Publications

OBJECTIVE: Eating in the absence of hunger (EAH) may be a genetically influenced phenotype of overweight children, but evidence is limited. This research evaluated the heritability (h(2)) of EAH and its association with overweight among Hispanic children 5 to 18 years old. Genetic and environmental associations of EAH with overweight, fat mass, and key hormonal regulators of food intake were also evaluated.

RESEARCH METHODS AND PROCEDURES: A family design was used to study 801 children from 300 Hispanic families. Weighed food intakes were used to measure EAH after an ad libitum dinner providing 50% of estimated energy needs. Fasting ghrelin, amylin, insulin, and leptin were measured by immunoassays. Measured heights, weights, and fat mass (using DXA) were obtained. Total energy expenditure (TEE) was measured by room respiration calorimetry.

RESULTS: On average, children consumed 41% of TEE at the dinner meal, followed by an additional 19% of TEE in the absence of hunger. Overweight children consumed 6.5% more energy at dinner (p < 0.001) and 14% more energy in the absence of hunger (p < 0.001) than non-overweight children. Significant heritabilities were seen for EAH (h(2) = 0.51) and dinner intake (h(2) = 0.52) and for fasting levels of ghrelin (h(2) = 0.67), amylin (h(2) = 0.37), insulin (h(2) = 0.37), and leptin (h(2) = 0.34). Genetic correlations were seen between eating behavior and fasting hormones, suggesting common underlying genes affecting their expression.

DISCUSSION: This research provides new evidence that overweight Hispanic children exhibit elevated levels of hyperphagic eating behaviors that are influenced by genetic endowment.

Birth weight has been shown to be associated with obesity and metabolic diseases in adulthood, however, the genetic contribution is still controversial. The objective of this analysis is to explore the genetic contribution to the relationship between birth weight and later risk for obesity and metabolic diseases in Hispanic children. Subjects were 1,030 Hispanic children in the Viva La Familia Study. Phenotypes included body size, body composition, blood pressure, fasting glucose, insulin, lipids, and liver enzymes. Birth weights were obtained from Texas birth certificates. Quantitative genetic analyses were conducted using SOLAR software. Birth weight was highly heritable, as were all other phenotypes. Phenotypically, birth weight was positively correlated to childhood body size parameters. Decomposition of these phenotypic correlations into genetic and environmental components revealed significant genetic correlations, ranging from 0.30 to 0.59. Negative genetic correlations were seen between birth weight and lipids. The genome scan of birth weight mapped to a region near marker D10S537 (LOD = 2.6). The bivariate genome-wide scan of birth weight and childhood weight or total cholesterol, improved the LOD score to 3.09 and 2.85, respectively. Chromosome 10q22 harbors genes influencing both birth weight and childhood body size and cardiovascular disease risk in Hispanic children.

OBJECTIVE: Cholecystokinin (CCK) is known to inhibit food intake and is an important signal for controlling meal volume, indicating a possible role in weight regulation. Our objective was to investigate genetic influences on plasma CCK in baboons.

RESEARCH METHODS AND PROCEDURES: Subjects were 376 baboons (males = 113, females = 263) from the Southwest National Primate Research Center, housed at the Southwest Foundation for Biomedical Research, San Antonio, Texas. Anthropometric and biochemical parameters were analyzed. Genetic effects on plasma CCK were estimated by the maximum likelihood-based variance components method implemented in the software program SOLAR (Sequential Oligogenic Linkage Analysis Routines).

RESULTS: Male baboons (32.7 +/- 6 kg) were much heavier than females (20.2 +/- 4 kg). Similarly, mean (+/- standard deviation) plasma CCK values were also higher in male baboons (13.8 +/- 6 pM) than female baboons (12.5 +/- 4 pM). Significant heritabilities were observed for plasma CCK (0.14 +/- 0.1, p < 0.05), body weight (h2 = 0.62 +/- 0.15, p < 10(-8)), and glucose (h2 = 0.68 +/- 0.17, p < 10(-7)). A genome-wide scan of plasma CCK detected a strong signal for a quantitative trait locus (QTL) on chromosome 17p12-13 [logarithm of the odds (LOD) = 3.1] near marker D17S804. Suggestive evidence of a second QTL was observed on chromosome 4q34-35 (LOD = 2.3) near marker D4S2374.

DISCUSSION: A substantial contribution of additive genetic effects to the variation in plasma levels of CCK was demonstrated in baboons. The identification of a QTL for plasma CCK on chromosome 17p is significant, as several obesity-related traits such as BMI, leptin, adiponectin, and acylation stimulating protein have already been mapped to this region.

OBJECTIVE: Obesity is a growing and important public health problem in Western countries and worldwide. There is ample evidence that both environmental and genetic factors influence the risk of developing obesity. Although a number of genes influencing obesity and obesity-related measures have been localized, it is clear that others remain to be identified. The rate of obesity is particularly high in American Indian populations. This study reports the results of a genome-wide scan for loci influencing BMI and weight in 963 individuals in 58 families from three American Indian populations in Arizona, Oklahoma, and North and South Dakota participating in the Strong Heart Family Study.

RESEARCH METHODS AND PROCEDURES: Short tandem repeat markers were genotyped, resulting in a marker map with an average spacing of 10 centimorgans. Standard multipoint variance component linkage methods were used.

RESULTS: Significant evidence of linkage was observed in the overall sample, including all three study sites, for a locus on chromosome 4q35 [logarithm of the odds (LOD)=5.17 for weight, 5.08 for BMI]. Analyses of the three study sites individually showed that the greatest linkage support for the chromosome 4 locus came from Arizona (LOD=2.6 for BMI), but that LOD scores for weight were >1 in all three samples. Suggestive linkage signals (LOD>2) were also observed on chromosomes 5, 7, 8, and 10.

DISCUSSION: The chromosome 4 locus detected in this scan is in a region lacking any obvious positional candidate genes with known functions related to obesity. This locus may represent a novel obesity gene.

Plasminogen is a hemostasis-related phenotype and is commonly implicated in thrombotic and bleeding disorders. In the San Antonio Family Heart Study (SAFHS), we performed to our knowledge the first genomewide linkage scan for quantitative trait loci (QTLs) that influence the level of plasminogen. The subset of the SAFHS population used for this study consists of 629 individuals distributed across 26 extended Mexican American families. Pedigree-based variance component linkage analyses were performed using SOLAR. The mean plasminogen level was 114.94% +/- 17.8 (range, 42-195). The heritability (h2) of plasminogen was 0.43 +/- 0.08 (p < 6.3 x 10(-13)). One region on chromosome 12 (12q14.1) showed suggestive evidence of linkage (LOD = 2.73, nominal p < 0.0002, genomewide p = 0.0786) near marker D12S1609. Because plasminogen has important effects in many human health problems, such as cancer and atherosclerosis, the role of this putative QTL in the regulation of plasminogen variability needs to be studied further.

Quantitative differences in gene expression are thought to contribute to phenotypic differences between individuals. We generated genome-wide transcriptional profiles of lymphocyte samples from 1,240 participants in the San Antonio Family Heart Study. The expression levels of 85% of the 19,648 detected autosomal transcripts were significantly heritable. Linkage analysis uncovered >1,000 cis-regulated transcripts at a false discovery rate of 5% and showed that the expression quantitative trait loci with the most significant linkage evidence are often located at the structural locus of a given transcript. To highlight the usefulness of this much-enlarged map of cis-regulated transcripts for the discovery of genes that influence complex traits in humans, as an example we selected high-density lipoprotein cholesterol concentration as a phenotype of clinical importance, and identified the cis-regulated vanin 1 (VNN1) gene as harboring sequence variants that influence high-density lipoprotein cholesterol concentrations.

This study was conducted to investigate genetic influence on serum ghrelin and its relationship with adiposity-related phenotypes in Hispanic children (n=1030) from the Viva La Familia study (VFS). Anthropometric measurements and levels of serum ghrelin were estimated and genetic analyses conducted according to standard procedures. Mean age, body mass index (BMI), and serum ghrelin were 11+/-0.13 y, 25+/-0.24 kg/m2 and 38+/-0.5 ng/mL, respectively. Significant heritabilities (p<0.001) were obtained for BMI, weight, fat mass, percent fat, waist circumference, waist-to-height ratio, and ghrelin. Bivariate analyses of ghrelin with adiposity traits showed significant negative genetic correlations (p<0.0001) with weight, BMI, fat mass, percent fat, waist circumference, and waist-to-height ratio. A genome-wide scan for ghrelin detected significant linkage on chromosome 1p36.2 between STR markers D1S2697 and D1S199 (LOD=3.2). The same region on chromosome 1 was the site of linkage for insulin (LOD=3.3), insulinlike growth factor binding protein 1 (IGFBP1) (LOD=3.4), homeostatic model assessment method (HOMA) (LOD=2.9), and C-peptide (LOD=2.0). Several family-based studies have reported linkages for obesity-related phenotypes in the region of 1p36. These results indicate the importance of this region in relation to adiposity in children from the VFS.

Although previous genome scans have searched for quantitative-trait loci (QTLs) influencing variation in blood pressure (BP), few have investigated the rate of change in BP over time as a phenotype. Here, we compare results from genomewide scans to localize QTLs for systolic, diastolic, and mean arterial BPs (SBP, DBP, and MBP, respectively) and for rates of change in systolic, diastolic, and mean arterial BPs (rSBP, rDBP, and rMBP, respectively), with use of the longitudinal data collected about Mexican Americans of the San Antonio Family Heart Study (SAFHS). Significant evidence of linkage was found for rSBP (LOD 4.15) and for rMBP (LOD 3.94) near marker D11S4464 located on chromosome 11q24.1. This same chromosome 11q region also shows suggestive linkage to SBP (LOD 2.23) and MBP (LOD 2.37) measurements collected during the second clinic visit. Suggestive evidence of linkage to chromosome 5 was also found for rMBP, to chromosome 16 for rSBP, and to chromosomes 1, 5, 6, 7, and 21 for the single-time-point BP traits collected at the first two SAFHS clinic visits. We also present results from fine mapping the chromosome 11 QTL with use of SNP-association analysis within candidate genes identified from a bioinformatic search of the region and from whole-genome transcriptional expression data collected from 1,240 SAFHS participants. Our results show that the use of longitudinal BP data to calculate the rate of change in BP over time provides more information than do the single-time measurements, since they reveal physiological trends in the subjects that a single-time measurement could never capture. Further investigation of this region is necessary for the identification of the genetic variation responsible for QTLs influencing the rate of change in BP.

OBJECTIVE: The prevalence of childhood obesity has increased dramatically in the United States. Early presentation of type 2 diabetes has been observed in children and adolescents, especially in the Hispanic population. The genetic contribution of glucose homeostasis related to childhood obesity is poorly understood. The objective of this study was to localize quantitative trait loci influencing fasting serum glucose levels in Hispanic children participating in the Viva La Familia Study.

DESIGN: Subjects were 1030 children ascertained through an overweight child from 319 Hispanic families. Fasting serum glucose levels were measured enzymatically, and genetic linkage analyses were conducted using SOLAR software.

RESULTS: Fasting glucose was heritable, with a heritability of 0.62 +/- 0.08 (P < 0.01). Genome-wide scan mapped fasting serum glucose to markers D13S158-D13S173 on chromosome 13q (LOD score of 4.6). A strong positional candidate gene is insulin receptor substrate 2, regulator of glucose homeostasis and a candidate gene for obesity. This region was reported previously to be linked to obesity- and diabetes-related phenotypes.

CONCLUSIONS: A quantitative trait locus on chromosome 13q contributes to the variation in fasting serum glucose levels in Hispanic children at high risk for obesity.