Published online before print September 27, 2004, doi: 10.1161/01.HYP.0000144294.28985.d5
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2004;44:681.)
© 2004 American Heart Association, Inc.
Scientific Contributions |
Interaction Between an 11ßHSD1 Gene Variant and Birth Era Modifies the Risk of Hypertension in Pima Indians
From the Phoenix Epidemiology and Clinical Research Branch (P.W.F., W.C.K., S.N., J.K., Y.-H.L., H.C.L., P.A.T., R.L.H.), National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Phoenix, Ariz; the Division of Cardiovascular and Medical Sciences (R.S.L.), Gardiner Institute, Western Infirmary, University of Glasgow, Glasgow, Scotland; the Endocrinology Unit (B.R.W.), University of Edinburgh, Western General Hospital, Edinburgh, Scotland; and the Carl T. Hayden VA Medical Center (P.A.P.), Phoenix, Ariz.
Correspondence to Paul W. Franks, Phoenix Epidemiology and Clinical Research Branch, National Institute of Diabetes and Digestive and Kidney Diseases, 1550 E Indian School Rd, Phoenix, AZ 85014. E-mail pfranks{at}niddk.nih.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
11ß-Hydroxysteroid dehydrogenase type 1 (11ßHSD1) is a candidate gene for hypertension, diabetes, and obesity through altered glucocorticoid production. This study explored the association of 11ßHSD1 gene variants with diabetes, hypertension, and obesity in a longitudinal population study of American Indians (N=918; exams=5508). In multivariate mixed models assuming an additive effect of genotype, a 5' upstream variant (rs846910) was associated with blood pressure (diastolic blood pressure ß=1.58 mm Hg per copy of the A allele, P=0.0008; systolic blood pressure ß=2.28 mm Hg per copy of the A allele, P=0.004; mean arterial blood pressure ß=1.83 mm Hg per copy of the A allele, P=0.0006) and hypertension (odds ratio=1.27 per copy of the A allele, P=0.02). However, birth date modified these associations (test for interaction: diastolic blood pressure P=0.16; systolic blood pressure P=0.007; mean arterial blood pressure P=0.01), such that the magnitude and direction of association between genotype and blood pressure changed with time. Finally, in models controlling for potential confounding by population stratification, we observed evidence of within-family effects for blood pressure (diastolic blood pressure ß=1.77 mm Hg per copy of the A allele, P=0.004; systolic blood pressure ß=2.04 mm Hg per copy of the A allele, P=0.07; mean arterial blood pressure ß=1.85 mm Hg per copy of the A allele, P=0.01) and for hypertension (odds ratio=1.26 per copy of the A allele; P=0.08). No association was observed for obesity. Associations with diabetes were similar in magnitude as reported previously but were not statistically significant. These data demonstrate association between genetic variability at 11ßHSD1 with hypertension, but these effects are modified by environmental factors.
Key Words: genetics • epidemiology • blood pressure • nutrition • gene expression
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The identification of genetic polymorphisms that predict the development of cardiovascular disease is important not only from the perspective of understanding disease etiology but also potentially in the targeted prevention of disease. Despite concerted efforts to identify such polymorphisms, there has been limited success in replicating genetic associations with complex disease.1 The possible explanations for why genetic association studies in complex disease have proven unreliable are many but include factors that are intrinsic to the original study (eg, inadequate control of confounding by population stratification,2,3 multiple hypothesis testing and selective reporting of results, and geneXenvironment interaction4) and factors that are extrinsic to the original study (eg, statistically underpowered replication studies5 and publication bias6).
The 11ß-hydroxysteroid dehydrogenase type 1 (11ßHSD1) gene, which is expressed predominantly in adipose tissue, liver, and the central nervous system, has been proposed as a candidate gene for diabetes and hypertension.7 The enzyme that it encodes converts biologically inactive cortisone to the active glucocorticoid cortisol. Cortisol is important in the tissue-specific control of glucose metabolism, blood pressure, and other aspects of cardiovascular function, and prolonged systemic overexposure to cortisol can cause severe metabolic dysfunction, as observed in Cushing’s syndrome. Thus, increased 11ßHSD1 activity has been proposed as a mechanism underlying the association between multiple cardiovascular risk factors, whereas inhibition of 11ßHSD1 has been proposed for the treatment of diabetes and central obesity.8–10 These hypotheses have been supported by experiments in animal models. For example, 11ßHSD1-null mice (which are unable to convert inactive 11-dehydrocorticosterone to active corticosterone, which represent the rodent homologues of human cortisone and cortisol, respectively) have enhanced hepatic and adipose insulin sensitivity and are protected from weight gain and hyperglycemia on a high-fat diet.11–13 By contrast, in transgenic mice overexpressing 11ßHSD1 within white adipose tissue under the AP2 promoter,14 intra-adipose corticosterone levels are 2- to 3-fold higher than in wild-type control mice, and the animals develop visceral obesity, hyperglycemia, and dyslipidemia. Furthermore, mice with overexpression of 11ßHSD1 in the liver under the apolipoprotein E promoter develop hyperinsulinemia and altered fat metabolism.15 Both adipose and hepatic overexpression of 11ßHSD1 results in hypertension, putatively related to increased glucocorticoid-dependent angiotensinogen synthesis in adipose tissue and liver, respectively.15,16 In obese humans, adipose 11ßHSD1 activity and mRNA are reportedly increased in subcutaneous adipose tissue to a similar degree as in the transgenic mice.7,17,18
In recent cross-sectional studies emanating from our group and elsewhere,7,19,20 11ßHSD1 enzyme activity and variability in the 11ßHSD1 gene have been associated with obesity, diabetes, and glucose intolerance. In the current study, longitudinal data collected between 1965 and 2003 were used to investigate the association between variants in the 11ßHSD1 gene and obesity, diabetes, and hypertension. The availability of multiple exams at different ages within the same individual increased statistical power and enabled exploration of gene by environment interaction. Because data were also collected within sibships, family-based analyses were also conducted to control for potential population stratification.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Study Population
The cohort comprised 918 (512 diabetic, 406 nondiabetic) men (n=413) and women (n=505) aged >18 years selected from the ongoing longitudinal study of the etiology of type 2 diabetes mellitus among the Gila River Indian Community in Arizona, where the majority of the residents are Pima or Tohono O’odham Indians.21,22 After their first examination, participants were invited to reattend every 2 years. However, because many participants attended at irregular intervals, data for a given individual may vary in the duration of follow-up and time between exams. All participants provided written, informed consent, and the Tribal Council of the Gila River Indian Community and the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases approved all studies.
Anthropometric Measurements and Metabolic Tests
After an overnight fast, participants underwent a 75-g oral glucose tolerance test for assessment of glucose tolerance according to World Health Organization diagnostic criteria.23 Type 2 diabetes was diagnosed when the 2-hour plasma glucose concentration was 
11.1 mmol/L (200 mg/dL). Standard anthropometric data were collected by trained observers with participants in light clothing and no shoes. Height and weight were measured by using a rigid stadiometer and calibrated scales. Systolic (SBP) and diastolic (DBP) blood pressures were measured in the supine position and on the right arm with an appropriately sized cuff for arm circumference. SBP and DBP were recorded with a mercury-gauge sphygmomanometer to the nearest 2 mm Hg at the first and fourth Korotkoff sounds, respectively. Mean arterial blood pressure (MAP) was computed as 2/3DBP+1/3SBP. Hypertension was defined as a DBP or SBP >90 and >140 mm Hg, respectively, or current use of antihypertensive medication, as determined by participant self-report or review of the medical record.
Genetic Analyses
In an earlier study, we reported 6 sequence variants in the 11ßHSD1 gene.20 However, only the 2 single nucleotide polymorphisms (SNPs) that we report in the current study were sufficiently prevalent for the study of association and/or were not in high linkage disequilibrium. Genotyping was performed by the allelic discrimination TaqMan assay (Applied Biosystems) for rs12086634 (N=918; SNP-5) and by direct sequencing with Big Dye Terminator as described previously for rs846910 (N=758; SNP-1).20 The reason fewer individuals were successfully genotyped for SNP-1 may relate to the structure of genomic DNA in this region. However, given that the means of the phenotypes do not differ significantly between those who were genotyped and those who were not (data not shown) and that the observed allele frequencies within those who were genotyped were highly concordant with the frequencies expected (see results), it is likely that the allelic distribution in those who were not genotyped was random and would not therefore bias the associations we report herein.
Statistical Analysis
All analyses were performed according to the procedures of the Statistical Analysis System software (SAS Institute). Participant characteristics are presented as the arithmetic mean (95% confidence interval [CI]) or when the variable was not of gaussian distribution, as the geometric mean (95% CI). Where necessary to reduce skewness, data were statistically normalized by natural logarithmic transformation. All data were analyzed in their continuous form when possible. The distribution across genotypes of potential confounding variables was assessed by ANOVA (for continuous traits) or a 
2 statistic (for categorical traits). To assess the relation between genotype and trait, a mixed model (PROC MIXED) was fitted that included genotype as a fixed effect along with age, sex, birth date, and "time since first examination"as covariates. All models, except where diabetes was an outcome, included diabetes as an additional covariate, and analyses that included mixed-heritage individuals also included ethnicity.
To account for the fact that many individuals attended >1 examination, the mixed models were fit with an unstructured covariance matrix that allowed for correlation in the trait between different examinations in the same individual. Because few individuals had attended >10 exams and the ability to estimate the variance components in the presence of sparse data are limited, no more than 10 exams were included for a given individual. When an individual had attended >10 exams, we selected 10 exams at random for analyses. Thus, in addition to the fixed effects, the mixed model estimated 55 components of variance (random effects) representing the relation between different exams in the same individual. An additional random effect representing sibship was included to account for the fact that some of the individuals were siblings. For continuous data, the parameters of these variables were estimated by a maximum-likelihood procedure (PROC MIXED). For dichotomous traits, similar models (without the sibship term) were fitted by using generalized estimating equations (PROC GENMOD).
Tests for genotypic association were undertaken assuming an additive effect of the alleles on the phenotype. We chose this model because it is the most robust of the possible models to misspecification of the true model. Additional analyses were conducted to test for interaction between genotype and birth date by fitting interaction terms (genotypexbirth date) to the models described earlier. The significance of these interaction terms was assessed by the likelihood-ratio test comparing the model with the interaction term to one containing only the main effects.
Additional mixed models designed to account for potential confounding by population stratification were computed by exploiting the familial relationships within the data. A modification of the method of Fulker et al24 was used to partition the association into between- and within-family components. In this model, the average number of copies of an allele of interest is determined among all individuals for each sibship. A term representing this sibship average and a term representing the deviation in the number of alleles of interest observed in an individual from this sibship average were both included. The coefficients associated with these terms represent the between- and within-family components, respectively. Although the between-family component is potentially confounded by population stratification,2,3 the within-family component is robust to such confounding and is suitable for use with dichotomous as well as continuous traits.24 The limitation of this method is that it is less statistically powerful than conventional tests of genetic association.
To further control for the potential for confounding by population stratification, we also conducted analyses in the total population, which includes American Indians with mixed ethnic heritage, and a subset of full-heritage Pima and Tohono O’odham Indians. Population-attributable risk for hypertension was calculated from a standard formula.25 Exposure was defined according to the number of minor alleles carried, assuming an additive effect of the allele. Disease was defined according to the definition of hypertension described earlier.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sex was not a significant effect modifier in any model. Therefore, we pooled data and statistically adjusted for sex in all analyses. Table 1 shows adjusted mean data by genotype (SNP-1) for the outcome variables assessed in this study. The minor allele frequencies of SNP-1 and SNP-5 in this study were 0.14 and 0.46, respectively. Both SNPs were in Hardy-Weinberg equilibrium (SNP-1: P=0.90,
2=0.015, df=1; SNP-5: P=0.09, 2=2.84, df=1) and were also in linkage disequilibrium (P<0.0001). We present D' as a measure of allelic association and r2 as a measure of concordance (D'=0.98, r2=0.22). The A allele at SNP-1 was associated with the T allele at SNP-5. In multivariate models, no statistically significant associations or genotypexbirth date interactions were observed for SNP-5 and the metabolic traits of diabetes or obesity (body mass index; Table 2, linear models; Table 3, family-based models). However, SNP-1 was associated with BP and hypertension. SNP-5 was not significantly associated with any metabolic traits (Tables 2 and 3 
). Therefore, the remainder of this section will focus on SNP-1.
|
|
|
Mixed Models for Quantitative Metabolic Traits
In the complete cohort that included mixed-heritage individuals, the adjusted associations of SNP-1 with SBP, DBP, and MAP were statistically significant (DBP ß=1.58 mm Hg per copy of the A allele, P=0.0008; SBP ß=2.28 mm Hg per copy of the A allele, P=0.004; MAP ß=1.83 mm Hg per copy of the A allele, P=0.0006). Birth date modified the association between SNP-1 and BP (tests for interaction: DBP P=0.16; SBP P=0.007; MAP P=0.01), where the magnitude of the association was less in more recent birth cohorts. Figure 1 shows MAP data stratified by decade-specific birth cohorts and by decade-specific age group within each birth cohort. Similar strength associations were observed in both the full-heritage subset and the complete cohort. However, owing to the smaller sample size in the full-heritage subset, the statistical significance of these relations was weaker (DBP ß=1.72 mm Hg per copy of the A allele, P=0.001; SBP ß=2.20 mm Hg per copy of the A allele, P=0.016; MAP ß=1.91 mm Hg per copy of the A allele, P=0.002). As in the complete cohort, there was evidence of birth datexgenotype interaction (DBP P=0.41; SBP P=0.02; MAP P=0.06), where the magnitude of the genetic effect on BP was less in more recent birth cohorts.
|
We did not observe any statistically significant associations between genotype and diabetes in either the complete cohort (odds ratio [OR]=1.11 per copy of the A allele, P=0.38) or the subset of full-heritage Pimas (OR=1.23 per copy of the A allele, P=0.12). Similarly, we observed no evidence of effect modification by birth date for either the complete sample (P=0.32) or the full-heritage subset (P=0.58).
Models With Allowance for Antihypertensive Medication
Next, we undertook 2 approaches to allow for the effects of treatment with antihypertensive medication, which may be more intensive in recent than in earlier birth cohorts, and thus explain the observed birth cohort interactions with genotype. In the first of these models, hypertension was defined as a dichotomous outcome, such that individuals taking medications were considered hypertensive, as described in Methods. In these models, the association between genotype and hypertension was significant in both the complete cohort (OR=1.27 per copy of the A allele, P=0.02) and in the full-heritage subset (OR=1.29 per copy of the A allele, P=0.03). Although the magnitude of effect was greater in earlier birth cohorts, there was no statistically significant interaction between genotype and birth date in either the complete cohort (test for interaction: P=0.19), or in the full-heritage subset (test for interaction: P=0.28).
In the second set of models, we excluded all exams wherein the individual was hypertensive and tested for interaction between genotype and BP. In these models (n=782; exams=3363), we observed significant main effects between SNP-1 and BP (DBP ß=1.23 mm Hg per copy of the A allele, P=0.005; SBP ß=1.27 mm Hg per copy of the a allele, P=0.029; MAP ß=1.16 mm Hg per copy of the a allele, P=0.007). We also observed evidence of birth datexgenotype interactions (tests for interaction: DBP, P=0.27; SBP, P=0.006; MAP, P=0.05), where the effects were strongest in earlier birth cohorts.
Mixed Models Controlling for Population Stratification
We controlled for confounding between genotype and outcome trait by population stratification by analyzing within-family effects (Table 3). Figure 2 illustrates that in 8 informative sibships (ie, those in which a carrier of the minor homozygous genotype and a carrier of the common allele were present), MAP averaged 11.4 mm Hg higher in those with the A/A genotype than in the other siblings. Under additive models for genotype, statistically significant within-family effects were observed for BP (DBP ß=1.77 mm Hg per copy of the A allele, P=0.004; SBP ß=2.03 mm Hg per copy of the A allele, P=0.07; MAP ß=1.86 mm Hg per copy of the A allele, P=0.01) and at a level of borderline statistical significance, for hypertension (OR=1.26 per copy of the A allele; P=0.08). No significant effect was observed for diabetes (OR=1.23 per copy of the A allele; P=0.15) or body mass index (ß=0.37 kg/m2 per copy of the A allele; P=0.63).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In Pima Indians, variability at the 11ßHSD1 gene may influence the risk of hypertension but does not appear to predispose to increased risk of obesity or diabetes. Although the frequency of the less common allele in this study is not high, the magnitude of effect in this group is sufficient to explain a clinically meaningful proportion of the risk of hypertension. For example, if the association is causal, the prevalence of hypertension attributable to the less common allele at SNP-1 or to alleles in tight linkage disequilibria with this allele is
3% in Pima Indians, based on data in this study. Birth date is an important modifier of the association between genotype and BP, where the magnitude of the association becomes progressively weaker in more recent birth cohorts, and this interaction does not rely on treatment with antihypertensive medication. Thus, the risk of hypertension attributable to the minor allele at SNP-1 is far greater in people born in the earlier part of the 20th century by comparison with those born in the latter part of the 20th century. Viewed in conjunction with previous epidemiologic and functional animal data, these findings support the notion that variability in 11ßHSD1 may play a role in human hypertension. Moreover, birth cohort–specific traits, such as early nutrition and birth weight, may be important modifiers of these effects. The 11ßHSD1 polymorphism explored in the present study, SNP-1, is a novel sequence variant in that its effects have not been widely investigated. Owing to its location upstream of the transcription initiation site of the 11ßHSD1 gene variant 1, it may affect the transcription rate of the gene. However, Nair et al20 observed no statistically significant association between 11ßHSD1 genotype (SNP-1 and SNP-5) and levels of 11ßHSD1 mRNA in isolated abdominal subcutaneous adipocyte or muscle (n=60), the latter of which is strongly correlated with 11ßHSD1 enzyme activity.17 Although low statistical power may help explain this lack of association, it is possible that variation in 11ßHSD1 expression is more evident in other tissue, such as pancreas. Alternatively, it may be that neither SNP is functional, and the associations that we observed are attributable to linkage disequilibrium with a hitherto-unknown functional locus.
Typically, genetic associations are assessed from cross-sectional data.1 However, it is probable that BP, like other complex traits, is a function of interactions between genetic and environmental factors.26 This may explain why no evidence of association between 11ßHSD1 genetic variability and hypertension has yet been reported in humans, despite evidence to this effect in animals.15,16 Thus, the successful detection of such genetic effects could depend greatly on the background level of these environment exposures during early development (chronic effects) and at the time of the examination (acute effects). An important aspect of the present analytical method is the use of longitudinal data with multiple observations per individual. These repeated measures not only increase the power of studies of genetic associations with disease but they also offer the opportunity to explore the extent to which genetic effects vary within individuals across time and between birth cohorts, neither of which are generally considered in studies of genetic association.
It is plausible that pathways linking obesity with high BP may involve 11ßHSD1-mediated elevation of glucocorticoid-dependent angiotensinogen production, as observed in transgenic mice overexpressing 11ßHSD1 and other models.13,16,27 Results from murine studies indicate that exposures such as dietary fat intake13 and salt consumption16 may be important factors that interact with the 11ßHSD1 gene to modulate glucocorticoid metabolism. Thus, where dietary salt consumption is excessive, as is typical in industrialized nations, upregulation of the pathways through which 11ßHSD1 modulates metabolic homeostasis could occur. Pima Indians in the present-day environment consume diets high in sugar, whereas several decades ago, the Pima diet may have contained more salt. It is plausible that this change in diet could explain why the detrimental effects of the 11ßHSD1 gene on BP, which are very strong in early birth cohorts, diminish in cohorts born in the latter half of the 20th century. An alternative explanation for this interaction could be that the treatment of hypertension is now more aggressive and effective than previously and that by treating those with hypertension, who are also likely to carry the variant 11ßHSD1 allele, the genetic association with disease is attenuated. However, we also observed modification by birth date of the gene-BP relation in normotensive people who were not taking antihypertensive medication, indicating that other factors may affect the penetrance of the 11ßHSD1 gene. A final explanation, which we were unable to test because of insufficient data, it is that in utero factors could modify expression of the 11ßHSD1 gene, because concomitant with reductions in BP, the treatment of diabetic mothers during pregnancy has improved in this population since the 1960s,28 potentially impacting on intrauterine growth.
A potential limitation of conventional genetic association studies is confounding by population stratification.2,3 If different subpopulations have different risks of disease and different frequencies of the allele in question, then spurious associations (ie, not because of linkage disequilibrium between a marker of a disease susceptibility locus) could be detected. Although the use of multiple examinations for each individual increases the power to detect an association, this method does not preclude the detection of spurious associations owing to population stratification. However, in the current study, statistically significant within-family effects were observed for SNP-1 with MAP and DBP. As these within-family association tests are robust to population stratification, our findings suggest that the association is due, at least in part, to a functional locus in the vicinity of 11ßHSD1, if not within 11ßHSD1 itself. It is noteworthy that although within-family tests are robust to this form of confounding, they are also less powerful than conventional tests that simultaneously assess both within- and between-family effects.24
In a recent study, we reported an association between SNP-1 and SNP-5 with diabetes in the same individuals.20 The findings of the current study are consistent with these earlier observations, inasmuch that they are of the same direction and of similar magnitude (ie, the CIs overlap), but they are not statistically significant. The reason for the difference in statistical significance may relate to the different sample structures and ascertainment of diabetes between studies. In the previous study, a single examination was included for each individual, and this examination coincided with the collation of pedigrees for our genome-wide scan, whereas the current study included several exams per participant, which were collected between 1965 and 2003. In the current study, only diabetes diagnoses that were ascertained in the longitudinal population study were included. However, in the previous study, some diagnoses were included that were ascertained from other research studies.
Perspectives
The current study provides additional evidence that variability at the 11ßHSD1 gene begets the development of hypertension and that during the past 40 years, changes in the environment have modified the effect of 11ßHSD1 on BP in Pima Indians. Our statistical approaches with repeated measures to identify these associations are not often applied in genetic association studies. These novel methods may be more reliable than conventional methods, particularly when observing associations with BP, which is especially prone to a regression dilution effect. Animal and human data illustrate that the expression of many genes, including 11ßHSD1, is modified by environmental factors such as diet, exercise, and smoking. Thus, geneXenvironment interactions such as those reported in the current study for 11ßHSD1 may partially explain why genetic associations with metabolic disease often fail to replicate. Moreover, the identification of genexenvironment interaction may give rise to strategies for targeted prevention.
|
Acknowledgments |
|---|
We thank members of the Gila River Indian Community and other volunteers for participation in this study. We express our gratitude to Sayuko Kobes for her invaluable assistance in organizing the data, to Dr Michal Prochazka for his comments on the manuscript, and to Prof John Lachin for his statistical advice on the mixed models.
Received July 1, 2004; first decision July 21, 2004; accepted August 27, 2004.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Hirschhorn JN, Lohmueller K, Byrne E, Hirschhorn K. A comprehensive review of genetic association studies. Genet Med. 2002; 4: 45–61.[Medline] [Order article via Infotrieve]
2. Knowler WC, Williams RC, Pettitt DJ, Steinberg AG. Gm3;5,13,14 and type 2 diabetes mellitus: an association in American Indians with genetic admixture. Am J Hum Genet. 1980; 43: 520–526.
3. Cardon LR, Palmer LJ. Population stratification and spurious allelic association. Lancet. 2003; 361: 598–604.[CrossRef][Medline] [Order article via Infotrieve]
4. Wareham NJ, Franks PW, Harding AH. Establishing the role of gene-environment interactions in the etiology of type 2 diabetes. Endocrinol Metab Clin North Am. 2002; 31: 553–566.[CrossRef][Medline] [Order article via Infotrieve]
5. Redden DT, Allison DB. Nonreplication in genetic association studies of obesity and diabetes research. J Nutr. 2003; 133: 3323–3326.
6. Olson CM, Rennie D, Cook D, Dickersin K, Flanagin A, Hogan JW, Zhu Q, Reiling J, Pace B. Publication bias in editorial decision making. JAMA. 2002; 287: 2825–2828.
7. Lindsay RS, Wake DJ, Nair S, Bunt J, Livingstone DE, Permana PA, Tataranni PA, Walker BR. Subcutaneous adipose 11ß-hydroxysteroid dehydrogenase type 1 activity and messenger ribonucleic acid levels are associated with adiposity and insulinemia in Pima Indians and Caucasians. J Clin Endocrinol Metab. 2003; 88: 2738–2744.
8. Seckl JR, Walker BR. Minireview: 11ß-hydroxysteroid dehydrogenase type 1, a tissue-specific amplifier of glucocorticoid action. Endocrinology. 2001; 142: 1371–1376.
9. Walker BR, Connacher AA, Lindsay RM, Webb DJ, Edwards CRW. Carbenoxolone increases hepatic insulin sensitivity in man: a novel role for 11-oxosteroid reductase in enhancing glucocorticoid receptor activation. J Clin Endocrinol Metab. 1995; 80: 3155–3159.[Abstract]
10. Bujalska IJ, Kumar S, Stewart PM. Does central obesity reflect ‘Cushing’s disease of the omentum’? Lancet. 1997; 349: 1210–1213.[CrossRef][Medline] [Order article via Infotrieve]
11. Kotelevtsev Y, Holmes MC, Burchell A, Houston PM, Schmoll D, Jamieson P, Best R, Brown R, Edwards CR, Seckl JR, Mullins JJ. 11ß-hydroxysteroid dehydrogenase type 1 knockout mice show attenuated glucocorticoid-inducible responses and resist hyperglycemia on obesity or stress. Proc Natl Acad Sci U S A. 1997; 94: 14924–14929.
12. Morton NM, Holmes MC, Fievet C, Staels B, Tailleux A, Mullins JJ, Seckl JR. Improved lipid and lipoprotein profile, hepatic insulin sensitivity, and glucose tolerance in 11ß-hydroxysteroid dehydrogenase type 1 null mice. J Biol Chem. 2001; 276: 41293–41300.
13. Morton NM, Paterson JM, Masuzaki H, Holmes MC, Staels B, Fievet C, Walker BR, Flier JS, Mullins JJ, Seckl JR. Novel adipose tissue-mediated resistance to diet-induced visceral obesity in 11ß-hydroxysteroid dehydrogenase type 1 deficient mice. Diabetes. 2004; 53: 931–938.
14. Masuzaki H, Paterson J, Shinyama H, Morton NM, Mullins JJ, Seckl JR, Flier JS. A transgenic model of visceral obesity and the metabolic syndrome. Science. 2001; 294: 2166–2170.
15. Paterson JM, Morton NM, Fievet C, Kenyon CJ, Holmes MC, Staels B, Seckl JR, Mullins JJ. Metabolic syndrome without obesity: hepatic overexpression of 11ß-hydroxysteroid dehydrogenase type 1 in transgenic mice. Proc Natl Acad Sci USA. 2004; 101: 7088–7093.
16. Masuzaki H, Yamamoto H, Kenyon CJ, Elmquist JK, Morton NM, Paterson JM, Shinyama H, Sharp MG, Fleming S, Mullins JJ, Seckl JR, Flier JS. Transgenic amplification of glucocorticoid action in adipose tissue causes high blood pressure in mice. J Clin Invest. 2003; 112: 83–90.[CrossRef][Medline] [Order article via Infotrieve]
17. Wake DJ, Rask E, Livingstone DE, Soderberg S, Olsson T, Walker BR. Local and systemic impact of transcriptional upregulation of 11ß-hydroxysteroid dehydrogenase type 1 in adipose tissue in human obesity. J Clin Endocrinol Metab. 2003; 88: 3983–3988.
18. Paulmyer-Lacroix O, Boullu S, Oliver C, Alessi MC, Grino M. Expression of the mRNA coding for 11ß-hydroxysteroid dehydrogenase type 1 in adipose tissue from obese patients: an in situ hybridization study. J Clin Endocrinol Metab. 2003; 87: 2701–2705.
19. Draper N, Echwald SM, Lavery GG, Walker EA, Fraser R, Davies E, Sorensen TIA, Astrup A, Adamski J, Hewison M, Connell JM, Pedersen O, Stewart PM. Association studies between microsatellite markers within the gene encoding human 11ß-hydroxysteroid dehydrogenase type 1 and body mass index, waist to hip ratio, and glucocorticoid metabolism. J Clin Endocrinol Metab. 2002; 87: 4984–4990.
20. Nair S, Lee YH, Lindsay RS, Walker BR, Tataranni PA, Bogardus C, Baier L, Permana PA. The 11ß-hydroxysteroid dehydrogenase type 1 gene: genetic polymorphisms are associated with type 2 diabetes in Pima Indians independently of obesity and expression in adipocyte and muscle. Diabetologia. 2004; 47: 1088–1095.[Medline] [Order article via Infotrieve]
21. Bennett PH, Burch TA, Miller M. Diabetes mellitus in American (Pima) Indians. Lancet. 1971; 2: 125–128.[Medline] [Order article via Infotrieve]
22. Knowler WC, Bennett PH, Hamman RF, Miller M. Diabetes incidence and prevalence in Pima Indians: a 19-fold greater incidence than in Rochester, Minnesota. Am J Epidemiol. 1978; 108: 497–505.
23. World Health Organization: Diabetes Mellitus: Report of a WHO Study Group. Geneva, Switzerland: World Health Organization: 1985. Technical report series No. 727.
24. Fulker DW, Cherny SS, Sham PC, Hewitt JK. Combined linkage and association sib-pair analysis for quantitative traits. Am J Hum Genet. 1999; 64: 259–267.[CrossRef][Medline] [Order article via Infotrieve]
25. Hennekens CH, Buring JE. Epidemiology in Medicine. Boston, Mass: Little Brown & Co; 1987.
26. Franks PW, Bhattacharya S, Luan J, Montague C, Brennard C, Challis B, Brage S, Ekelund U, Middleberg RP, O’Rahilly S, Wareham NJ. Association between physical activity and blood pressure is modified by variants in the G-protein coupled receptor 10. Hypertension. 2004; 43: 224–228.
27. Morton NM, Ramage L, Seckl JR. Down-regulation of adipose 11ß-hydroxysteroid dehydrogenase type 1 by high-fat feeding in mice: a potential adaptive mechanism counteracting metabolic disease. Endocrinology. 2004; 145: 2707–2712.
28. Lindsay RS, Hanson RL, Bennett PH, Knowler WC. Secular trends in birth weight, BMI, and diabetes in the offspring of diabetic mothers. Diabetes Care. 2000; 23: 1249–1254.
This article has been cited by other articles:
 |
|
 |
|
  Y. Liu, D. Mladinov, J. L. Pietrusz, K. Usa, and M. Liang Glucocorticoid response elements and 11{beta}-hydroxysteroid dehydrogenases in the regulation of endothelial nitric oxide synthase expression Cardiovasc Res, January 1, 2009; 81(1): 140 - 147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Liu, R. J. Singh, K. Usa, B. C. Netzel, and M. Liang Renal medullary 11{beta}-hydroxysteroid dehydrogenase type 1 in Dahl salt-sensitive hypertension Physiol Genomics, December 12, 2008; 36(1): 52 - 58. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Vimaleswaran, J. Luan, G. Andersen, Y. L. Muller, E. Wheeler, E. C. Brito, S. O'Rahilly, O. Pedersen, L. J. Baier, W. C. Knowler, et al. The Gly482Ser genotype at the PPARGC1A gene and elevated blood pressure: a meta-analysis involving 13,949 individuals J Appl Physiol, October 1, 2008; 105(4): 1352 - 1358. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P Macfarlane, S. Forbes, and B. R Walker Glucocorticoids and fatty acid metabolism in humans: fuelling fat redistribution in the metabolic syndrome J. Endocrinol., May 1, 2008; 197(2): 189 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R Walker Glucocorticoids and Cardiovascular Disease Eur. J. Endocrinol., November 1, 2007; 157(5): 545 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gambineri, V. Vicennati, S. Genghini, F. Tomassoni, U. Pagotto, R. Pasquali, and B. R. Walker Genetic Variation in 11{beta}-Hydroxysteroid Dehydrogenase Type 1 Predicts Adrenal Hyperandrogenism among Lean Women with Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., June 1, 2006; 91(6): 2295 - 2302. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||