reported p values are two-sided and nominal. Corrections for multiple testing of 12 SNPs were performed by Bonferroni adjustment. Assuming an effect of NEGR1 on weight regain after lifestyle intervention similar to that reported for MC4R mutations (8), we had a power of > 0.99 to detect differences in weight changes after lifestyle intervention for NEGR1 risk alleles in a two-sided comparison with an α error at based on the frequencies of the risk alleles at NEGR1 in our study. Since no studies of NEGR1 and weight regain after lifestyle intervention exist in childhood, we used data concerning MC4R mutations and weight regain after lifestyle intervention which are available in childhood (8). Using BMI and weight changes as in the study of Delahanty et al. (7) is not adequate in childhood, since BMI and weight are age and gender dependent (12). Therefore, BMI-SDS calculations are necessary to analyze weight changes (10). Results The genotype distributions for the GWAS-derived obesity risk alleles for the analyzed SNPs are shown in Table 1. The 282 overweight children reduced their weight significantly (mean change of BMI-SDS 0.28 ± 0.34; p < 0.001) during the 1-year lifestyle intervention. The children increased their BMI-SDS significantly between end of intervention and 1 year later (+ ± 0.34 BMI-SDS; p = 0.005). Compared to baseline, BMI-SDS at the 2-year follow-up was still significantly reduced ( 0.24 ± 0.45; p < 0.001). The obesity risk alleles at the SNP within the gene SDCCAG8 were associated with a lower degree of BMI-SDS reduction during the lifestyle intervention (see Table 2). The amount of weight reduction differed according to the obesity risk genotype at SDCCAG8: Homozygotes for the obesity risk allele lost less weight than carriers of the other genotypes (homozygous obesity risk allele carriers: 0.23 ± 0.32 BMI-SDS; heterozygote obesity risk allele
Hinney et al.: Genes and weight regain 1211 Table 1 Genotype distributions of obesity risk (and non-risk) alleles in 282 overweight children undergoing a 1-year lifestyle intervention. Nearest gene NEGR1 SDCCAG8 TMEM18 TNKS-MSRA PTER MTCH2 SH2B1 FTO MAF NPC1 MC4R KCTD15 Homozygotes 123 (43.6%) 8 (2.8%) 5 (1.8%) 225 (80.1%) 5 (1.8%) 128 (45.4%) 81 (28.9%) 73 (25.9%) 69 (24.6%) 86 (30.6%) 142 (50.5%) 137 (48.6%) for the obesity non-risk allele Heterozygotes 120 (42.3%) 76 (27.0%) 75 (26.6%) 50 (17.8%) 37 (13.1%) 116 (41.1%) 137 (48.9%) 140 (49.6%) 135 (48.0%) 147 (52.3%) 115 (40.9%) 115 (40.8%) Homozygotes for the obesity risk allele 39 (13.8%) 198 (70.2%) 202 (71.6%) 6 (2.1%) 240 (85.1%) 38 (13.5%) 62 (22.1%) 69 (24.5%) 77 (27.4%) 48 (17.1%) 24 (8.5%) 30 (10.6%) NEGR1, Neuronal growth regulator 1 gene; TNKS, tankyrase gene; SDCCAG8, serologically defined colon cancer antigen 8 gene; FTO, fat mass and obesity-associated gene; MC4R, melanocortin 4 receptor gene; TMEM18, transmembrane protein 18 gene; PTER, phosphotriesterase related gene; MTCH2, mitochondrial carrier 2 gene; SH2B1, SH2B adaptor protein 1 gene; MAF, v-maf musculoaponeurotic fibrosarcoma oncogene homolog (avian) gene; NPC1, Niemann-Pick disease, type C1, gene; KCTD15, potassium channel tetramerization domain containing 15 genes. Table 2 Effect of 12 obesity risk alleles on body weight loss (as ΔBMI-SDS) after a 1-year lifestyle intervention in 282 overweight children. Nearest gene NEGR1 SDCCAG8 TMEM18 TNKS-MSRA PTER MTCH2 SH2B1 FTO MAF NPC1 MC4R KCTD15 ΔBMI-SDS, 0.00 0.16 0.06 0.01 0.02 0.02 β-estimator 95% CI to 0.06 0.24 to 0.09 to 0.10 0.13 to 0.16 to p-value 0.939 7.90 10 6 0.522 0.314 0.192 0.376 0.754 0.563 0.578 0.394 0.179 0.326 0.06 to 0.10 0.09 to For each SNP effect size (β estimators in units of ΔBMI-SDS), their 95% confidence interval (CI), and the p values are listed. The β estimators were derived from linear regression models [additive model of inheritance adjustment for age (linear), sex, and baseline BMI-SDS]. Results that were significant after correction for multiple testing are italicized. Positive values of the effect sizes for the changes (Δ) indicate a reduction of BMI-SDS.
1212 Hinney et al.: Genes and weight regain carriers: 0.38 ± 0.33 BMI-SDS; homozygotes carriers of the other allele: 0.66 ± 0.50 BMI-SDS). None of the obesity risk alleles including the SNP at NEGR1 was significantly associated with weight regain 1 year after the end of a 1-year lifestyle intervention (see Table 3). Discussion This is the first study in childhood analyzing the impact of different obesity susceptibility loci on weight regain after a lifestyle intervention. We did not find significant associations in any of the analyzed GWAS-based obesity SNPs with weight regain 1 year after the lifestyle intervention. One might argue that the study population was too small to detect significant differences. However, the study sample was large enough to detect a significant impact of the SNP in SDCCAG8 on weight loss during lifestyle intervention in obese children according to our previous much larger study (3). Our findings are in contrast to a study in adults demonstrating an impact of a SNP in NEGR1 on weight regain (7). Therefore, the impact of this SNP on weight regain seems to differ between adults and children. Additionally, the study of Delahanty et al. (7) was performed in adults with disturbances in glucose metabolism, which may also explain the difference to our study. It is of note that Delahanty et al. (7) analyzed SNP rs2815752 and not SNP rs10789336 which was used in our study. However, as linkage disequilibrium between these two SNPs is extremely high (r 2 = 0.962, D = 1.000) (15), the equivocal results cannot readily be explained by this difference. The strengths of this study are its longitudinal design and the analysis of different variants in one study sample. However, some limitations have to be kept in mind. Due to the moderate sample size, the findings have to be interpreted with caution. Our sample was too small to detect minimal or moderate influences of SNP genotypes on weight regain. Follow-up studies after lifestyle interventions in pediatric obesity are very difficult to perform, and genetic data with long-term follow-up are lacking. However, the study sample was large enough to detect a significant impact of the SNP in SDCCAG8 on weight loss during the lifestyle intervention in obese children. We had previously reported that effect in a much larger study group (3). Power analysis ruled out an effect of NEGR1 comparable to MC4R mutations on weight regain. However, much larger sample sizes would be required Table 3 Effect of 12 obesity risk alleles on body weight regain (as ΔBMI-SDS) 1 year after a 1-year lifestyle intervention (2 years after admission) in 282 overweight children. NEGR1 SDCCAG8 TMEM18 TNKS-MSRA PTER MTCH2 SH2B1 FTO MAF NPC1 MC4R KCTD15 Nearest gene ΔBMI-SDS, 0.01 0.00 0.06 0.02 0.01 0.01 0.02 β-estimator to 0.09 to 0.10 to 0.02 0.09 0.15 to 0.11 0.14 95% CI to 0.07 p-value 0.671 0.964 0.123 0.630 0.325 0.745 0.270 0.345 0.683 0.157 0.360 0.429 For each SNP effect size (β estimators in units of ΔBMI-SDS), their 95% CI and the p values are listed. The β estimators were derived from linear regression models [additive model of inheritance adjustment for age (linear), sex, and baseline BMI-SDS]. Positive values of the effect sizes for the changes (Δ) indicate a decrease in BMI-SDS.
Hinney et al.: Genes and weight regain 1213 to ultimately rule out smaller effects and to validate our observations. Finally, we have analyzed only a subgroup of obesity susceptibility loci on weight regain and other SNPs may be related to weight regain. For example, a SNP in PPARγ has been reported to be related to weight regain after lifestyle intervention in adults (7). However, this SNP was not related to BMI in the largest GWAS meta-analysis (1). In summary, we did not find evidence for an association of any of the analyzed obesity risk genotypes on weight regain 1 year after a 1-year lifestyle intervention (2 years after admission to the program), suggesting that other factors such as the known obesity-risk genes are also related to weight regain. Author contributions Anke Hinney, Johannes Hebebrand, and Thomas Reinehr developed the study design. Nina Lass and Barbara Wolters performed the anthropometrical measurements. Carolin Pütter performed the statistical analyses. Harald Grallert and Thomas Illig genotyped the included SNPs. Thomas Reinehr wrote the first draft of the paper. All authors discussed the findings and were involved in the writing of the manuscript. Acknowledgments: We thank the children who participated in this study. Conflict of interest statement Authors conflict of interest disclosure: All authors declare that there is no conflict of interest. Funding: This work was supported by grants from the German Federal Ministry of Education and Research (BMBF), the National Genome Research Network (NGFNplus; grant number 01GS0820), and the Competence Network Obesity (LARGE; grant number 01 GI0839). The funder had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Received May 11, 2013; accepted June 12, 2013; previously published online July 10, 2013 References 1. Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 2010;42:937 48. 2. Reinehr T, Hinney A, Nguyen TT, Hebebrand J. Evidence of an influence of a polymorphism near the INSIG2 on Xweight loss during a lifestyle intervention in obese children and adolescents. Diabetes 2008;57/3:623 6. 3. Scherag A, Kleber M, Boes T, Kolbe AL, Ruth A, et al. SDCCAG8 Obesity alleles and reduced weight loss after a lifestyle intervention in overweight children and adolescents. Obesity (Silver Spring) 2011;20:466 70. 4. Franks PW, Jablonski KA, Delahanty LM, McAteer JB, Kahn SE, et al. Assessing gene-treatment interactions at the FTO and INSIG2 loci on obesity-related traits in the Diabetes Prevention Program. Diabetologia 2008;51:2214 23. 5. Reinehr T, Hinney A, Toschke AM, Hebebrand J. Aggravating effect of INSIG2 and FTO on overweight reduction in a one-year lifestyle intervention. Arch Dis Child 2009;94:965 7. 6. Moleres A, Rendo-Urteaga T, Zulet MA, Marcos A, Campoy C, et al. Obesity susceptibility loci on body mass index and weight loss in Spanish adolescents after a lifestyle intervention. J Pediatr 2012;161:466 70. 7. Delahanty LM, Pan Q, Jablonski KA, Watson KE, McCaffery JM, et al. Genetic predictors of weight loss and weight regain after intensive lifestyle modification, metformin treatment, or standard care in the Diabetes Prevention Program. Diabetes Care 2012;35:363 6. 8. Reinehr T, Hebebrand J, Friedel S, Toschke AM, Brumm H, et al. Lifestyle intervention in obese children with variations in the melanocortin 4 receptor gene. Obesity (Silver Spring) 2009;17:382 9. 9. Farooqi IS, O Rahilly S. Recent advances in the genetics of severe childhood obesity. Arch Dis Child 2000; 83:31 4. 10. Cole TJ. The LMS method for constructing normalized growth standards. Eur J Clin Nutr 1990;44:45 60. 11. Kromeyer-Hauschild K, Wabitsch M, Geller F, Ziegler A, Geiss HC, et al. Percentiles of body mass index in children and adolescents evaluated from different regional German studies. Monatsschr Kinderheilkd 2001;149:807 18. 12. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. Br Med J 2000;320:1240 3. 13. Reinehr T, Kersting M, Wollenhaupt A, Alexy U, Kling B, et al. Evaluation of the training program OBELDICKS for obese children and adolescents. Klin Padiatr 2005;217:1 8. 14. Reinehr T, Temmesfeld M, Kersting M, de Sousa G., Toschke AM. Four-year follow-up of children and adolescents participating in an obesity intervention program. Int J Obes (Lond) 2007;31:1074 7. 15. http://www.broadinstitute.org/mpg/snap/ldsearchpw.php. Accessed on June 20, 2013.