| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Kinesiology (A.T., L.B., V.D., L.P.), Department of Social and Preventive Medicine, Faculty of Medicine, Laval University, Ste-Foy, Québec, Canada, G1K 7P4; Pennington Biomedical Research Center (C.B.), Louisiana State University, Baton Rouge, Louisiana 70808; and Laval Hospital Research Center (J.-P.D.), Québec Heart Institute, Ste-Foy, Québec, Canada G1V 4G5
Address all correspondence and requests for reprints to: Angelo Tremblay, M.D., Division of Kinesiology, Department of Social and Preventive Medicine, Faculty of Medicine, Laval University, Ste-Foy, Québec, Canada G1K 7P4. E-mail: angelo.tremblay{at}kin.msp.ulaval.ca.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
The literature provides support for the hypothesis that sequence variation in the glucocorticoid receptor (GRL) may be involved in the gene-environment interaction effect on adiposity (8). Hormones binding this receptor, i.e. cortisol and corticosterone, exert orexigenic (9) and antithermogenic effects (10, 11), which are concordant with the well-established effect of adrenalectomy on the development of obesity in most animal models (12, 13, 14). The importance of the glucocorticoid receptor polymorphisms on adiposity is also emphasized by one of our recent clinical interventions, which showed a positive relationship between changes in desire to eat and plasma cortisol in obese individuals experiencing resistance to further loss of body fat (15). Finally, recent studies focusing on the specific contribution of the glucocorticoid receptor polymorphism to variations in both body fatness and visceral fat also provide support that this marker could be relevant. Indeed, Buemann et al. (16) examined a subset of data collected in phase 2 of the QFS and found a significant higher abdominal visceral fat area in the homozygous for the 4.5-kb allele in the lower tertile of percent body fat. In another study, Rosmond et al. (17) found significant GRL between-genotype differences for both body mass index and visceral fat accumulation among glucocorticoid receptor genotypes.
Taken together, these observations suggest that genetic variation at the glucocorticoid receptor should be considered in the search of a specific gene-environment interaction effect. Therefore, we have examined data collected in phases 1 and 2 of the QFS to determine whether genetic variation within the intron 2 of the glucocorticoid receptor [GRL IVS2-BclI restriction fragment length polymorphism (RFLP)] is associated with significant changes in adiposity over time.
| Subjects and Methods |
|---|
|
|
|---|
The subjects of this study were healthy male and female children who participated in phase 1 (19781982) of the QFS and who were retested in phase 2 (19901992), an average of 12 yr later. We specifically selected subjects who were children at the beginning of the QFS because they experienced a substantial weight gain in the transition between preadolescence or adolescence and adulthood over the 12-yr follow-up. The children who were selected were those for whom DNA was available. Some of the descriptive characteristics of the subjects are presented in Tables 1
and 2
. Each subject gave his/her written consent to participate in this study. In phase 1, the consent was also given by parents. The procedures of both phases 1 and 2 of the QFS were approved by the Laval University Ethics Committee.
|
|
Male and female subjects were considered separately, and they were classified on the basis of the glucocorticoid receptor polymorphism (GRL IVS2-BclI). Thus, for each gender, data were compared between the three following groups: 1) the homozygous for the 4.5-kb allele; 2) the heterozygous carrying the 4.5- and 2.3-kb alleles; and 3) the homozygous for the 2.3-kb allele. The outcome variables included phenotypes measured at phase 1 and phase 2 and the changes over time between the two phases.
Measurements
A Harpenden skinfold caliper was used to measure sc skinfold thickness at the following sites: biceps, triceps, calf, subscapular region, abdomen, and suprailiac crest. The sum of the three latter skinfolds was used as a marker of trunk adiposity, whereas the sum of the other three skinfolds was calculated as an indicator of limb adiposity. The methodology recommended by the International Biological Program (18) was followed for these measurements.
Hydrostatic weighing was performed in a subset of subjects in phase 1 and in all subjects in phase 2. Percent body fat was estimated from body density by using the Siri equation (19). Fat mass was derived by multiplying body weight by the estimated percent body fat.
Food intake was measured with a three-dimensional dietary record as previously described (20). The tables of Dubuc and Lahaie (1978) (21) were used to calculate energy and nutrient intake in phase 1, whereas these calculations were performed with the Canadian nutrient file (1988) in phase 2 of the study. Dietary fat intake and alcohol consumption were the variables specifically considered for the present study.
A physical activity diary was used to assess participation in vigorous physical activity (22). This diary was completed during the same days (2 weekdays and 1 weekend day) as those selected to record food intake. This method classifies activities on a scale from 1 (sleep) to 9 (vigorous exercise). For each 15-min period, the subject coded the dominant activity of this period. In phase 1 of the study, children were occasionally assisted by their parents in completing both dietary and activity records. The focus of the present study is on the impact of vigorous activities that have been shown to influence body fatness (6). The number of 15-min periods coded as 9 during the three-dimensional observation period was therefore used.
Determination of the genotypes
Genomic DNA was extracted from permanent lymphoblastoid cells (23) using the proteinase K and phenol/chloroform technique (24). In 50-µl total vol, 5 µg genomic DNA were digested with 50 U of the BclI restriction enzyme (New England Biolabs, Inc., Mississauga, Canada) at 50 C for 1620 h. After electrophoresis on 1.2% agarose gel for 1620 h at 50 mAmp, the separated digested DNA was transferred to a nylon membrane by alkaline transfer. DNA fragments were hybridized with OB7 (25), a 4.3-kb cDNA probe complementary to the
-form of the GRL transcript except for the first 500 bp. The probe was labeled with (
-32P)deoxy-CTP by random priming using a T7 DNA polymerase kit (Quick Prime; Pharmacia Biotech, Mississauga, Canada). Hybridization was at 65 C for 12 h in a solution containing 0.25 M Na2HPO4, 1 mM EDTA, 7% sodium dodecyl sulfate, and 0.1 µg/µl sonicated salmon sperm (SSPE) DNA (Stratagene, La Jolla, CA). Subsequently, membranes were washed two to four times in a solution of SSPE (20 mM NaCl, 1 mM NaH2PO4, 0.1 mM EDTA) plus 0.1% sodium dodecyl sulfate and were placed in a cassette with an X-OMAT AR film (Eastman Kodak Co., New Haven, CT) for 72120 h at 70 C. On some occasions, membranes were washed in a more stringent solution containing 10x less SSPE.
Statistical analysis
The comparison between genotypes for each gender was performed by using first an analysis of covariance analysis adjusting for age at the beginning of phase 1. Once significant differences were detected, the Duncan multiple range test was used to determine which genotypes were differing from each other. In a second step of analysis, the analysis of covariance was repeated by correcting further for the change in fat-balance-related variables, i.e.
percent energy intake as fat,
alcohol consumption, and
number of 15-min periods coded as 9 (vigorous activity). A
represents the value observed in phase 2 minus that obtained in phase 1. To evaluate whether genotype and allele frequency were in Hardy-Weinberg equilibrium, a
2 test was applied. Power calculations were made to determine whether a sample size of 173 subjects provided a sufficient statistical power to detect a significant association between adiposity changes and the GRL IVS2-BclI polymorphism. Power calculations were performed using a SAS Institute, Inc. (Cary, NC) macro (26). Power calculations indicate that, for a gene with a genotype frequency of 5%, the power to detect association at
= 0.05 is greater than 80%.
| Results |
|---|
|
|
|---|
In general, the same trends were observed for male subjects (Table 2
). Indeed, most morphological indicators were comparable between genotypes, both before and after the follow-up, despite the fact that substantial time effect on adiposity was noted.
To more directly assess the time effect on adiposity in each genotype, data were also analyzed by considering changes that occurred between phase 1 and phase 2 of the study. As illustrated in Fig. 1
, the increase in sc skinfolds observed over the 12-yr follow-up was considerably greater in the heterozygous subgroup of female subjects, compared with the two homozygous groups. Specifically, the increase in the sum of six skinfolds and the sum of three trunk and extremity skinfolds observed in the 4.5/2.3 genotype subgroup was more than twice that seen in the 4.5/4.5 and the 2.3/2.3 genotype subgroups. In male subjects, the same trend was observed, but it did not reach statistical significance for any sc adiposity marker. A more pronounced increase in body fat mass was also found in the heterozygous male and female groups, but this effect was not statistically significant, partly because of the reduced statistical power for these variables (results not illustrated).
|
2 analysis, P = not significant with 2 degrees of freedom; n = 173). Similar results were obtained when each gender was considered separately. | Discussion |
|---|
|
|
|---|
There are many factors that can influence body adiposity over a 12-yr follow-up. It is therefore difficult to characterize this interaction effect on the basis of the present results. However, we measured several phenotypes that may help in this regard. These phenotypes include percent energy as dietary fat, alcohol consumption, and participation in vigorous physical activities, which all have been shown to explain a small part of the variance in adiposity in the context of the QFS (3, 4, 6, 29). To take into account the contribution of these factors to variations in adiposity over time, the comparison of genotype subgroups was repeated by adjusting for the changes in these variables. The results showed that the difference in adiposity between the heterozygous females and the two homozygous subgroups was essentially unchanged by this statistical correction. We would rather believe that other factors more closely related to mechanism involving the glucocorticoid receptor might be key environmental factors explaining the observed gene-time interaction effect. In this regard, stress-related factors whose biological effects are modulated by the hypothalamic-pituitary-adrenal axis can also be realistically considered as another potential explanation of the increase in adiposity that occurs over time. For instance, a survey of 243 obese women revealed that factors such as psychological stress and medications could account for up to 75% of the total weight gain over time (30). A study of Swedish children led to the conclusion that rapid weight gain during school years is an indicator of psychological stress (31). This is also in agreement with the demonstration that weight fluctuations are associated with negative psychological attributes, including stress (32). These observations are concordant with the fact that the products of stress-related activation of the hypothalamic-pituitary-adrenal, i.e. corticosterone and cortisol, promote positive energy balance by increasing energy intake (9) and by decreasing thermogenesis (10, 11). Therefore, even though the present study has not identified the environmental factor(s) accounting for the more pronounced predisposition of heterozygous females at the glucocorticoid receptor marker to gain weight over time, the impact of stress-related factors as determinants of the gene-time (age) interaction effect should be investigated.
The fact that fat gain was greater in the heterozygous subjects at the GRL IVS2-BclI marker, compared with the homozygous subjects, raises questions as to the mechanisms underlying this genetic effect. Association studies reporting significant differences between heterozygotes and homozygotes are not infrequent. Why heterozygous women gained more than twice the sc adiposity as homozygous women can be explained by a number of hypotheses. First, GRL IVS2-BclI variant could be in linkage disequilibrium with another polymorphism that is responsible for the observed phenotype. Second, we cannot rule out an interaction effect between the GRL IVS2-BclI marker and environmental factors, such as maternity, or other genetic factors that could influence the phenotype. Third, GRL is a highly regulated transcription factor, comprising three known isoforms that have different activity. These receptors are ubiquitously expressed and exert different functions, depending on their localization (33). GRL can form homodimer or heterodimer with mineralocorticoid receptor molecules and interact with other proteins, such as activating protein-1, to activate or repress gene expression under its control (33). Mutations in proteins that can form homodimer and heterodimer can impair dimerization and alter normal function (34). It is possible that changes in GRL isoforms ratio, the presence of one mutated allele, or compound heterozygosity could modify GRL responses to glucocorticoid, which is something that would not be seen in homozygous women. This is not obvious for the homozygous women because alternative pathways could be activated to compensate for the effect of carrying two deleterious alleles (4.5/4.5). Two mice models can support the heterozygous phenotype. One model, heterozygous for cyclooxygenase-2±) (35), develops obesity; and the other model, heterozygous for glucose transporter-4± (36), develops diabetes. These two models, when homozygous for the null allele, were not developing obesity or diabetes.
In summary, the results of this study demonstrate that the transition from adolescence to adulthood is accompanied by substantial body fat gain, both in females and males. The main finding of the study is the observation of a much greater increase in adiposity, with age, in female subjects displaying the 4.5/2.3 BclI genotype at the glucocorticoid receptor gene. The statistical significance of this effect was unchanged by correcting for changes, over time, in percentage dietary energy as fat, alcohol consumption, and vigorous physical activity participation. This suggests that other environmental factors interact with the glucocorticoid receptor to cause a more pronounced fat gain, with age, in heterozygous females. The study of the impact of stress-related factors on energy balance should be pursued if genotype-age interaction effects on body fatness are to be defined.
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: QFS, Québec Family Study; RFLP, restriction fragment length polymorphism; SSPE, sonicated salmon sperm.
Received September 30, 2002.
Accepted March 21, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B. R Walker Glucocorticoids and Cardiovascular Disease Eur. J. Endocrinol., November 1, 2007; 157(5): 545 - 559. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Vogels, E. C. Mariman, F. G Bouwman, A. D. Kester, K. Diepvens, and M. S Westerterp-Plantenga Relation of weight maintenance and dietary restraint to peroxisome proliferator-activated receptor {gamma}2, glucocorticoid receptor, and ciliary neurotrophic factor polymorphisms Am. J. Clinical Nutrition, October 1, 2005; 82(4): 740 - 746. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Bouchard, V. Drapeau, V. Provencher, S. Lemieux, Y. Chagnon, T. Rice, D. Rao, M.-C. Vohl, A. Tremblay, C. Bouchard, et al. Neuromedin {beta}: a strong candidate gene linking eating behaviors and susceptibility to obesity Am. J. Clinical Nutrition, December 1, 2004; 80(6): 1478 - 1486. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. P. Bagby Obesity-Initiated Metabolic Syndrome and the Kidney: A Recipe for Chronic Kidney Disease? J. Am. Soc. Nephrol., November 1, 2004; 15(11): 2775 - 2791. [Full Text] [PDF] |
||||
![]() |
E. F.C. van Rossum and S. W.J. Lamberts Polymorphisms in the Glucocorticoid Receptor Gene and Their Associations with Metabolic Parameters and Body Composition Recent Prog. Horm. Res., January 1, 2004; 59(1): 333 - 357. [Abstract] [Full Text] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |