| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Pediatrics (J.C.P., H.S., F.A.T.), Georgia Prevention Institute, Augusta, Georgia; and the Department of Psychiatry (F.A.T.), Medical College of Georgia, Augusta, Georgia.
Address correspondence and reprint requests to Frank A. Treiber, PhD, Medical College of Georgia, Georgia Prevention Institute, Building HS-1640, Augusta, GA 30912. E-mail: ftreiber{at}mcg.edu
| ABSTRACT |
|---|
|
|
|---|
Methods: Four hundred fifty adolescents (mean age = 18.5 ± 2.7 years; 228 [124 males] whites and 222 [110 males] blacks completed two stressors (video game challenge, forehead cold pressor). Hemodynamic measures were taken before, during, and after each stressor. Stressors were separated by a 20-minute rest period.
Results: Frequency of detrimental haplotype (Gly16/Glu27) carrier status was greater among whites than blacks (p < .05). A significant three-way interaction among haplotype, BMI, and race for resting systolic blood pressure (SBP) found the highest BP level to be among high BMI carriers, but only for whites. A separate three-way interaction was found to be significant for haplotype, anger-in and race such that high anger-in carriers showed the highest level of resting SBP (p < .05) and total peripheral resistance (TPR) (p < .05) and the greatest TPR reactivity to the cold pressor task (p < .01). Post hoc analyses revealed these interactions with anger-in were only present among blacks. No significant interactions with anger-out for either ethnic group were observed.
Conclusions: This study demonstrates modulating influences of BMI and anger expression styles on ADRB2 gene associations with hemodynamic function at rest and in response to laboratory stress. These findings support the hypothesis that consideration of geneenvironment interactions may better characterize the role of ADRB2 variation in the development of stress-induced essential hypertension.
Key Words: ADRB2 stress anger cardiovascular reactivity race BMI
Abbreviations: EH = essential hypertension; BP = blood pressure; SBP = systolic blood pressure; DBP = diastolic blood pressure; TPR = total peripheral resistance; CO = cardiac output; ADRB2 = beta-adrenergic receptor; SNP = single nucleotide polymorphism; BMI = body mass index; ET-1 = endothelin 1; AI = anger-in; AO = anger-out; AC = anger control.
| INTRODUCTION |
|---|
|
|
|---|
Exaggerated hemodynamic reactivity to behavioral stress has been shown to be predictive of hypertension (23). Stress exposure protocols may be particularly relevant for studying beta-adrenergic-mediated changes. Two studies have examined the relationship between ADRB2 gene polymorphisms and hemodynamic reactivity to acute behavioral stress (24,25). McCaffery et al. (24) found an association between the Arg16Gly polymorphism and resting BP, but not in BP reactivity to mental arithmetic or the Stroop Color Interference test. Li et al. (25) reported a positive association between the Arg16Gly polymorphism with BP both at rest and in response to mental arithmetic and cold pressor task. Possible explanations for this inconsistency include the differential influence of specific modulating factors and/or inherent differences between the white (24) and European populations (25). These findings further emphasize the need for in-depth consideration of potential geneenvironment interactions in the link between ADRB2 and altered risk for EH.
Associations between anger expression constructs (i.e., anger-out, anger-in, and anger control) and increased risk for EH have been reported (2631). High anger suppression (represented by "anger-in") has been shown to be associated with increased blood pressure at rest (27,28,30,31) and greater blood pressure reactivity to behavioral stress (32). By contrast, some studies have reported "anger-out" (rather than anger-in) to be the toxic component of anger expression, being associated with higher resting BP (29) and increased carotid atherosclerosis (26). Recently, an association between low anger management skills and exaggerated vasoconstrictive reactivity to stress was observed in young adults, but only among subjects who were carriers of a detrimental endothelin-1 (ET-1) allele (32a). In light of this finding, the link between anger expression and cardiovascular reactivity may be explained by changes in specific BP control systems, including alterations in beta-2 receptor-mediated vasodilation.
Because beta-2 adrenergic receptors act on adipocytes to mediate fat breakdown and energy balance, it has been suggested that ADRB2 variants may be relevant to the etiology of obesity-related hypertension (15). BMI may, therefore, be an important factor that interacts with a detrimental ADRB2 variant to predict EH risk. Further, an accumulating body of evidence indicates that adiposity is often associated with increased blood pressure reactivity (3336), supporting the use of cardiovascular stress reactivity for studying potential interactions between ADRB2 variation and adiposity.
A complicating feature of ADRB2 gene association studies is the fact that the Arg16Gly and Gln27Glu polymorphisms occur in specific allele combinations more frequently than would be expected by chance. This linkage disequilibrium has been reported between the Arg16Gly and Gln27Glu loci, as well as with additional ADRB2 SNPs (10). Indeed, Drysdale et al. (10) found that all known ADRB2 SNPs comprise four common haplotypes in blacks and three common haplotypes in whites, suggesting that combinations of ADRB2 SNPs may be more pertinent to BP control than single SNPs. Based on a review of the relevant literature (8,17,37,38), including our own work (39), we contend that the haplotype 16Gly/27Glu (containing both minor alleles) may be particularly detrimental for BP control.
We hypothesized that the coupling of a detrimental ADRB2 haplotype with high BMI or poor anger coping skills would result in exaggerated hemodynamic function at rest and in response to acute laboratory stress. We focused on the outcome variables that best captured ADRB2s role as a peripheral mediator of resistance (i.e., systolic blood pressure [SBP], diastolic blood pressure [DBP], and total peripheral resistance [TPR]). We tested this hypothesis in a multiethnic cohort of young adults who participated in a competitive video game challenge and a forehead cold stressor.
| METHODS |
|---|
|
|
|---|
Protocol
The Institutional Review Board from the Medical College of Georgia approved the study. Informed consent was obtained from each subject, and in the case of subjects <18 years old, parents provided consent. Subjects underwent a battery of anthropometric assessments including height (in centimeters) and weight (in kilograms) according to established protocols. BMI was calculated as kilograms per meters squared and standardized for age and sex (i.e., BMI Z-score). Blood pressure measures were obtained with a Dinamap/Pediatric Vital Signs Monitor (model 1846SX; Critikon). Heart rate and cardiac output (CO) were measured with the thoracic bioimpedance monitor (NCCOM-3, version 6; BoMed Medical Manufacturing Ltd.), and total peripheral resistance was calculated as mean arterial pressure (MAP)/CO (mm Hg/L per minute). MAP was estimated by (SBP + 2 DBP)/3.
After being instrumented with a blood pressure cuff and bioimpedance electrodes, subjects were placed in the supine position for an initial 20-minute rest period. Hemodynamic measurements were collected at minutes 11, 13, and 15. Baseline hemodynamics were calculated as the average of minutes 13 and 15. Reactivity was defined as change scores (peak resting level). Following the first rest period, subjects engaged in a competitive video game challenge (Atari "Breakout") for 10 minutes according to a standardized protocol (41). Hemodynamics were recorded every 2 minutes during the video game challenge as well as during each 20-minute recovery period. After the video game challenge and the postvideo game recovery period, subjects completed a 1-minute forehead cold stressor, "Chilly Willy," according to a previously established procedure (41,42). A large, zip-locking bag filled with six cups of crushed ice and 1.5 cups of water (3°5°C) was placed across the subjects forehead for 1 minute. CV responses were recorded during the final 30 seconds of the stressor. Subjects concluded the reactivity protocol with a final recovery period.
Genotyping
Extraction of genomic DNA from plasma buffy coats or buccal swabs was performed with the QiaAmp DNA Blood Mini Kit (Qiagen). For haplotype analysis, two common SNPs (allele frequency >0.05) were selected for genotyping: G46A (Gly16Arg) and C79G (Gln27Glu). Genotyping was completed by polymerase chain reaction/restriction-enzyme digestion, as described elsewhere (39). Carriers of the detrimental haplotype were identified as those individuals having Gly at AA position 16 and Glu at AA position 27. Haplotype frequencies for ADRB2 SNPs were estimated by PHASE 2.0 (43).
Anger Expression
The Spielberger Anger Expression Scale was used to measure the multidimensional construct of anger expression. This 24-item Likert scale captures three distinct components of anger expression: anger-in (AI) (i.e., suppression or inhibition of anger), anger-out (AO) (i.e., outward expression of anger), and anger control (AC) (i.e., effective anger restraint). Based on general feelings or actions to anger-provoking situations, subjects assigned each item a value ranging from 1 ("almost never") to 4 ("almost always"). Because the literature review indicated that AI and AO were strongly associated with BP control, only those subscales were used in the data analysis. The Spielberger AI and AO subscales have been found to have adequate internal consistency (0.73 < alpha < 0.84) and 1-year testretest reliabilities of 0.52 and 0.75, respectively (44,45).
| RESULTS |
|---|
|
|
|---|
|
|
To avoid inflation of the type 1 error rate owing to the consideration of multiple dependent measures, a multivariate analysis was done using the following terms: SBP, DBP, and TPR at rest; change scores in SBP, DBP, and TPR to the video game stressor; and change scores in SBP, DBP, and TPR to the forehead cold stressor. Haplotype status (carrier or noncarrier of Gly16/Glu27), race, BMI, anger-in, anger-out, and all two- and three-way interactions involving haplotype were entered as independent variables. Gender, age, and baseline hemodynamic values were entered into the model as control variables. This model was reduced by eliminating nonsignificant interactions and main effects not included in significant higher-order interactions. The final multivariate model had main effects for gender, race, BMI, haplotype, AI, two-way interaction terms for haplotype by race, haplotype by BMI, haplotype by AI, race by AI, and three-way interaction terms for haplotype by race by BMI and haplotype by race by AI.
The three-way interaction, haplotype by race by AI, was significant for resting SBP and TPR and for TPR reactivity to forehead cold (all ps < .05). To better interpret the importance of these three-way interactions, we further analyzed the haplotype by AI interaction separately by ethnic group. These post hoc analyses revealed the haplotype by AI interactions in whites to be nonsignificant (all p > .05). In blacks, the haplotype by AI interaction was significant such that carriers of the detrimental haplotype who were also high in AI showed the greatest resting SBP (Fig. 1) and the greatest TPR reactivity to the forehead cold stressor (all ps < .05) (Fig. 2). The haplotype by AI interaction for resting TPR in blacks was not significant (p > .05). The haplotype by BMI by race interaction was significant for resting SBP and DBP (ps < .03). In post hoc analyses, the haplotype by BMI interaction for resting SBP remained significant for whites only and revealed that high BMI carriers showed significantly higher resting SBP compared with low BMI individuals and high BMI noncarriers (p < .001) (Fig. 3). To distinguish between "high" and "low" BMI, we used the values 1 standard deviation above and below the mean, respectively.
|
|
|
Significant main effects and two-way interactions were generally subsumed within significant three-way interactions. The only significant effects not subsumed within these three-ways included: a significant main effect of BMI for resting TPR and video DBP indicating that high BMI individuals had lower TPR at rest and higher DBP reactivity to the video game stressor compared with their low BMI counterparts; a significant haplotype by AI interaction for resting DBP indicating that carriers of the Gly16/Glu27 allele who were the lowest in anger suppression showed the lowest DBP at rest compared with all other groups; and a significant race by AI interaction for SBP reactivity to forehead cold (p < .04) indicating that blacks who were high in AI were more reactive to the ice stressor compared with whites and blacks who were low in AI.
| DISCUSSION |
|---|
|
|
|---|
A key finding from this study was a significant interaction between haplotype carrier status and anger suppression in blacks for resting SBP. Carriers of the detrimental haplotype who reported high levels of anger suppression had significantly higher SBP than carriers with low anger suppression and noncarriers, irrespective of anger suppression status. Interestingly, low anger-suppressing carriers displayed the lowest SBP at rest. That is, this group had lower resting SBP than noncarriers and high anger-suppressing carriers. In subsequent analyses, this same group (low anger-suppressing carriers) displayed lower DBP and TPR at rest. These findings suggest a potential protective function of low anger suppression among carriers of an otherwise harmful haplotype.
An important issue stemming from our findings involves the differential impact of race, anger suppression, and ADRB2 haplotype on TPR reactivity between the two stressors. The forehead cold stressor elicits extremely high vasoconstrictive responses compared with the video game, particularly among blacks (46). Regarding the role of anger on stressor response, we speculate that individuals who are prone to suppress feelings of anger may have done so in response to the pain component of the forehead cold stressor. This may have partially contributed to their vasoconstrictive-mediated stress responses. We further hypothesize that the difference in degree of TPR reactivity between the video game (approximately 2 mm Hg/L per minute) and forehead cold stressor (approximately 56 mm Hg/L per minute) may partially explain the differential impact of the ADRB2 gene on effect size. That is, the compensatory, vasodilatory function of ADRB2 would become more important under conditions of greater vasoconstriction (i.e., the forehead cold stressor). As a result, differences in effect size owing to an individuals physical capacity to vasodilate would be more likely to be observed with the forehead cold stressor. Thus, blacks with high anger suppression who were also carriers of a detrimental ADRB2 haplotype would be expected to show the greatest TPR reactivity, particularly to the forehead cold stressor.
The interaction between ADRB2 haplotype carrier status and BMI on resting SBP in whites is an interesting finding given the mixed literature on the link between ADRB2 alteration and adiposity measures (4749). Notably, we found no statistically significant associations between haplotype carrier status and BMI or additional adiposity-related measures (i.e., waist circumference, waist-to-hip ratio, skinfold thickness). These findings lend support to the viewpoint that ADRB2 is more directly related to BP control than adiposity (50). In whites, the significant interaction between haplotype and BMI indicated that the coincident occurrence of both factors is particularly harmful, although this finding was not replicated in the black subjects.
It is important to note that in both ethnic groups haplotype-specific analyses generated statistically significant findings, although no associations involving the ADRB2 gene at the SNP level reached statistical significance. The two SNPs (Gly16Arg and Gln27Glu) are in fairly high linkage disequilibrium; however, they are not mutually exclusive. Therefore, individuals that are carriers of both "detrimental" alleles of the two SNPs represent a unique subgroup with distinct characteristics (both genetic and environmental) that may result in altered BP control.
These results carry important implications for the primary and secondary prevention of EH. Early behavioral interventions aimed at reducing ineffective anger expression (especially anger suppression) may hold particular value as a preventive measure in blacks with a family history of EH. Similarly in whites, and particularly carriers of the detrimental ADRB2 haplotype, the preemptive use of exercise and dietary interventions in individuals at risk for obesity may decrease the likelihood of future blood pressure dysregulation and EH. In addition to these primary prevention approaches, the prudent use of antihypertensive medications (e.g., the use of beta-1 specific beta blockers or drugs other than beta blockers) in individuals with known genetic and environmental features may increase the efficacy of treatment and reduce the incidence of EH-related sequelae, including stroke, ischemic heart disease, and renal failure.
Although we are excited by the potential implications of this study, the findings must be interpreted cautiously as a result of limitations of the study design. All subjects participating in this study had medically documented family history of cardiovascular disease. Whether these results translate to the general population is unknown. The ADRB2 allele frequencies in our subjects were in line with studies that used random sampling from the population at large, but important differences in BP risk profiles between our subjects and the general population are possible. However, the prevalence of cardiovascular disease, particularly EH, is quite high in the geographical region from which the cohort was recruited (i.e., "stroke belt") (51). Thus, our findings involving these subjects are likely to generalize to other individuals in this region. A second limitation of this study is the cross-sectional design that provides a one-time "snapshot" of each subject. The advantages of using a longitudinal design for determining EH risk profiles have been well established. Longitudinal assessment would help in determining when significant differences between subgroups first appear, allowing for the development of time-sensitive interventions.
Findings from this study indicate that adiposity and anger suppression may play important race-specific roles as modulators of EH risk among carriers of a detrimental ADBR2 haplotype. The physiological pathways by which these interactive effects may occur are largely unknown. Mills and Dimsdale have suggested a potential beta-adrenergic receptor-mediated mechanism linking anger suppression and increased likelihood for cardiovascular disease (32). They found high beta-adrenergic receptor sensitivity and increased SBP reactivity to a math stressor among subjects who routinely suppressed their anger. Altered beta-2 adrenergic sensitivity has been linked to carriers of the Gly16 and Glu27 alleles (5,6,8). Both in vitro and in vivo work (in animal models and humans) is necessary to determine the extent to which inherited ADRB2 dysfunction affects overall risk for hypertension and how the additional factors of BMI, anger, and race affect this relationship. Additional evaluation of these relationships within a longitudinal cohort framework would go far in elucidating the importance of these factors in the etiology of EH.
| NOTES |
|---|
|
|
|---|
This work was supported in part by grant HL 69999 from the National Heart, Lung, and Blood Institute.
DOI:10.1097/01.psy.0000204925.18143.4f
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |