SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods and Procedures
  4. SUPPLEMENTARY MATERIAL
  5. ACKNOWLEDGEMENT
  6. DISCLOSURE
  7. References
  8. Supporting Information

Recent genome wide association studies (GWAS) have revealed a number of genetic variants robustly associated with bone mineral density (BMD) and/or osteoporosis. Evidence from epidemiological and clinical studies has shown an association between BMD and BMI, presumably as a consequence of bone loading. We investigated the 23 previously published BMD GWAS-derived loci in the context of childhood obesity by leveraging our existing genome-wide genotyped European American cohort of 1,106 obese children (BMI ≥95th percentile) and 5,997 controls (BMI <95th percentile). Evidence of association was only observed at one locus, namely Osterix (SP7), with the G allele of rs2016266 being significantly over-represented among childhood obesity cases (P = 2.85 × 10−3). When restricting these analyses to each gender, we observed strong association between rs2016266 and childhood obesity in females (477 cases and 2,867 controls; P = 3.56 × 10−4), which survived adjustment for all tests applied. However, no evidence of association was observed among males. Interestingly, Osterix is the only GWAS locus uncovered to date that has also been previously implicated in the determination of BMD in childhood. In conclusion, these findings indicate that a well established variant at the Osterix locus associated with increased BMD is also associated with childhood obesity primarily in females.

Osteoporosis is a systemic skeletal disease characterized by low bone mass and micro-architectural deterioration of bone tissue, with a consequent increase in bone fragility and susceptibility to fracture (1). This disorder is most common in women after menopause due to a rapid drop in estrogen levels.

In adults, bone mineral density (BMD) is the single best predictor of fragility fractures and is used as an important diagnostic criterion for osteoporosis (2). It is now widely recognized from twin studies (3) that BMD is highly heritable and recent discoveries resulting from genome wide association studies (GWAS) of this trait further confirm a genetic component. Since the first such report in 2008, outlining OPG (TNFRSF11B) and LRP5 (4), additional loci have been reported at RANKL, ESR1, 1p36 (ZBTB40), MHC SOST, MARK3, SP7 (Osterix), and TNFRSF11A (RANK) (5,6). In early 2009, a meta-analysis of five GWAS of femoral neck and lumbar spine BMD in a total of 19,195 subjects of northern European decent added another 13 new loci, namely GPR177, SPTBN1, CTNNB1, MEPE, MEF2C, STARD3NL, FLJ42280, LRP4, DCDC5, SOX6, FOXL1, HDAC5, CRHR1 (7). Altogether, variants at 23 loci have been established to date with GWAS of adult BMD or related traits in populations of European ancestry.

In contrast to the studies of adult BMD, there have been few reports of genetic studies of BMD determination in childhood. Familial resemblance of BMD is expressed in prepuberty and is greater in adolescence and early adulthood than in later life (8). Bone accrual during childhood and adolescence is a determinant of peak bone size and mass, which are traits that are crucial for bone structural strength later in life and ultimately the degree of risk of presenting with osteoporosis (9). To date, the only GWAS of BMD in children revealed a single locus near the Osterix (SP7) gene (10), which was also observed in the adult studies.

During childhood, BMI is positively associated with increased bone mass and bone dimensions (11,12,13). A recent report from a pediatric cohort in the Avon Longitudinal Study of Parents and Children (ALSPAC) showed that total fat mass was strongly correlated with total body, spinal, and upper and lower limb bone mineral content and suggested that fat mass influences bone mass in children (14).

In this study, we have leveraged our existing genome-wide genotyped pediatric cohort with height and weight measures to examine the known BMD GWAS loci in the context of childhood obesity. In our analyses, single-nucleotide polymorphisms (SNPs) corresponding to the 23 BMD loci discovered in adult studies (4,5,6,7,10) were investigated in our childhood obesity cohort (Table 1). Evidence of association was only observed with rs2016266 at the Osterix locus (P = 2.85 × 10−3), the only GWAS-discovered BMD locus to date that has also been reported to impact BMD specifically in children (10). The minor G allele of rs2016266 was associated with childhood obesity; this is the same allele that has been previously associated with increased BMD. We did not observe evidence of association at any of the other loci investigated which may be due to limited power, but probably indicates that they do not play a major role in the pathogenesis of childhood obesity. Additionally, our positive control of FTO lends credibility to our results. The corresponding results when treating BMI as a quantitative trait are summarized in Supplementary Table S1 online.

Table 1.  Case-control analysis results for bone mineral density associated loci in the context of childhood obesity
inline image

Our findings lend an intriguing perspective on the association between BMI and BMD (and other bone outcomes) in children, since both are associated with the same genetic variant. These findings suggest that the increased BMD-associated with obesity may be due to more than simply mechanical loading of bone through excess weight, since the minor G allele of rs2016266 confers a higher BMD in both children and adults and we observe that it confers risk for childhood obesity.

Timpson et al. suggested that Osterix may influence bone development via longitudinal and periosteal bone growth (10). Adipocytes and osteoblasts are both derived from mesenchymal stem cells, so it is possible that Osterix influences both BMD and BMI in children by its effect on this cell lineage.

At least one study previously suggested that fat mass influences cortical bone mass more in female children than in males (15). As such, we examined rs2016266 in each gender separately (males: 629 cases and 3,130 controls; females: 477 cases and 2,867 controls) with respect to our phenotypes of interest. We observed strong association between rs2016266 and obesity in female children (P = 3.56 × 10−4) (Table 2), which survived adjustment for all tests applied. Breaking this observation down further with respect to age, the association is strongly driven by females aged 10 years or older (see Supplementary Table S2 online). Conversely, we did not observe association between this locus and obesity in male children (P = 0.42). In addition, we found no consistent gender-specific association at any of the other known BMD loci (see Supplementary Tables S3 and S4 online).

Table 2.  Association of the Osterix locus (rs2016266) with childhood obesity in each gender separately
inline image

As such, our study finds that the Osterix association with childhood obesity is very gender-specific i.e., the association was primarily observed in females, despite more males being present in this cohort. Analysis with respect to gender was not reported in the original GWAS report of this locus with respect to BMD in either adults or children (6,10). Despite not knowing that outcome, this observation is in keeping with the concept that females are more susceptible to osteoporosis in later life; however, it is not known how this locus confers its influence in a gender-specific manner. The fact that the association is observed primarily in females over the age of 10 years old suggests that pubertal hormones may be involved, influencing the action of this zinc-containing transcription factor which is already known to be essential for osteoblast differentiation and bone formation (16). Alternatively, since changes in BMD are very slow, it is possible that the age association is simply caused by the time that it takes to create a measurable change in BMD. It will be important in later studies to explore and refine the molecular significance of the present correlation between increased BMD and obesity in female children.

In conclusion, variation at the only GWAS locus implicated in the determination of pediatric BMD to date, i.e., Osterix, is also associated with childhood obesity. This association was primarily driven by an influence on female children; however, it remains to be determined how Osterix confers this gender-specific effect and future experiments investigating the repertoire of genes that are transcriptionally regulated by Osterix are underway that could shed light on its dual role of impacting bone density and adiposity in children.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Methods and Procedures
  4. SUPPLEMENTARY MATERIAL
  5. ACKNOWLEDGEMENT
  6. DISCLOSURE
  7. References
  8. Supporting Information

Research subjects

Our study cohort consisted of 7,103 children of European ancestry with BMI information. All subjects were biologically unrelated and aged between 2 and 18 years old. Blood was drawn and DNA subsequently extracted for genotyping. This study was approved by the institutional review board of the Children's Hospital of Philadelphia. Parental informed consent was given for each study participant.

BMI was compared to the Centers for Disease Control reference values (17) and age- and sex-specific Z-scores (standard deviation units) were calculated. Childhood obesity was defined as BMI ≥95th percentile (Z-score = 1.645) (18) and controls were children below this threshold. The age-range was restricted to subjects between 2 and 18 years of age due to the age-range limit of the BMI reference range. Subjects with BMI Z-scores >3 or <−3 were excluded as outliers to avoid the consequences of potential measurement error or Mendelian causes of extreme obesity (see Supplementary Table S5 online for clinical characteristics).

Genotyping

We performed high throughput genome-wide SNP genotyping, using the Illumina Infinium II HumanHap550 or Human 610 BeadChip technology(Illumina, San Diego, CA), at Children's Hospital of Philadelphia's Center for Applied Genomics, as described previously (19). The SNPs analyzed survived the filtering of the genome-wide dataset for SNPs with call rates <95%, minor allele frequency <1%, missing rate per person >2%, and Hardy-Weinberg equilibrium P < 10−5.

We used previously reported SNPs or the best surrogate SNP available based on HapMap data for CEPH (Centre d'Etude du Polymorphisme Humain) reference individuals from Utah (CEU) (see Supplementary Table S6 online). rs3751812 at FTO was included as a positive control as it has previously been reported to yield very strong association with pediatric obesity.

Analysis

We queried the data for the indicated SNPs in our pediatric samples. All statistical analyses were carried out using plink (20). European ethnicity for our cohort was derived using the multi-dimensional scaling feature within plink. The single marker association analysis was carried out using the 1-df allelic χ2 test. Odds ratios and the corresponding 95% confidence intervals were calculated for each SNP.

ACKNOWLEDGEMENT

  1. Top of page
  2. Abstract
  3. Methods and Procedures
  4. SUPPLEMENTARY MATERIAL
  5. ACKNOWLEDGEMENT
  6. DISCLOSURE
  7. References
  8. Supporting Information

We would like to thank all participating subjects and families. Elvira Dabaghyan, Hope Thomas, Kisha Harden, Andrew Hill, Kenya Fain, Crystal Johnson-Honesty, Cynthia Drummond, Shanell Harrison, and Sarah Wildrick provided expert assistance with genotyping or data collection and management. We would also like to thank Smari Kristinsson, Larus Arni Hermannsson, and Asbjörn Krisbjörnsson of Raförninn ehf for their extensive software design and contribution. This research was financially supported by the Children's Hospital of Philadelphia. The study is supported by an Institute Development Award from The Children's Hospital of Philadelphia, a Research Development Award from the Cotswold Foundation and NIH grant 1R01HD056465-01A1.

References

  1. Top of page
  2. Abstract
  3. Methods and Procedures
  4. SUPPLEMENTARY MATERIAL
  5. ACKNOWLEDGEMENT
  6. DISCLOSURE
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods and Procedures
  4. SUPPLEMENTARY MATERIAL
  5. ACKNOWLEDGEMENT
  6. DISCLOSURE
  7. References
  8. Supporting Information

supporting Information

FilenameFormatSizeDescription
oby_2254_sm_1.doc69KSupporting info item
oby_2254_sm_2.doc31KSupporting info item
oby_2254_sm_3.doc76KSupporting info item
oby_2254_sm_4.doc76KSupporting info item
oby_2254_sm_5.doc35KSupporting info item
oby_2254_sm_6.doc30KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.