Monday, May 24, 2004
Institution: CTRO HOS COIMBRA HOSP GRAL Sign In as Member/Individual (Non-Member) Subscription Activation
Abstract of this Article ( )
Email this article to a friend
Similar articles found in: Hypertension Online PubMed
This Article has been cited by:
other online articles
Search PubMed for articles by: Schneider, R. H. || King, C. G.
Alert me when: new articles cite this article
Download to Citation Manager
© 1995 American Heart Association, Inc.
A Randomized Controlled Trial of Stress Reduction for Hypertension in Older African Americans
Robert H. Schneider; Frank Staggers; Charles N. Alexander; William Sheppard; Maxwell Rainforth; Kofi Kondwani; Sandra Smith; Carolyn Gaylord King
From the Center for Health and Aging Studies, Department of Physiological and Biological Sciences (R.H.S.) and Department of Psychology (C.N.A., M.R., K.K., C.G.K.), Maharishi University of Management, Fairfield, Iowa; the Hypertension and Stress Management Research Clinic, West Oakland Health Center, Oakland, Calif (F.S., W.S., K.K., S.S.); the Department of Social and Behavioral Sciences, University of Arkansas, Pine Bluff (C.G.K.); and the Haight-Ashbury Free Medical Clinic, San Francisco, Calif (F.S.).
Correspondence to Robert H. Schneider, MD, Center for Health and Aging Studies, Maharishi University of Management FB 1028, Fairfield, IA 52557-1028.
Top Abstract Introduction Methods Results Discussion References
Abstract We tested the short-term efficacy and feasibility of two stress education approaches to the treatment of mild hypertension in older African Americans. This was a randomized, controlled, single-blind trial with 3 months of follow-up in a primary care, inner-city health center. Of 213 African American men and women screened, 127 individuals (aged 55 to 85 years with initial diastolic pressure of 90 to 109 mm Hg, systolic pressure of 189 mm Hg, and final baseline blood pressure of 179/104 mm Hg) were selected. Of these, 16 did not complete follow-up blood pressure measurements. Mental and physical stress-reduction approaches (Transcendental Meditation and progressive muscle relaxation) were compared with a lifestyle modification education control program and with each other. The primary outcome measures were changes in clinic diastolic and systolic pressures from baseline to final follow-up, measured by blinded observers. The secondary measures were linear blood pressure trends, changes in home blood pressure, and intervention compliance. Adjusted for significant baseline differences and compared with control, Transcendental Meditation reduced systolic pressure by 10.7 mm Hg (P<.0003) and diastolic pressure by 6.4 mm Hg (P<.00005). Progressive muscle relaxation lowered systolic pressure by 4.7 mm Hg (P=.054) and diastolic pressure by 3.3 mm Hg (P<.02). The reductions in the Transcendental Meditation group were significantly greater than in the progressive muscle relaxation group for both systolic blood pressure (P=.02) and diastolic blood pressure (P=.03). Linear trend analysis confirmed these patterns. Compliance was high in both stress-reduction groups. Home systolic but not diastolic pressure changes were similar to clinic changes. Selected mental and physical stress-reduction techniques demonstrated efficacy in reducing mild hypertension in this sample of older African Americans. Of the two techniques Transcendental Meditation was approximately twice as effective as progressive muscle relaxation. Long-term effects and generalizability to other populations require further evaluation.
Key Words: hypertension, stress • relaxation • blacks • aged
Top Abstract Introduction Methods Results Discussion References
Age- and gender-specific mortality rates for African Americans are 50% higher than for white Americans.1 2 Disproportionately high rates of hypertension contribute to the excess rates of mortality and morbidity from cardiovascular and renal disease in this group.2 Hypertension in African Americans has a higher prevalence, incidence, and severity; earlier onset; more target-organ damage; and is generally treated later and less adequately than in white Americans.3 Furthermore, more than 70% of elderly African Americans are hypertensive, and hypertensive diseases cause four to seven times more mortality than in older white Americans.4 5 For these reasons hypertension has been considered the number one health problem among adult African Americans.5 6
Numerous controlled clinical trials have demonstrated that lowering blood pressure (BP) reduces morbidity and mortality in the general population7 and in African Americans in particular.8 However, the efficacy of conventional drug therapy in preventing the most frequent complication of hypertension—coronary heart disease—is substantially less than predicted.9 This may be due to adverse side effects of antihypertensive drug therapy.10 11 In addition, antihypertensive pharmacotherapy has been associated with impaired quality of life,12 low compliance,13 and high cost.14 15 These limitations of conventional drug therapy have been noted in the elderly16 and African American17 patient populations. For these and other reasons the Fifth Joint National Committee (JNC V), National High Blood Pressure Education Program, and the Working Group on Hypertension in the Elderly recommended that nondrug lifestyle modification approaches be used as first-line definitive or adjunctive treatment for hypertension, particularly in African Americans with high BP.
Chronic psychosocial stress has been implicated in the etiology of hypertension.6 18 19 In African Americans socioenvironmental and psychosocial stress have been associated with higher BP.20 Increasing social and economic disparities in later life may further increase psychosocial stress and the risk for hypertension in older African Americans.5 Heightened cardiovascular and sympathetic nervous system reactivity may be a mechanism for the stress-hypertension link in this population.20 21 Therefore, stress reduction may be useful for treating hypertension in older African Americans, yet there has been a lack of controlled clinical trials of stress reduction for hypertension in this group.22 However, there have been numerous clinical studies of stress-reduction approaches for hypertension in nonminority populations, with inconsistent results.23 24 These studies have generally suffered from serious methodological weaknesses, including inadequate sample sizes and baseline determinations; absence of appropriate control groups; possible confounding by expectancy or other cointerventions, such as changes in antihypertensive medication; and lack of comparison or differentiation of distinct approaches to stress reduction.24 25
The question of heterogeneity of effects of different approaches to stress management has been investigated in several quantitative meta-analyses which suggest that the Transcendental Meditation (TM) technique, a mental technique for stress reduction,26 may be particularly effective in reducing physiological arousal,27 anxiety,28 and smoking, alcohol, and drug abuse29 and improving psychological health.30 Previous studies on TM and BP in nonminority samples have suggested significant antihypertensive effects of the program.31 32 33 Progressive muscle relaxation (PMR),34 a widely used physical-based approach, has also been used for reducing psychological stress28 and BP.23 24 25 Both TM and PMR have been previously studied in young adult, normotensive African Americans.35
Therefore, the objectives of the present study were to conduct a well-controlled clinical trial comparing mental (TM) and physical-based (PMR) techniques for stress reduction in the treatment of mild hypertension in a community-based sample of older African Americans to assess the short-term efficacy of these techniques in reducing hypertension and determine the feasibility of longer-term implementation of these stress-reduction programs in primary care settings with this patient population.
Top Abstract Introduction Methods Results Discussion References
The design consisted of a randomized, controlled, single-blind, primary care center–based trial that included three study groups: the TM program, PMR, and a lifestyle modification education control (EC) program. The two active stress-management conditions (TM and PMR) were closely matched for external characteristics. The EC group partially controlled for treatment attention, time, and monitoring. Eligible participants underwent a baseline period of monitoring and then were randomly assigned to one of the study groups for 3 months of intervention. The primary outcome variables were changes in clinic systolic and diastolic BP values measured blindly. The secondary outcomes were changes in self-monitored home BP and compliance. Psychological and behavioral characteristics were assessed at baseline to determine the equivalence of groups. The study was carried out between January 1989 and June 1991 at the West Oakland Health Center, a primary care community health center in Oakland, Calif.
Eligibility and Randomization
The target population consisted of men and women, self-identified as African American, aged 55 years or older, with a history of mild hypertension. Subjects were recruited from local community clinics, senior citizen centers, and other community organizations and were reimbursed $6 per visit for travel expenses to the clinic. Initial BP eligibility criteria were 90 to 109 mm Hg diastolic BP and less than or equal to 189 mm Hg systolic BP based on three successive measurements at the initial screening visit. Subjects were eligible for entry into the trial whether or not they were taking antihypertensive medications as long as they met the BP eligibility criteria noted above. If potential subjects were taking BP medications and agreed to discontinue them, they were tapered off BP drugs by their primary physicians with a washout period of 4 to 8 weeks before entry into the baseline. BP medication dosages during the baseline and treatment periods were kept stable by participant and physician consent. Candidates were excluded if they had medical evidence of life-threatening or disabling diseases. All subjects participated with the approval of their primary physicians and gave informed consent. The study was approved by institutional review boards at the West Oakland Health Center and Maharishi University of Management and followed institutional guidelines at the collaborating centers.
After the initial screening visit and verification of eligibility, subjects returned to the study clinic every 1 to 2 weeks for an average of four baseline visits. At these subsequent baseline visits BP measurement and psychometric testing were performed. Final baseline BP level was based on three measurements at each of the last two baseline visits. After completion of baseline, subjects were randomly assigned to one of the treatment groups according to a number adaptive allocation model.36 37 This procedure was used because it allowed for adaptation of the ratio of experimental to control subjects based on the availability of eligible participants while maintaining randomization to the three study groups. If participants’ BP exceeded 104 mm Hg diastolic or 179 mm Hg systolic at any two successive visits, they were excluded from the trial and referred for standard care.
Data for all participants were subsequently collected monthly for 3 months. The final 3-month follow-up determinations were measured at a series of two visits 1 week apart. Clinic BP was measured by a research technician who was blinded to the treatment status of the subjects. All clinic BPs were measured with a stationary automated BP monitor with digital readout (model 300S, Vitastat Medical Services) that was calibrated against a mercury sphygmomanometer at regular intervals.38 Readings were taken with subjects in the seated position after they had sat for 5 minutes at rest while not practicing any stylized relaxation technique. Three readings using the first and fifth Korotkoff sounds were recorded at each visit, with the last two averaged to give the final reading for that visit.
For the home BP determinations a standard self-monitoring procedure was used that had been previously studied for validity and reliability.18 39 Subjects were taught to record their own BP by a trained research technician using a semiautomatic, auscultatory device (model UA 731, Takeda Medical, Inc). A subject was considered proficient when his or her readings were within 5 mm Hg of the technician’s readings. The home BP monitoring equipment was periodically checked by the study staff for accuracy and proper operation. Participants recorded their home BP twice a day (in the morning and afternoon or evening) for 1 week at the end of the baseline and intervention phases.
A battery of psychosocial and behavioral assessments was administered verbally to each participant individually by a trained interviewer. These measures included (1) the National Survey of Black Americans subscales on personal efficacy, stress impact, and social support40 ; (2) State-Trait Personality Inventory for trait anxiety and anger subscales41 ; (3) Multidimensional Health Locus of Control Scale for the internal control subscale42 ; (4) self-esteem40 ; (5) well-being43 ; (6) Nottingham Health Profile44 ; (7) Generalized Outcome Expectancy45 ; (8) expectancy of outcome for these specific treatments31 ; and (9) health habit questionnaires for exercise and diet.46 Compliance with the active interventions was determined by a regularity questionnaire assessing frequency of practice of the stress-reduction technique, which the participants completed at monthly intervals during the treatment phase.47 Perceived efficacy of the treatments was determined by questionnaire during the final posttest session.48
The two active stress-reduction conditions, TM and PMR, were matched to each other for teaching format, instructional time, home practice requirements, and expectancy of beneficial outcomes based on the standard TM course format.26 Neither active stress-management intervention required any change in personal beliefs, philosophy, or lifestyle other than daily practice. Instruction in both active interventions included an introductory presentation and discussion, brief personal interview, personal instruction meeting, and three follow-up small group seminars. The instructional meetings lasted about 1.5 hours each and took place over the course of 1 week. Thereafter, each stress-reduction group met for a 1.5-hour session every month. Each of the instructors for the active interventions was African American and professionally qualified and experienced in teaching either PMR or TM. Participants were instructed to practice their respective techniques for 20 minutes twice daily (morning and evening) while seated comfortably with eyes closed. They were also requested not to reveal details of their program to individuals outside their treatment group.
The distinctive features of the three treatments were as follows: The TM program is the principal approach for stress management and self-development of Maharishi Ayur-Veda, a system of natural health care derived from the ancient Vedic approach to health by Maharishi Mahesh Yogi.49 50 The TM technique has been described as a simple yet precise mental technique whereby the ordinary thinking process becomes quiescent and a distinctive psychophysiological state of restful alertness appears to be gained.50 51 Details of the instructional protocol and practice have been previously described.26
The PMR technique followed the previously published procedure of Bernstein and Borkovic34 that is based on Jacobson’s classic muscle relaxation program.52 This physical relaxation technique involves directing the participants’ attention to tensing and relaxing the various muscle groups throughout the body systematically to achieve deep relaxation.
The partial attention control, termed lifestyle modification EC, included a set of educational instructions and materials modeled after the usual community practice recommendations for the nondrug management of mild hypertension. These included specific guidelines for reduction of dietary sodium and caloric intake as well as aerobic exercise based on the lifestyle modification recommendations of the Joint National Committee.6 Participants in this group met with the treatment provider for individual or small group sessions once every month for 0.5 to 1 hour during the treatment phase. Participants in this group were given expectations that their BP could be reduced by adoption of these lifestyle modifications.
Baseline characteristics of the three groups were compared by MANOVA and univariate ANOVA. The baseline factors included age, sex, weight, BP, medication status, and psychological and behavioral/lifestyle characteristics.
Treatment outcomes were assessed by ANCOVA, with BP change as the outcome variable and baseline BP and other baseline characteristics that significantly differed between the groups as covariates. Change in BP was defined as 3-month posttreatment BP minus baseline BP. Significance was set at a value of P<.05.
Planned contrasts allowed pairwise comparisons of the three treatment groups on BP outcomes. These contrasts were one tailed because of the directionality of predictions.53 On the basis of previous research discussed above, it was hypothesized that both active interventions would be superior to the EC program and that the TM group would show greater reductions in BP than the comparison relaxation program.29 31 32 33 For analysis of clinic BP trends over time, each of the three monthly follow-up visits was included in a repeated-measures ANCOVA (linear trend analysis). In addition, analyses by intention-to-treat37 were performed by two methods. First, missing BP change scores were treated as missing at random, and the BMDP 5V routine was applied for estimation of their values.54 Second, a more conservative approach was applied by assignment of the maximal increase in BP observed in any subject to all the missing values for all subjects. Effects of the interventions on psychological and behavioral variables will be reported in a future publication on quality of life.
Top Abstract Introduction Methods Results Discussion References
Recruitment and Attrition
Over an 18-month period 213 people were screened for eligibility for the study. Of these, 127 were randomized to treatment. The most common reason for exclusion of candidates during the baseline period was BP outside the study range. Of the 127 participants, 16 attrited over the course of the 3-month intervention period. Of those who attrited, 11 (2 TM, 3 PMR, 6 EC) were eliminated because of changes in their antihypertensive medication by their primary care physician. The remaining 5 subjects (2 TM, 2 PMR, 1 EC) withdrew for a variety of personal reasons, most commonly change of residence. Thus, 111 subjects completed the study.
Table 1 presents baseline demographic and BP characteristics of participants across randomized groups. The mean age was 67 years; 57% were female and 43% male. Fifty percent were not taking antihypertensive medications. Mean weight was 82 kg. Mean BP was 147/92 mm Hg. MANOVA indicated no significant difference between groups on baseline characteristics. Among the 19 baseline physiological and psychological characteristics assessed by univariate analysis, only age was significantly different between groups (TM, 64 years; PMR, 69 years; and EC, 67 years; P<.01).
View this table:[in this window][in a new window] Table 1. Baseline Characteristics for Stress-Reduction and Control Groups
Compliance Rates and Expectancy
Compliance with the active stress-reduction interventions was high. Monthly reports by the participants indicated that 97.1% of the TM group and 81.1% of the PMR group practiced their techniques "twice a day" or "almost twice every day." There was no significant difference between the active groups on compliance rates by a test of proportion. Participants in both the TM and PMR groups rated their instructors "excellent" (mean was 3.9 and 3.92, respectively, for TM and PMR instructors, with 4=excellent and 3=good). There were no differences between the active groups on the measures of outcome expectancy.
Clinic BP Results
Table 2 presents for each group (total n=111) adjusted change scores, with age and baseline BP as covariates. Both active intervention groups showed significant reductions in systolic and diastolic BP values compared with the lifestyle modification EC group. Compared with the EC group, the TM group showed an adjusted reduction of 10.7 mm Hg in systolic BP (P=.0002) and 6.4 mm Hg diastolic BP (P=.00005). Compared with the EC group, the PMR group showed adjusted reductions of 4.7 mm Hg in systolic BP (P=.054) and 3.3 mm Hg in diastolic BP (P=.02). The reductions in the TM group were significantly larger than in the PMR group for systolic BP (P=.02) and diastolic BP (P=.03). Clinic heart rates did not change significantly in any of the groups.
View this table:[in this window][in a new window] Table 2. Change From Baseline in Clinic BP for Stress-Reduction and Control Groups
The results of the intention-to-treat analyses did not substantially differ from the primary analyses above. Moreover, when dropouts from the study were compared with the continuing participants, there were no significant differences in baseline BP or demographic, lifestyle, or psychological characteristics.
Figs 1 and 2 display the systolic and diastolic BP changes adjusted for baseline BP and age for the 104 participants, with complete data for each of the three monthly follow-up visits. Repeated-measures ANCOVA confirmed significant reductions in both the TM and PMR groups compared with the EC group in systolic and diastolic BP values (TM, P<.0005 for systolic BP and P<.0001 for diastolic BP; PMR, P<.025 for both systolic BP and diastolic BP). Linear trend analysis for the repeated-measures ANCOVA showed there was a consistently greater downward linear trend over the 3 months in the TM group compared with the EC group for both systolic BP (P<.005) and diastolic BP (P<.01) but no significant trend differences for the PMR group compared with the EC group.
View larger version (18K):[in this window][in a new window] Figure 1. Line graph shows mean changes in clinic systolic pressure over 3 months (follow-up minus baseline) with SEM. Probability values are for repeated-measures ANCOVA comparing each experimental group (TM and PMR) with control (EC). TM indicates Transcendental Meditation (n=36); PMR, progressive muscle relaxation (n=33); and EC, lifestyle modification education control (n=35).
View larger version (16K):[in this window][in a new window] Figure 2. Line graph shows mean changes in clinic diastolic pressure over 3 months (follow-up minus baseline) with SEM. Probability values are for repeated-measures ANCOVA comparing each experimental group (TM and PMR) with control (EC). TM indicates Transcendental Meditation (n=36); PMR, progressive muscle relaxation (n=33); and EC, lifestyle modification education control (n=35).
Home BP Results
The results of home BP monitoring were calculated from an average of 14 readings during 7 days for the 92 participants with home BP data. Nineteen subjects did not comply with home BP recording and thus were not included in this analysis. Both active treatment groups demonstrated reductions in systolic BP compared with the lifestyle modification EC group. The TM group showed an adjusted reduction of 9.6 mm Hg (P=.0004) and the PMR group a reduction of 4.3 mm Hg (P=NS). The reduction in the TM group was significantly greater than in the PMR group (P<.05). Neither active intervention group demonstrated significant reductions in home diastolic BP.
BP Results in Medicated and Nonmedicated Subgroups
Although age was the only baseline characteristic that significantly differed between groups, inspection of baseline characteristics in Table 1 suggested the potential for other confounding differences between groups (eg, medication status, sex, and weight). Sex, weight, and medication status were then evaluated as additional covariates to the primary analysis, with no appreciable change in the results. We also examined whether baseline characteristics interacted with the BP outcomes. The only significant interaction was between medication status and diastolic BP (P=.02). To further evaluate the effects of medication status on BP change, we analyzed post hoc the medication and nonmedication subgroups for baseline characteristics and responses to treatment. Baseline BP values did not differ significantly between the medicated and nonmedicated subgroups. Within the nonmedicated subgroups the TM participants showed a reduction in systolic BP of 13.0 mm Hg, the PMR participants showed a reduction of 3.8 mm Hg, and the EC group showed an increase of 5.2 mm Hg. Within the medicated subgroups the TM participants showed a reduction in systolic BP of 7.9 mm Hg, the PMR participants showed a reduction of 6.0 mm Hg, and the EC group showed a decrease of 3.0 mm Hg. Regarding diastolic BP, in the nonmedicated subgroups there were reductions of 6.6 mm Hg in the TM group and 2.3 mm Hg in the PMR group and an increase of 4.7 mm Hg in the EC group. In the subgroup continuing to take antihypertensive medications, clinic diastolic BP was reduced 4.1 mm Hg in the TM group, 2.5 mm Hg in the PMR group, and 1.2 mm Hg in the EC group. Also, there were no significant changes in weight, self-reported exercise, or self-reported diet when the groups were compared.
Top Abstract Introduction Methods Results Discussion References
This trial demonstrated the feasibility and short-term efficacy of using selected stress-reduction approaches in the treatment of mild hypertension in older African Americans in an inner-city primary care health center setting. After 3 months of follow-up both TM and PMR significantly reduced systolic and diastolic BP values compared with a control program. Furthermore, TM practice was associated with BP reductions that were approximately twice as large as those with PMR. Compliance with both active interventions was relatively high. These effects were clinically and statistically significant, as discussed below.
This study met the design characteristics demanded for a well-controlled trial of behavioral stress-reduction approaches for hypertension24 25 : (1) The baseline period of four visits minimized reductions in BP during the treatment period caused by regression to the mean or habituation effects. (2) Randomization minimized, although it did not entirely eliminate, initial differences between the groups. (3) BP and other study data were collected in a single-blind design. Although the study participants were aware of the treatment they received, there were no differences between the groups in expectancy of beneficial outcome from the study. (4) Assessment at multiple follow-up visits (ie, monthly) demonstrated a consistent pattern of BP reductions. (5) The power was large enough to detect clinically significant BP differences between the groups. (6) The project was conducted in a primary care health center in an urban African American community, which enhances generalizability of findings to community settings. (7) Potential confounders or cointerventions such as changes in weight, self-reported exercise, diet, or antihypertensive medications were methodologically or statistically controlled.
This was one of the few randomized controlled trials to compare the relative efficacy of mental and physical-based techniques for stress reduction in hypertensive individuals within the same experimental setting25 55 and to our knowledge the only one with older African Americans. In addition, the utilization of two behavioral interventions in the same trial allowed each to serve as an active control for the other. Both TM and PMR engendered similar expectancies of positive outcome and required the same frequency and duration of instructional attention and daily home practice. Both interventions were taught by African American instructors from the community who were professionally trained and enthusiastic about their programs. Each was well received by the participants, as evidenced by high teacher ratings and compliance rates. Thus, the differences in BP-lowering effects between the two stress-reduction groups observed in this study were likely due to differences in efficacy between the two approaches rather than to nonspecific factors such as a placebo effect.
Therefore, these findings support the suggestion of heterogeneity of stress-reduction approaches; that is, different techniques produce different results.25 29 These findings do not support the conclusion of a recent trial that stress-reduction approaches in general lack efficacy in reducing BP.23 The Trial of Hypertension Prevention (TOHP),23 which used a multimodality stress-management program (including progressive relaxation) with high normal BP subjects, found no significant changes in BP compared with controls. Interestingly, in TOHP the stress-management intervention was associated with increased self-reported stress, which may explain its lack of efficacy in reducing BP. Also, two recent quantitative meta-analyses24 25 and the qualitative review of JNC V6 concluded that when clinical trials were well controlled, stress-reduction therapies generally showed little or no effect on lowering BP. However, none of the trials reported in these reviews included the TM program. In contrast, the current well-controlled trial, specifically comparing the efficacy of the TM technique with that of a widely used physical approach to relaxation, clearly showed their differential effectiveness in BP reduction over the short term.
The differential effectiveness of TM compared with PMR is consistent with the findings of prior meta-analyses in which the effects of TM practice on health-related outcomes, including mental and behavioral health (eg, arousal, anxiety, and smoking), were on average twice as large as other stress-reduction techniques, including PMR,28 30 56 and with a previous study of TM and relaxation in young adult African Americans.35 The results found with the TM program in this trial with hypertensive African Americans were similar in magnitude to results of other trials of TM in predominantly white samples.33 For example, in a randomized controlled trial in an elderly white sample, similar differential effects between TM and relaxation on systolic BP were observed.31
The reduction of 11/6 mm Hg with the TM program in this relatively short-term trial is also similar to the average reduction in BP reported in a meta-analysis of controlled trials of antihypertensive drug therapy of several years’ duration.9 If the results seen in this short-term feasibility study were maintained over at least 2 to 3 years, reductions in associated morbidity and mortality might be observed. For example, a meta-analysis of randomized trials of antihypertensive treatment in the elderly, with BP reductions similar to the present trial, found that cardiovascular mortality decreased 22%, stroke mortality 33%, and coronary mortality 26%.57
The finding of similar reductions in the clinic and at home for systolic BP in both active intervention groups suggests that TM and PMR produced effects that generalized to the participants’ natural environment.58 59 The lack of significant change in home diastolic BP may have been due to the relative insensitivity of the electronic home monitoring device to diastolic BP or other limitations of the procedure. Home BP monitoring has been reported to systematically underestimate changes compared with clinic BP, especially diastolic BP.39 Further investigation into out-of-clinic BP patterns with ambulatory BP monitoring may clarify this question.
Baseline differences between groups did not appear to account for the primary results of the study. First, all results were adjusted for age, the only statistically significant baseline difference between the groups. A second analysis also adjusted for the nonsignificant differences in weight, sex, and medication status, with no appreciable change in the results. Third, although there was a slightly larger proportion of medicated subjects in the control group, this could not account for the response differences between the groups because post hoc analysis indicated that medicated control participants’ BP decreased more during the trial than their nonmedicated counterparts. Therefore, the overrepresentation of medicated subjects in the EC group would have tended to decrease the differences in BP reduction between the active and control groups. Thus, the primary results may actually be more conservative than otherwise. Further studies comparing the antihypertensive responses of medicated and nonmedicated groups with larger sample sizes designed for this purpose may help to further clarify this question.60 Finally, the two active treatment groups were generally similar to each other in baseline and demand characteristics of the intervention. Yet comparison of these active groups to each other on outcomes indicated that the TM subjects showed BP reductions about twice the magnitude of those of the PMR group despite their close matching.
The results with the active interventions of this trial, particularly TM, compare favorably with those of other lifestyle modification approaches in average BP reductions reported for weight loss, dietary sodium restriction, and aerobic exercise. For example, MacMahon et al61 summarized five randomized trials of weight reduction in hypertensive patients and reported that a mean weight loss of 9.2 kg (20 lb) was associated with a 6.3/3.1 mm Hg reduction in BP. In an overview of sodium reduction trials and BP, Cutler et al62 pooled results of 18 trials and found a weighted average decrease of 4.9/2.6 mm Hg in hypertensive subjects. Exercise training in hypertensive individuals has been associated with an average reduction of 5 to 6 mm Hg in systolic BP.63
Regarding the potential mechanisms for the effects observed in this study, cross-sectional, case-control, and prospective studies in a variety of populations provide evidence for a role of chronic stress in the development, maintenance, and progression of hypertension that persists after controlling for age, weight, sodium and alcohol intakes; physical activity; and family history of hypertension.18 19 64 Proposed physiological mechanisms for the etiologic link between stress and hypertension include excessive sympathetic nervous system activation and cardiovascular reactivity.65 66 These may be coupled with other neuroendocrine alterations involving the hypothalamic-pituitary axis. In African Americans there is evidence that disproportionately high levels of psychosocial and environmental stress are associated with increased sympathetic tone, cardiovascular reactivity, peripheral vasoconstriction, and renally mediated sodium retention.67 It may be that the cumulative effects of stress also contribute to the age-related increases in BP in African Americans that are not observed in traditional African societies.5 68 Although there is evidence for reduced sympathetic activation in TM practitioners,33 69 it has been proposed that the practice of TM may lower high BP through an integrated set of adaptive responses involving cortical, autonomic, neuroendocrine, and cardiovascular systems. These mechanisms may be part of an integrated neurophysiological homeostatic response.29 33 51
Although JNC V acknowledged that stress may contribute to the etiology of hypertension and that stress management is an "appealing concept," it concluded that the role of stress-reduction techniques in the treatment of hypertension has not yet been adequately demonstrated.6 However, the JNC did not report findings from previous studies including the TM program,31 32 33 nor did they have access to the results of the present trial. Furthermore, as discussed above, recent findings do not support the homogeneity assumption of earlier reviews but instead indicate heterogeneous responses for different stress-reduction techniques.28 29 30 31 56
The use of an effective stress-reduction technique such as that described in this report either alone or in combination with other lifestyle modifications suggested by JNC could contribute several distinct advantages for clinical care. These might include reduced adverse side effects, improved quality of life, enhanced compliance, reduced health-care costs over the long term, and potential improvements in morbidity and mortality from cardiovascular diseases.29 31 70 The latter hypothesis is supported by reports that TM has been associated with substantially lower morbidity from heart disease and lower all-cause mortality.31 70 This may be related in part to simultaneous improvements in other cardiovascular risk factors, such as smoking, serum cholesterol, and alcohol abuse.29 56 71
The approach of prevention and health promotion with behavioral stress reduction is concordant with cultural theories in African American psychology72 and recommendations for health empowerment in African American individuals and communities.73 This approach is also consistent with current recommendations for cost-effective prevention and health promotion.74 75
In conclusion, the results from the present short-term study suggest that stress reduction, particularly with the TM program, may be feasible and efficacious in treating mild hypertension in older African Americans. The findings of the current trial require confirmation from future controlled trials with longer-term intervention periods, larger subgroup sizes (for evaluation of the interaction of medication and treatment effects), and diverse ethnic samples to determine generalizability to other populations.
This work was supported by grants from the Retirement Research Foundation (No. 88-95 and No. 88-96), Chicago, Ill, and the Lancaster Foundation, Bethesda, Md. Preparation of this manuscript was supported in part by National Institutes of Health grant 5R01HL-48107. The authors are grateful to the following individuals for their helpful advice and/or editorial comments: Brian Hofland, PhD; Paul Gelderloos, PhD; Norman Anderson, PhD; R. Keith Wallace, PhD; David Orme-Johnson, PhD; Robert Cooper, MD; and for technical assistance, Bruce Smith, Linda Heaton, and John Salerno, PhD.
Preliminary portions of this study were presented at the Third International Conference on Race, Ethnicity and Health, Salvador, Bahia, Brazil, July 1991, and at the 52nd Annual Scientific Meeting of the American Psychosomatic Society, Boston, Mass, April 13-16, 1994.
Received December 21, 1994; first decision January 20, 1995; accepted July 10, 1995.
Top Abstract Introduction Methods Results Discussion References
1. Hale C. A demographic profile of African Americans. In: Taylor S, ed. Health Issues in the Black Community. San Francisco, Calif: Jossey-Bass Inc Publishers; 1992:6-19.
2. US Department of Health and Human Services. Report of the Task Force on Black and Minority Health, Volume IV: Cardiovascular and Cerebrovascular Disease, Part 2. Washington, DC: US Government Printing Office; 1986:229-284.
3. Saunders E. Cardiovascular Diseases in Blacks. Philadelphia, Pa: FA Davis Co; 1991:83-192.
4. Gillum R. Cardiovascular diseases in the United States: an epidemiological overview. In: Saunders E, ed. Cardiovascular Diseases in Blacks. Philadelphia, Pa: FA Davis Co; 1991:3-14.
5. Anderson NB. Aging and hypertension among blacks: a multidimensional perspective. In: Jackson J, ed. The Black American Elderly: Research on Physical and Psychosocial Health. New York, NY: Springer; 1988:190-214.
6. Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC). The Fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC V). Arch Intern Med.. 1993;153:154-183. [Medline]
7. MacMahon S, Peto R, Cutler J. Blood pressure, stroke, and coronary heart disease, I: prolonged differences in blood pressure—prospective observational studies corrected for the regression dilution bias. Lancet.. 1990;335:765-774. [Medline]
8. Ooi WL, Budner NS, Cohen H. Impact of race on treatment response and cardiovascular disease among hypertensives. Hypertension.. 1989;14:227-234. [Abstract]
9. Collins R, Peto R, MacMahon S. Blood pressure, stroke and coronary heart disease, part II: short-term reductions in blood pressure: overview of randomized drug trials in their epidemiological context. Lancet.. 1990;335:827.[Medline]
10. Leren P. The hypertension-CHD dilemma. Acta Med Scand.. 1985;217:145-147. Editorial.
11. Multiple Risk Factor Intervention Trial Research Group (MRFIT). Baseline test electrocardiographic abnormalities, antihypertensive treatment, and mortality in the Multiple Risk Factor Intervention Trial. Am J Cardiol.. 1985;55:1-15. [Medline]
12. Lyles MF. Adverse effects of antihypertensive therapy. Clin Geriatr Med.. 1989;5:691-702. [Medline]
13. Dunbar-Jacob J, Dwyer K, Dunning EJ. Compliance with antihypertensive regimen: a review of the research in the 1980s. Ann Behav Med.. 1991;13:31-39.
14. Stason WB. Cost and quality trade-offs in the treatment of hypertension. Hypertension. 1989;13(suppl 1):I-145-I-148.
15. Shulman NB. Economic issues relating to access to medications. In: Saunders E, ed. Cardiovascular Diseases in Blacks. Philadelphia, Pa: FA Davis Co; 1991:79-82.
16. Applegate WB. Hypertension in elderly patients. Ann Intern Med.. 1989;110:901-915. [Medline]
17. Hildreth CJ, Saunders E. Heart disease, stroke and hypertension in blacks. In: Braithwaite RL, Taylor S, eds. Health Issues in the Black Community. San Francisco, Calif: Jossey-Bass Inc Publishers; 1992:90-105.
18. Schneider RH, Egan BM, Johnson EH, Drobny H, Julius S. Anger and anxiety in borderline hypertension. Psychosom Med.. 1986;48:242-248. [Abstract]
19. Markovitz JH, Matthews KA, Kannel WB, Cobb JL, D’Agostino RB. Psychological predictors of hypertension in the Framingham Study: is there tension in hypertension? JAMA.. 1993;270:2439-2443. [Abstract]
20. Anderson NB, Myers HF, Pickering T. Hypertension in blacks: psychosocial and biological perspectives. J Hypertens.. 1989;7:161-172. [Medline]
21. Calhoun DA, Mutinga ML, Colins AS, Wyss JM, Oparil S. Normotensive blacks have heightened sympathetic response to cold pressor test. Hypertension.. 1993;22:801-805. [Abstract]
22. Magnus M. Cardiovascular health among African Americans: a review of the health status, risk reduction, and intervention strategies. Am J Health Promotion.. 1991;5:282-290.
23. Trials of Hypertension Prevention Collaborative Research Group (TOHP). The effects of non-pharmacologic interventions on blood pressure of persons with high normal levels: results of the Trials of Hypertension Prevention, phase 1. JAMA.. 1992;267:1213-1220. [Abstract]
24. Eisenberg DM, Belbanco TL, Berkey CS, Kaptchuk TJ, Kupelnick B, Julh I, Chalmers TC. Cognitive behavioral techniques for hypertension: are they effective? Ann Intern Med.. 1993;118:964-972. [Medline]
25. Jacob RG, Chesney MA, Williams DM, Ding Y, Shapiro AP. Relaxation therapy for hypertension: design effects and treatment effects. Ann Behav Med.. 1991;13:5-17.
26. Roth R. Maharishi Mahesh Yogi’s Transcendental Meditation. Washington, DC: Primus; 1994:1-194.
27. Dillbeck MC, Orme-Johnson DW. Physiological differences between Transcendental Meditation and rest. Am Psychol.. 1987;42:879-881.
28. Eppley KR, Abrams AI, Shear J. Differential effects of relaxation techniques on trait anxiety: a meta-analysis. J Clin Psychol.. 1989;45:957-974. [Medline]
29. Alexander CN, Robinson P, Orme-Johnson D, Schneider RH, Walton KG. The effects of Transcendental Meditation compared to other methods of relaxation and meditation in reducing risk factors, morbidity, and mortality. Homeostasis. 1994a;35:243-264.
30. Alexander CN, Rainforth M, Gelderloos P. Transcendental Meditation, self-actualization and psychological health: a conceptual overview and statistical meta-analysis. J Soc Behav Pers.. 1991;6:189-247.
31. Alexander CN, Chandler HM, Langer EJ, Newman RI, Davies JL. Transcendental Meditation, mindfulness, and longevity: an experimental study with the elderly. J Pers Soc Psychol.. 1989;57:950-964. [Medline]
32. Wallace RK, Silver J, Mills PJ, Dillbeck MC, Wagoner DE. Systolic blood pressure and long term practice of the Transcendental Meditation and TM-Sidhi program: effects of TM on systolic blood pressure. Psychosom Med.. 1983;45:41-46. [Abstract]
33. Schneider RH, Alexander CN, Wallace RK. In search of an optimal behavioural treatment for hypertension: a review and focus on Transcendental Meditation. In: Johnson EH, Gentry WD, Julius S, eds. Personality, Elevated Blood Pressure, and Essential Hypertension. Washington, DC: Hemisphere Publishing Corp; 1992:291-312.
34. Bernstein DA, Boerkovic TD. Progressive Relaxation Training. Champaign, Ill: Research Press; 1973:1-66.
35. Gaylord C, Orme-Johnson D, Travis F. The effects of the Transcendental Meditation technique and progressive muscle relaxation on EEG coherence, stress reactivity, and mental health in Black adults. Int J Neurosci.. 1989;46:77-86. [Medline]
36. Simon R. Adaptive treatment assignment methods and clinical trials. Biometrics.. 1977;33:743-749. [Medline]
37. Meinert CL. Clinical Trials: Design, Conduct, and Analysis. New York, NY: Oxford University Press; 1986:295.
38. Hunt JC, Frohlich ED, Moser M, Rocella FJ, Keighley FA. Devices used for self-measurement of blood pressure: revised statement of the National High Blood Pressure Education Program. Arch Intern Med.. 1985;145:2231-2234. [Medline]
39. Mengden T, Vetter W. An evaluation of self-measured blood pressure versus casual and ambulatory blood pressure: implications for clinical trials. Cardiovascular Reviews and Reports.. 1994;15:35-48.
40. Jackson J, Gurien G, Bowman P, Hatchett S. National Survey of Black Americans: A Study of Black American Life. Ann Arbor, Mich: Survey Research Center, University of Michigan; 1983:1085.
41. Spielberger CD, Jacobs G, Barker L. Preliminary Manual for the State-Trait Personality Inventory (STPI). Tampa, Fla: Center for Research in Behavioral Medicine and Community Psychology. University of South Florida: 1979.
42. Walston KA, Walston BS, DeVeliss MB. Development of the Multidimensional Health Locus of Control (MHLC) scales. MB Health Education Monograph.. 1978;6:160-170.
43. Bradburn N. The Structure of Well-Being. Chicago, Ill: Aldine; 1969:1-120.
44. Hunt SM, McKenna SP, McEwen J. The Nottingham Health Profile: subjective health status and medical consultations. Soc Sci Med. 1981;15A:221-229.
45. Scheier MF, Carver CS. Optimism, coping, and health: assessment and implications of generalized outcome expectancies. Health Psychology.. 1985;4:219-247. [Medline]
46. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol.. 1986;124:453-469. [Abstract]
47. Hoelscher TJ, Lichstein KL, Rosenthal TL. Objective versus subjective assessment of relaxation compliance in anxious individuals. Behav Res Ther.. 1984;22:187-193. [Medline]
48. Agras WS, Thorne M, Taylor CB. Expectation and the blood pressure lowering effects of relaxation. Psychosom Med.. 1982;44:389-395. [Abstract]
49. Schneider RH, Cavanaugh KJ, Kasture HS, Rothenberg S, Averbach R, Robinson D, Wallace RK. Health promotion with a traditional system of natural health care: Maharishi Ayur-Veda. J Soc Behav Pers.. 1990;5:1-27.
50. Maharishi Mahesh Yogi. On the Bhagavad Gita: A New Translation and Commentary. Baltimore, Md: Penguin; 1969:470-471.
51. Jevning R, Wallace RK, Beidebach M. The physiology of meditation: a review: A wakeful hypometabolic response. Neurosci Biol Behav Rev.. 1992;168:415-424.
52. Jacobson E. Progressive Relaxation. Chicago, Ill: University of Chicago Press; 1992:28-100.
53. Rosenthal R, Rosnow RL. Essentials of Behavioral Research: Methods of Data Analysis. New York, NY: McGraw-Hill Publishing Co; 1984:477.
54. Dixon WJ, Brown MB, Engelman L, Hill MA, Jennrich RI. BMDP Statistical Software Manual. Berkeley, Calif: University of California Press; 1990;11:1207-1244.
55. Jacob RG, Chesney MA. Psychological and behavioral methods to reduce cardiovascular reactivity. In: Matthews KA, Weiss SM, Detre T, Dembroski TM, Falkner B, Manuck SB, Williams RB Jr, eds. Handbook of Stress, Reactivity, and Cardiovascular Disease. New York, NY: John Wiley & Sons; 1986:417-457.
56. Alexander CN, Robinson P, Rainforth M. Treating alcohol, nicotine, and drug abuse through Transcendental Meditation: a review and statistical meta-analysis. Alcoholism Treatment Quarterly. 1994b;11:11-84.
57. Thijs L, Fagard R, Lijnen P, Staessen J, Van Hoof R, Amery A. A meta analysis of outcome trials in elderly hypertensives. J Hypertens.. 1992;10:1103-1109. [Medline]
58. Mejia A, Julius S. Practical utility of blood pressure readings obtained by self-determination. J Hypertens. 1989;7(suppl 3):S53-S57.
59. World Hypertension League. Self-measurement of blood pressure. J Hypertens.. 1988;6:257-261. [Medline]
60. Kaufmann PG, Jacob RG, Ewart CK. Hypertension intervention pooling project. Health Psychol. 1988;7(suppl):209-224.
61. MacMahon S, Cutler J, Brittain E, Higgins M. Obesity and hypertension: epidemiological and clinical issues. Eur Heart J. 1987;8(suppl B):57-70.
62. Cutler JA, Follmann D, Elliot P, Suh I. An overview of randomized trials of sodium reduction and blood pressure. Hypertension. 1991;17(suppl I):I-27-I-33.
63. World Hypertension League. Physical exercise in the management of hypertension: a consensus statement by the World Hypertension League. J Hypertens.. 1991;9:283-287. [Medline]
64. Pieper C, LaCroix AZ, Karasak RA. The relation of psychosocial dimensions of work with coronary heart disease risk factors: a meta-analysis of 5 United States databases. Am J Epidemiol.. 1989;129:483-494. [Abstract]
65. Julius S, Fitzpatrick MA, Egan B, Schneider RH. Neural mechanisms in clinical hypertension. In: Guthrie GP, Kotchen TA., eds. Hypertension and the Brain. New York, NY: Futura Publications; 1984:305-318.
66. Fredrikson N, Matthews KA. Cardiovascular responses to behavioral stress and hypertension: a meta-analytic review. Ann Behav Med.. 1990;12:30-39.
67. Anderson NB, McNeilly M, Myers H. Autonomic reactivity and hypertension in blacks: a review and proposed model. Ethn Dis.. 1991;1:154-170. [Medline]
68. Carvalho JJM, Baruzzi RG, Howard PF. Blood pressure in four remote populations in the INTERSALT study. Hypertension.. 1989;14:238-246. [Abstract]
69. Mills PJ, Schneider RH, Hill D, Walton KG, Wallace RK. Beta-adrenergic receptor sensitivity in subjects practicing Transcendental Meditation. J Psychosom Res.. 1990;34:29-33. [Medline]
70. Orme-Johnson D. Medical care utilization and the Transcendental Meditation program. Psychosom Med.. 1987;49:493-507. [Abstract]
71. Cooper MJ, Aygen MM. Transcendental Meditation in the management of hypercholesterolemia. J Hum Stress.. 1979;5:24-27.
72. Airhihenbuwa CO. Health promotion and disease prevention strategies for African Americans: a conceptual model. In: Braithwaite RL, Taylor SE, eds. Health Issues in the Black Community. San Francisco, Calif: Jossey-Bass Inc Publishers; 1992:267-280.
73. Neighbors HW, Braithwaite RL, Thompson E. Health promotion and African-Americans: from personal empowerment to community action. Am J Health Promotion.. 1995;9:281-287.
74. Fries JF, Koop CE, Beadle CE. Reducing health care costs by reducing the need and demand for medical services. N Engl J Med.. 1993;329:321-325. [Free Full Text]
75. Leaf A. Preventive medicine in our ailing health care system. JAMA.. 1993;269:616-618. [Medline]
Institution: CTRO HOS COIMBRA HOSP GRAL Sign In as Member/Individual (Non-Member) Subscription Activation
Abstract of this Article ( )
Email this article to a friend
Similar articles found in: Hypertension Online PubMed
This Article has been cited by:
other online articles
Search PubMed for articles by: Alexander, C. N. || Egan, B.
Alert me when: new articles cite this article
Download to Citation Manager
© 1996 American Heart Association, Inc.
Trial of Stress Reduction for Hypertension in Older African Americans
II. Sex and Risk Subgroup Analysis
Charles N. Alexander; Robert H. Schneider; Frank Staggers; William Sheppard; B. Mawiyah Clayborne; Maxwell Rainforth; John Salerno; Kofi Kondwani; Sandra Smith; Kenneth G. Walton; Brent Egan
the Department of Psychology (C.N.A., B.M.C., M.R., K.K., K.G.W.) and Center for Health and Aging Studies, Department of Physiological and Biological Sciences (C.N.A., R.H.S., J.S., K.K., K.G.W.), Maharishi University of Management, Fairfield, Iowa; Hypertension and Stress Management Research Clinic, West Oakland (Calif) Health Center (F.S., W.S., S.S.); Haight-Ashbury Free Medical Clinic, San Francisco, Calif (F.S.); and Medical University of South Carolina, Charleston (B.E.). Preliminary portions of this study were presented at the Society of Behavioral Medicine, San Diego, Calif, March 22-25, 1995, and at the Centennial Conference of the National Medical Association, Atlanta, Ga, July 29 to August 3, 1995.
Top Abstract Introduction Methods Results Discussion References
Our objective was to test the short-term efficacy and feasibility of two stress-reduction approaches for the treatment of hypertension in older African Americans, focusing on subgroup analysis by sex and by high and low risk on six measures of hypertension risk: psychosocial stress, obesity, alcohol use, physical inactivity, dietary sodium-potassium ratio, and a composite measure. The study involved a follow-up subgroup analysis of a 3-month randomized, controlled, single-blind trial conducted in a primary care, inner-city health center. Subjects were 127 African American men and women, aged 55 to 85 years, with diastolic pressure of 90 to 104 mm Hg and systolic pressure less than or equal to 179 mm Hg. Of these, 16 did not complete follow-up blood pressure measurements. Mental and physical stress-reduction approaches—the Transcendental Meditation technique and progressive muscle relaxation, respectively—were compared with a lifestyle modification education control and with each other. Both systolic and diastolic pressures changed from baseline to follow-up for both sexes and for high and low risk level (defined by median split) on the six measures of hypertension risk. Compared with education control subjects, women practicing the Transcendental Meditation technique showed adjusted declines in systolic (10.4 mm Hg, P<.01) and diastolic (5.9 mm Hg, P<.01) pressures. Men in this treatment group also declined in both systolic (12.7 mm Hg, P<.01) and diastolic (8.1 mm Hg, P<.001) pressures compared with control subjects. Women practicing muscle relaxation did not show a significant decrease compared with control subjects, and men declined significantly in diastolic pressure only (6.2 mm Hg, P<.01). For the measure of psychosocial stress, both the high and low risk subgroups using the Transcendental Meditation technique declined in systolic (high risk, P=.0003; low, P=.06) and diastolic (high risk, P=.001; low, P=.008) pressures compared with control subjects, whereas for muscle relaxation, blood pressure dropped significantly only in the high risk subgroup and only for systolic pressure (P=.03) compared with control subjects. For each of the other five risk measures, Transcendental Meditation subjects in both the high and low risk groups declined significantly in systolic and diastolic pressures compared with control subjects. Effects of stress reduction on blood pressure were found to generalize to both sexes and diverse risk factor subgroups and were significantly greater in the Transcendental Meditation treatment group. These effects (along with high compliance) even in individuals with multiple risk factors for hypertension clearly warrant longer-term investigation in this and other populations.
Key Words: stress • risk factors • meditation • relaxation • blacks • blood pressure
Top Abstract Introduction Methods Results Discussion References
The age-adjusted rate of hypertension in African American adults is more than 50% higher than for white Americans,1 and African Americans die from hypertensive diseases at substantially higher rates than do whites.2 3 These disparities in hypertension and related mortality may be partly due to the higher prevalence and severity of several hypertension risk factors in African Americans, including psychosocial stress, obesity, alcohol use, physical inactivity, and sodium sensitivity.1 When these risk factors occur in combination, they appear to produce a multiplicative effect, substantially increasing the likelihood of hypertension and subsequent cardiovascular disease.4 African Americans are 50% more likely than whites to display multiple risk factors for cardiovascular disease.1
There are also important sex differences in hypertension and related risk factors in African Americans. Elderly black women show the highest prevalence of hypertension (>75%) of any race/age subgroup,5 and their age-adjusted total mortality rate for heart disease is two thirds higher than for white women. Moreover, black women are twice as likely as both white women and black men to be obese.6 With respect to black men, the age-adjusted total mortality rate from heart disease is 33% higher than in white men. Moreover, compared with black women, there is evidence that black men have a higher drinking rate,7 lower compliance with antihypertensive medication,8 a lower leisure-time physical activity rate,1 and higher psychosocial stress on some indicators, as suggested, for example, by higher rates of suppressed hostility and suicide.9 10
Not only are African Americans exposed to more extreme socioenvironmental stressors than whites,11 but they also may show greater cardiovascular and sympathetic reactivity to such stressors.12 Psychosocial stress also may contribute indirectly to high BP by increasing the intensity and clustering of other maladaptive behavioral responses associated with an increased risk of hypertension (eg, obesity and alcohol abuse). Stress related to mounting socioeconomic and health problems later in life may particularly place elderly African Americans at multiple jeopardy for hypertensive disease.7 Moreover, there may be significant sex differences in the contribution of stress to hypertension. A recent study13 found that high levels of anxiety/tension predicted the subsequent development of hypertension in older men but not in older women in the general population; it was recommended that future studies assess this sex difference specifically for African Americans.
Given the contribution of stress to the etiology of hypertension in older African Americans, the application of behavioral methods for stress reduction would seem particularly relevant to this population. Yet with the exception of our own recent study,14 no trials have been reported on stress management in the treatment of hypertension in elderly African Americans.15 In our original randomized controlled trial in hypertensive older African Americans, we found that compared with a lifestyle education control (EC) group, the Transcendental Meditation (TM) technique, a mental technique for stress reduction, reduced systolic and diastolic BPs approximately twice as much as a physical relaxation approach, progressive muscle relaxation (PMR), after 3 months of follow-up.14 These results were consistent with recent statistical meta-analyses in primarily white populations indicating that the TM technique reduced several risk behaviors—including acute and chronic physiological arousal,16 alcohol and cigarette use,17 anxiety,18 and poor self-esteem19 —significantly more than other forms of relaxation.
In light of the significant contribution of sex differences and the increased frequency of hypertension risk behaviors in hypertensive African Americans, we reanalyzed data from our original study to address two clinically important questions: Will application of these stress-reduction techniques in older African Americans produce similar BP reductions for (1) both sexes and (2) individuals at high as well as low levels on several hypertension risk factors, including psychosocial stress, obesity, alcohol use, physical inactivity, sodium-potassium ratio, and a composite measure of risk?
Top Abstract Introduction Methods Results Discussion References
More detailed methods are described in the original study.14
Criteria for subject selection included self-identification as African American, age 55 years and older, and hypertension initially defined as 90 to 109 mm Hg DBP and less than or equal to 189 mm Hg SBP based on three measures taken at the screening visit. BP and quality of life measures were assessed during four subsequent baseline visits over a 1- to 2-month period. Subjects were subsequently excluded from the study if their BP exceeded 104 mm Hg DBP or 179 mm Hg SBP on two successive baseline visits. Volunteers taking antihypertensive medications were eligible for the trial, as were individuals not taking BP medications, if they met the entry criteria noted above and if they did not change their medication regimen during either the baseline or treatment phase of the study. The trial was conducted at the West Oakland Health Center, a primary care community health center in Oakland, Calif. This study was approved by the institution's Committee for Protection of Human Subjects. All subjects gave informed consent to participate.
After completing the baseline period, subjects were randomly assigned to one of three treatment conditions: (1) The TM program, the principle technology of consciousness of the Maharishi's Vedic Approach to Health, which has been recently described by Nader20 as a comprehensive, prevention-oriented system of natural health care brought to light by Maharishi Mahesh Yogi from the ancient Veda and Vedic literature21 22 ; (2) PMR, a widely studied physical approach, adapted by Bernstein and Borkovec23 from Jacobson's original technique for systematically relaxing the major muscle groups of the body24 ; and (3) a lifestyle-modification education control (EC) based on guidelines for nondrug treatment of hypertension recommended by the Joint National Committee.25 The two active stress-reduction procedures were matched for expectancy of benefits, teaching format, instructional time, and home practice (20 minutes twice a day), and neither required change in personal beliefs or lifestyle. Initial instruction occurred over 1 week, with 1.5-hour monthly follow-up meetings. To partially control for attention and expectancy, EC subjects also met with treatment providers for 1-hour follow-up sessions monthly and were given the expectation that their BP could be managed by adopting the lifestyle changes that were recommended.
BP follow-up took place monthly for 3 months. The final posttest assessment was based on the average of two BP measurements taken 1 week apart during the 11th and 12th weeks. BPs were measured with an automated BP monitor (model 300S, Vitastat Medical Services) with digital readout calibrated against a mercury sphygmomanometer at regular intervals. BP measurements (by testers blind to the subject's treatment status) were taken after subjects sat for 5 minutes without practicing any relaxation techniques. During each visit, three assessments were taken at 1-minute intervals using the first and fifth Korotkoff sounds, with the final reading based on the average of the last two measurements.
Six factors related to hypertension risk were measured at pretest. Subjects were not stratified on these variables before randomization. Instead, they were divided into high and low risk levels on each of these parameters for subsequent subgroup analysis. This method not only controlled for potential differences between groups on each risk factor but allowed assessment of the generalizability of treatment effects across risk level for the various factors (see "Data Analysis"). The six risk factor measures were as follows: (1) Obesity: Subjects were weighed without shoes or outdoor garments on a balance scale that also measured height (Healthometer, 402KL). BMI was calculated from weight and height (kilograms per meter squared). (2) Alcohol consumption: The number of drinks per week (with one drink defined as 12 oz of beer, 5 oz of wine, or 1.5 oz of liquor) was assessed by a short questionnaire adapted from the Multiple Risk Factor Intervention Trial (MRFIT).26 (3) Physical exertion level: The number of hours per day of moderate or vigorous physical activity was assessed by a questionnaire adapted from MRFIT.26 (4) Dietary sodium-potassium: Intake of sodium and potassium was estimated from the Trials of Hypertension Prevention Health Habits Questionnaire and was calculated with a normative database of standard portion sizes and nutrient contents.27 Because some researchers consider low potassium intake (in addition to high sodium intake) to be a contributor to hypertension risk28 and to control for response differences in the ability of subjects to recall all foods eaten, the ratio of sodium intake to potassium intake was used as the outcome measure. (5) Psychosocial stress: A battery of psychosocial measures of quality of life was administered verbally and individually to each participant over three pretest sessions and again at posttest. The battery included the personal efficacy subscale from the National Survey of Black Americans29 ; trait anxiety and trait anger subscales from the State-Trait Personality Inventory30 ; the Rosenberg Self-Esteem Scale; well-being as measured by the Affect Balance Scale31 ; global score on the Nottingham Health Profile, which contains four subscales measuring physical health and two measuring psychosocial health32 ; total score on the short-form Duke University Multidimensional Health Profile, which contains five subscales assessing general health, symptoms, and physical, social, and emotional functioning; and total score on the health problems and impact subscale of the Index of Illness. Because of the large number of quality of life measures, we performed a principle components analysis to identify a smaller number of underlying quality of life factors, thus reducing the probability of a type 1 error.33 The first factor, which accounted for the largest proportion of variance (28%), was taken as an indicator of psychosocial stress. It was composed of high scores on trait anger and anxiety and low scores on self-esteem, well-being, and personal efficacy (with factor loadings 0.50). Both anger and anxiety have been implicated in high BP,34 35 and more recently, psychosocial "cushioning" factors such as self-esteem have been seen to protect against hypertension.36 37 (6) Multiple risk factors: The prevalence of multiple risk factors was determined by score on a weighted composite index for the above five risk factors as determined by multiple regression (see "Data Analysis").
Baseline characteristics of the three groups were compared by MANOVA and ANOVA. The baseline characteristics included age, sex, BMI, BP, medication status, and psychological and behavioral/lifestyle characteristics. Significance was set at a value of P<.05 for these and all other statistical comparisons.
Treatment outcomes were assessed by ANCOVA, with treatment status as the independent variable, changes in SBP and DBP as the dependent variables, and baseline SBP and DBP and other baseline characteristics that significantly differed between the groups as covariates. Change in BP was defined as 3-month posttreatment BP minus baseline BP. Additionally, two types of intent-to-treat analyses were performed for determination of whether lost data due to dropouts altered the results. First, missing BP change scores were treated as missing at random, and the BMDP 5V routine33 was applied for estimation of group means from all available data. Second, more conservatively, the maximum increase in BP observed in any subject was applied as the missing value for all subjects.
For assessment of treatment outcomes in relation to sex, separate ANCOVAs were performed within sex subgroups using the same independent variable, dependent variables, and covariates as in the analysis of outcomes for the entire sample. Also, repeated measures ANCOVAs based on all three posttest visits examined monthly BP change within the sex subgroups.
Similar to the within-subgroups analyses for sex, treatment effects were separately assessed by ANCOVA comparing each high and low risk subgroup, formed by median split on each risk factor. In addition to examination of the five individual risk factors taken separately, these factors were used in combination to create an index of overall hypertension risk. This was accomplished by performing regression analysis on the data for the control group alone to predict BP change over 3 months. Separate regression models were constructed for SBP and DBP as the dependent variables, with the pretest levels of the five risk factors as the explanatory or independent variables. The two regression models accounted for 29.8% of the SBP change score and 25.9% of the variance in DBP change, respectively. The coefficients from these models yielded two regression equations from which were constructed two composite indexes for relative risk of increase in SBP and DBP. As with the individual risk factors, the sample was divided by median split on the two composite indexes. The same ANCOVAs as for the separate risk factors were performed on the overall high and low risk subgroups, based on the composite SBP risk index for SBP change and the composite DBP index for DBP change.
Planned contrasts allowed pairwise comparisons of the three treatment groups on BP outcomes. These contrasts were one-tailed because of directionality of predictions. Given our initial findings on the effects of stress reduction on BP reduction in African Americans14 and previous research suggesting the generalizability of effects (especially of the TM technique) across sex and in various high-risk populations,38 it was hypothesized that for each sex and for each risk level (ie, high and low) on the various risk factors, both active interventions would be more effective than the EC intervention and that the TM group would show greater reductions in BP than the PMR group. Two-tailed tests were used for statistical comparisons that did not evaluate effects of treatment (eg, on outcome differences by sex or risk level within each treatment group).
We performed power analyses to assess the statistical power to conduct the above subgroup analyses. In the original study,14 with 37 subjects per treatment group, effect size differences between treatments were as follows: TM-EC=0.88 for SBP and 0.94 for DBP; PMR-EC=0.38 for SBP and 0.49 for DBP; TM-PMR=0.49 for SBP and 0.46 for DBP. Assuming that effect size differences between treatments were the same for both women and men; that cell sizes were 21.3 and 15.7 subjects, respectively; and that ANCOVA with one-tailed planned comparisons at the 5% significance level were used, power for treatment comparisons in women was as follows: for TM versus EC, 88% to 92%; TM versus PMR, 43% to 47%; and PMR versus EC, 34% to 47%. In men it would be as follows: TM versus EC, 77% to 83%; TM versus PMR, 35% to 38%; and PMR versus EC, 28% to 38%. Power for comparisons at different risk levels was similar to that for these sex comparisons. Assuming identical effect size differences between treatments for both the high and low risk subgroups and all cell sizes of 18.5 due to median split on each risk factor, the power for treatment comparisons for both risk levels would be as follows: for TM versus EC, 83% to 88%; TM versus PMR, 39% to 43%; and PMR versus EC, 31% to 42%. Thus, given the large effect size differences between TM and EC, there was ample power for detection of differences between these treatment conditions in the various subgroup analyses. Given the smaller effect size differences for the other treatment comparisons, the power for detection of differences was correspondingly reduced.
Top Abstract Introduction Methods Results Discussion References
As previously described,14 127 subjects were randomized to treatment groups. There was a 12.6% attrition rate over 3 months of treatment, resulting in a final sample of 111 subjects. Of the 16 subjects lost to follow-up, 11 dropped out because of change in hypertensive medication (2 in the TM group; 3, PMR; 6, EC); the 5 others dropped out primarily because of change in residence (2 in the TM group; 2, PMR; 1, EC). There was no sex difference in attrition rates. Pooling across sex, MANOVA showed no overall significance between treatment groups on the 19 baseline physiological and quality of life variables. Univariate ANOVA indicated that only age significantly differed between groups (TM=63.7 years, PMR=69.2 years, EC=67.4 years; P<.01). However, in this sample, age was not found to be significantly correlated with SBP or DBP either at baseline or with change in BP over time.
Table 1 presents baseline BP and demographic characteristics separately for women and men for each treatment group. In the total sample, 58% were women (n=74) and 42% (n=53) were men. Women (mean age, 67.9 years) were significantly older (P<.05) than men (mean age, 64.8 years). Also, BMI was significantly higher for women than for men (P=.015), and as assessed by 2 analysis, significantly more women (45%) were obese (BMI >30.0 kg/m2) than men (23%). Approximately half of the total sample was taking hypertensive medications (55.4% of the women, and 41.5% of the men). Baseline SBP for women (150.1 mm Hg) was significantly higher (P<.05) than for men (143.5 mm Hg); DBP did not differ significantly between sexes.
View this table:[in this window][in a new window] Table 1. Comparison of Women and Men at Pretest by Treatment Group
Participants' monthly reports indicated high compliance rates for the active treatments: 97.1% of the TM group and 81.1% of the PMR group practiced their programs twice daily or almost twice daily. By test of proportion, compliance rates did not significantly differ between these two groups. Both groups rated their instructors as "excellent" (3.9 on a 4-point scale).39 The groups also did not differ on a measure of expectancy of positive outcomes from treatment. All three groups expected to change substantially, as indicated by high scores on a 4-point scale (3.2 for the TM group; 3.1 for PMR; 3.1 for EC). Also, there were no sex differences for compliance rate or expectancy.
BP Changes in Women and Men
Pooling across treatment groups, ANCOVA indicated that the change in SBP in women (-6.7 mm Hg) was significantly greater (P<.05, two-tailed) than in men (-3.3 mm Hg) over the 3-month period; however, there was no significant sex difference in DBP change. For each treatment group, Table 2 presents, separately for women and men, the raw SBP and DBP change scores and change scores adjusted for age and baseline SBP and DBP covariates. Compared with EC women, TM women showed an adjusted decrease of 10.4 mm Hg in SBP (P<.01) and 5.9 mm Hg in DBP (P<.01). Compared with EC women, PMR women showed a nonsignificant decrease of 4.8 mm Hg in SBP and 1.9 mm Hg in DBP. Compared with PMR women, TM women showed nonsignificant decreases of 5.6 mm Hg in SBP and 4.0 mm Hg in DBP. Compared with EC men, TM men showed significant adjusted decreases of 12.7 mm Hg in SBP (P<.01) and 8.1 mm Hg in DBP (P<.001). Compared with EC men, PMR men showed a nonsignificant reduction of 6.1 mm Hg in SBP and a significant reduction of 6.2 mm Hg in DBP (P<.01). Compared with PMR men, TM men showed a significant reduction of 6.6 mm Hg in SBP (P<.05) and a nonsignificant decrease of 1.9 mm Hg in DBP. Probability values for the two intent-to-treat analyses were consistent with those from the primary BP analyses. Also, continuing subjects and dropouts did not differ significantly in baseline BP, demographics, and quality of life variables.
View this table:[in this window][in a new window] Table 2. Change From Baseline in Clinic SBP and DBP for Stress-Reduction and Control Groups by Sex
Fig 1 shows SBP (top) and DBP (bottom) changes adjusted for baseline levels and age for the 60 women with complete data for each of the three monthly follow-up visits. Fig 2 presents the same BP changes for the 44 men with complete data. Compared with EC women, TM women showed a significant decrease in SBP (P<.01) and DBP (P<.009), whereas PMR women did not. Compared with PMR women, TM women showed a decline in DBP (P<.07, trend) but not SBP. Compared with EC men, TM men showed significant decreases in SBP (P<.0004) and DBP (P<.00005), and PMR men in DBP (P<.02) but not SBP. Compared with PMR men, TM men showed a significant decline in SBP (P<.01) and DBP (P<.08, trend).
View larger version (16K):[in this window][in a new window] Figure 1. Adjusted changes in clinic SBP (top) and DBP (bottom) in women over the 3-month treatment period. Probability values are for repeated measures ANCOVA comparing each experimental group (TM [n=18] and PMR [n=18]) with EC (n=24). Probability trend for TM vs PMR: DBP, P<.07.
View larger version (17K):[in this window][in a new window] Figure 2. Adjusted changes in clinic SBP (top) and DBP (bottom) in men over the 3-month treatment period. Probability values are for repeated measures ANCOVA comparing each experimental group (TM [n=18] and PMR [n=15]) with EC (n=11). Probabilities for TM vs PMR: SBP, P<.01; DBP, P<.08, trend.
Moderating Role of Risk Factors for Hypertension
In Table 3 , change scores in SBP and DBP for the three treatment groups are separately presented (covarying for baseline BP and age) by low and high level on the various hypertension risk factors.
View this table:[in this window][in a new window] Table 3. Adjusted Mean Change Scores in SBP and DBP in Elderly African American Men and Women in Low or High Levels on Hypertension Risk Factors
The composite psychosocial stress score (Z-score sum of scores) for the high risk subgroup was greater than or equal to -0.13 SD units. Fig 3 shows adjusted SBP (top) and DBP (bottom) change scores for the high and low risk subgroups in each treatment grouping. For both high and low risk subgroups, the TM group had significantly lower SBP (ie, adjusted mean change; high, P=.0003; low, P=.06, trend) and lower DBP (high, P=.001; low, P=.008) compared with the EC group. For the PMR group, only the high risk subgroup had significantly lower SBP (P=.03) compared with the EC group. Compared with PMR subjects, TM subjects had lower SBP (high, P=.06, trend; low, P=.05) and lower DBP (high, P=.05; low, P=.06, trend).
View larger version (30K):[in this window][in a new window] Figure 3. Adjusted changes in clinic SBP (top) and DBP (bottom) in high and low psychosocial stress subgroups after 3-month treatment. Probability values are for comparisons of adjusted mean change for each experimental group (TM [n=36] and PMR [n=33]) with EC (n=35). Probabilities for TM vs PMR: SBP, high risk, P=.06, trend; low risk, P=.05; DBP, high risk, P=.05; low risk, P=.06, trend.
As determined by median split, BMI for the high risk subgroup was greater than or equal to 28.8 kg/m2 for women and greater than or equal to 26.1 kg/m2 for men. For both the high and low risk subgroups, TM subjects had significantly lower SBP (high, P=.002; low, P=.007) and lower DBP (high, P=.006; low, P=.0004) compared with EC subjects. For the PMR group, the low risk subgroup had significantly lower SBP (P=.04) and the high risk subgroup had lower DBP (high, P=.03; low, P=.08, trend) compared with the EC group. Compared with PMR subjects, TM subjects had significantly lower SBP in the high risk subgroup (P=.01) and lower DBP in the low risk subgroup (P=.02). When obese versus nonobese subgroups were formed according to standard BMI criteria for women and men (rather than using the BMI medians), similar results were obtained.
Because about half the members of each group reported total abstinence from alcohol, the high risk alcohol-use subgroup was composed of all reported drinkers. For the high and low risk subgroups, TM subjects had significantly lower SBP (high, P=.00005; low, P=.02) and lower DBP (high, P=.00005; low, P=.02) compared with EC subjects. PMR subjects had significantly lower SBP (P=.0004) and DBP (P=.0002) in only the high risk subgroup compared with EC subjects. Compared with PMR subjects, TM subjects had significantly lower SBP (P=.01) and lower DBP (P=.02) in the low risk but not the high risk subgroup.
The number of hours per day spent in either moderate or vigorous physical activity by the high risk subgroup (ie, less physically active) was 2 hours or fewer. For both the high and low physical inactivity subgroups, TM subjects had lower SBP (high, P=.07, trend; low, P=.0001) and lower DBP (high, P=.005; low, P=.001) compared with EC subjects. Only the low physical inactivity subgroup of the PMR group had significantly lower SBP (P=.04) and lower DBP (P=.04) compared with the EC group. Compared with PMR subjects, the low physical inactivity TM subgroup had lower SBP (P=.03) and DBP (high, P=.09, trend; low, P=.09, trend).
The ratio of dietary sodium to potassium intake was greater than or equal to 1.11 for the high risk subgroup. For both high and low risk subgroups, TM subjects had significantly lower SBP (high, P=.06, trend; low, P=.0009) and lower DBP (high, P=.001; low, P=.003) compared with EC subjects. Compared with EC subjects, PMR subjects had significantly lower SBP (P=.02) and DBP (P=.01), but for the low risk subgroup only. Compared with PMR subjects, TM subjects showed a significant decrease in both SBP (P=.04) and DBP (P=.003) in the high risk but not the low risk subgroup.
Multiple Risk Factor Index
Subjects at high risk for a rise in BP were those with a predicted increase, based on all five risk factors, of greater than or equal to 0.94 mm Hg for SBP and greater than or equal to 0.37 mm Hg for DBP. For both high and low combined risk levels, TM subjects showed significant decreases in both SBP (high, P=.00005; low, P=.06, trend) and DBP (high, P=.0001; low, P=.004) compared with EC subjects. Compared with EC subjects, PMR subjects also showed significant decreases in SBP (P=.001) and DBP (P=.0004) but only for the high risk subgroup. Compared with PMR subjects, TM subjects showed significantly lower SBP (P=.04) and DBP (P=.002) for the low risk but not the high risk subgroup alone (see Table 3 ).
Although smoking is not a major risk factor for hypertension, it is for congestive heart disease. Thus, for the purposes of exploratory analysis, a high risk (for congestive heart disease) smoking subgroup was composed of all self-reported smokers. For both high and low risk smoking subgroups, TM subjects showed significant decreases compared with EC subjects in both SBP (high, 15.4 mm Hg, P=.013; low, 10.0 mm Hg, P=.003) and DBP (high, 9.7 mm Hg, P=.02; low, 5.2 mm Hg, P=.002). Compared with EC subjects, PMR subjects showed a significant decrease only for the low risk subgroup in DBP (2.4 mm Hg, P=.05). Compared with PMR subjects, TM subjects showed decreases in SBP for high risk (P=.09, trend) and low risk (P=.05) smoking subgroups and in DBP for the low risk subgroup (P=.07, trend).
To further assess the generalizability of treatment effects across high and low risk levels, we determined within each treatment group whether the effects on BP were similar (or different) for high and low risk subjects for each risk behavior. Within the TM group, the reduction in SBP or DBP did not differ significantly for subjects at either high risk or low risk on five of the six hypertension risk factor outcomes. The only exception was on the overweight factor: TM subjects who were of normal weight reduced DBP significantly more (P=.01) than those who were overweight; however, there was no difference in SBP. Also, for PMR subjects, there was no difference between subjects at high and low risk on the various risk factors. In contrast, for the EC group, subjects at high levels on four of the six risk variables showed a significant increase in SBP and DBP compared with subjects at low levels, including psychosocial stress (SBP, P=.01; DBP, P=.07, trend), physical inactivity (SBP, P=.004; DBP, P=.004), alcohol (SBP, P=.0002; DBP, P=.0003), and the overall index (SBP, P=.0004; DBP, P=.0003). For the obesity and sodium-potassium risk factors, SBP/DBP reduction did not differ by risk level for the EC group.
Top Abstract Introduction Methods Results Discussion References
Our original study was the first randomized controlled trial directly comparing the effects of mental and physical stress-reduction methods on hypertension in African Americans.14 To our knowledge, this follow-up study is the first to systematically investigate the applicability of stress-reduction techniques to different hypertension risk factor subgroups among African Americans. Moreover, few, if any, such risk factor subgroup studies have been conducted with behavioral approaches, even in the general elderly population.
The results of our subgroup analyses indicate the feasibility and short-term efficacy of the use of stress-reduction approaches in the treatment of hypertension in older African Americans of both sexes who are at high as well as low risk for six hypertension-related measures of risk: obesity, alcohol use, psychosocial stress, dietary sodium-to-potassium ratio, physical inactivity, and presence of multiple risks. After 3 months of follow-up, the TM technique significantly decreased SBP and DBP in both men and women compared with the EC group, whereas PMR significantly decreased DBP in both women and men but did not reduce SBP in either sex. For both women and men, the TM stress-reduction approach appeared to have an effect size approximately twice that of PMR. Furthermore, pooling across sexes, for both high and low risk factor subgroups, the TM technique decreased both SBP and DBP significantly more than the EC group for all six hypertension risk factors. For both high and low risk subgroups, PMR decreased SBP and DBP significantly more than the EC group for approximately half of the risk subgroup outcomes. For both high and low risk levels, the TM technique again reduced SBP and DBP approximately twice as much as PMR across the various risk factors.
These findings are likely to be internally valid because a randomized, single-blind design was used and multiple BP measures were taken during four baseline visits to minimize regression to the mean or white coat effects. Moreover, the use of two behavioral approaches in the same experimental setting allowed control for nonspecific intervention effects, with both active groups given similar expectancy of benefits, attention from trainers, and time allotted for daily practice. External validity was enhanced by conducting the trial in a major primary care center in a large inner-city African American community, thus increasing the likelihood of the generalizability of results to other such community settings. In addition, the baseline sex differences found in this study, including higher BMI and SBP in women and lower use of antihypertensive medication in men, are consistent with those seen in the larger older African American population.40 41 However, it should be noted that within each sex taken separately, ie, for men as well as women, the TM technique reduced BP more than PMR or EC. This was not simply because of differences in adherence to treatment, because compliance was high in both active treatment groups. Also, all subjects were included in the analysis irrespective of compliance levels.
The potential effects of methodological constraints in this study deserve examination. First, despite randomization, age significantly differed across groups at pretest. On the other hand, statistically covarying for age did not significantly alter treatment outcomes. Also, with 19 demographic and quality of life variables, one variable would be expected to be significant because of chance alone. Second, with the exception of obesity, all other risk factors were assessed by self-report, which may be less reliable than more objective methods of assessment. Although the "absolute" levels on the self-report measures may not be fully accurate, we used median splits to divide subjects into subgroups that were at least "relatively" higher or lower on the risk factor in question. The efficacy of this approach was supported by the finding that baseline differences on these risk measures tended to predict subsequent change in BP, as assessed by the multiple regression model. Third, because of the division into subgroups for analysis, sample size and hence statistical power was reduced in some cases. Power to detect differences between the TM and EC groups remained high (77% to 92%) but was lower for detecting differences between the stress-reduction approaches (35% to 47%) and between the PMR and EC groups (28% to 47%). However, significant differences in BP declines between the stress-reduction approaches and between PMR and EC were nevertheless found. Repeated measures ANCOVA of the subgroup data over the 3-month treatment period tended to yield more significant results because this statistical analysis enhances statistical power by reducing within-subject variance.
A final methodological concern is the relatively short (3 months) duration of the study. A critical question is whether these treatments will remain effective in reducing BP in both sexes and in the various risk factor subgroups over the long term. Studies of lifestyle modification approaches, particularly strict dietary regimens, have found it difficult to sustain generally more modest BP reductions over the long term.42 Intensive exercise programs, on the other hand, produce BP reductions similar to those found here in the TM group, but such programs appear to have lower compliance rates.43 44 We are currently conducting randomized controlled trials on the longer-term effects of these stress-reduction techniques on both hypertension and hypertensive heart disease in African Americans.
The results of many of the subgroup analyses are not only statistically significant, they are also potentially clinically significant. For example, the BP reductions through TM in the high psychosocial stress subgroup (11.3/5.3 mm Hg) and in the low psychosocial stress subgroup (11.4/6.4 mm Hg) are similar to those reported in antihypertensive drug trials.45 Comparable BP reductions in drug trials over the long term are associated with 35% to 40% less stroke and 20% to 45% less congestive heart disease morbidity.46 This is consistent with the results of a recent follow-up study47 of our randomized controlled trial with white elderly subjects39 showing lower cardiovascular and all-cause mortality rates and longer survival time over 8-year and 15-year periods in the TM group compared with two other mental techniques and usual care.
The results in this study also compare well with those reported for lifestyle-modification approaches. For example, meta-analyses show BP reductions of approximately 6/3 mm Hg in controlled trials of weight loss,48 dietary sodium reduction,49 and aerobic exercise.43 44 50 However, although these lifestyle modifications are known to at least moderately reduce BP if practiced regularly, adherence to such lifestyle regimens tends to be low over the long term.51 If overeating and alcohol misuse represent maladaptive attempts to cope with stress, then difficulties in altering such behaviors over the long term without enhancing underlying psychosocial well-being would not be surprising.52 53 54 In contrast, at least over the short term, compliance with the stress-reduction techniques studied here was quite high—97% for TM and 81% for PMR.
Possibly the most clinically relevant finding of these subgroup analyses was that BP reductions through the TM program, and to a lesser degree PMR, generalized across a wide range of risk conditions. Despite substantial differences in hypertension risk patterns, both African American men and women showed significant decreases in BP. Also, the high risk subgroups—ie, subjects who were psychologically distressed, obese, or physically inactive; used higher amounts of alcohol; consumed high sodium relative to potassium; or were at high risk on multiple factors—appeared to benefit from these treatments approximately as much as the low risk subgroups. It is noteworthy that even on the multiple risk factor variable, which accounted for 26% to 29% of the variance in BP change, high combined risk subjects changed as much as low risk subjects through the TM technique. (It should be noted that in the regression model, the regression coefficient for each risk factor was in the predicted direction except for the physical inactivity variable. Contrary to findings with other populations,50 higher physical activity was associated with an increase in BP in this minority elderly sample. Nevertheless, because our regression model was empirically derived, this variable was included. This made little difference in the outcome, however, because when the physical inactivity variable was excluded from the model, the BP differences between [and within] treatment groups remained essentially the same. The unexpected direction of association between physical inactivity and BP change may have been an artifact of the measure used and should be investigated more thoroughly in future studies.) The only exception was that DBP changed less in obese than in nonobese TM subjects. DBP nevertheless decreased significantly in obese TM subjects compared with EC subjects. In contrast, within the EC group, high risk subjects for most of the risk variables, including the multiple risk variable, significantly increased in BP relative to low risk subjects even over a 3-month period. This suggests that stress reduction may be most useful for preventing the progression of hypertension and subsequent cardiovascular disease in those who are at greatest risk.
The Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC V) recognized the contributions of sex, obesity, physical inactivity, alcohol consumption, dietary sodium, and psychosocial stress to the risk for hypertension and subsequent cardiovascular disease.25 The report also recommended lifestyle modification for all hypertensive individuals as either monotherapy or adjunctive therapy. JNC V additionally recommended that lifestyle modification be targeted to the particular behavioral disorder (eg, weight reduction for obese hypertensive individuals). Although the present findings do not negate these recommendations, they do suggest that stress reduction, at least with the TM technique, may be sufficient to reduce BP in older African Americans displaying various risk profiles.
Although JNC V acknowledged the contribution of stressors to hypertension, it questioned whether stress-reduction techniques have been shown to reduce or prevent high BP. One reason for the apparent disparity between our particular findings and their general conclusion is that JNC V did not focus on potential differences in the effects of specific approaches to stress reduction. Their conclusions were based on a narrative (ie, nonquantitative), selective review that emphasized the results of two prior randomized trials showing that muscle relaxation (in combination with one or more other stress-management approaches) did not significantly lower BP compared with controls.55 56 In contrast, the current study showed differential effects for TM compared with both PMR and EC. Thus, our data support the hypothesis that stress-reduction approaches are not homogeneous in their effects; that is, different techniques may produce different effects on BP, as well as on other clinical outcomes. Our findings are consistent with recent quantitative meta-analyses indicating that the TM technique is associated with effects approximately twice as large as found with other approaches—including PMR, other forms of muscle relaxation, electromyographic biofeedback, and techniques explicitly devised to induce a "relaxation response"—for chronic anxiety,18 poor self-esteem,19 and alcohol misuse,17 controlling for the strength of experimental design.
It is unlikely that changes in BMI, exercise, and dietary sodium-potassium accounted for the BP reductions in the present study because (1) subjects with low levels of these risk conditions at baseline showed BP reductions similar to those in subjects with high risk levels, and (2) other data from this study indicate that these factors did not change significantly over the 3-month period.57 With regard to changes in alcohol use, consumption did significantly decline among alcohol users in the TM group (by 6.5 and 4.6 drinks per week compared with EC and PMR, respectively), and this may have contributed to the decrease in BP in the alcohol users. However, the TM nonusers of alcohol also showed significant BP reductions, indicating that other mechanisms are likely to be involved.
With regard to psychosocial stress as a mechanism for the effects of stress reduction on BP, the finding that the subjects in low stress subgroups also showed significant declines in BP does not preclude the possibility that people who perceive that they are low in stress can still experience further declines in BP. Indeed, it was found that for the subjects who reported fewer stressful problems at baseline (as determined by median split), there was still a significant decline on the psychosocial stress factor for TM subjects compared with EC and PMR subjects combined. Decreased stress as a mechanism was also supported by the finding of a significant decrease on a second psychosocial factor, termed perceived health, in TM subjects compared with EC and PMR subjects combined. This factor was composed of total scores on three measures of perceived physical and mental health (see "Measures") that have been shown in the literature to be stress related.58 59
Thus, a decrease in psychosocial stress appears to have contributed to the effect of the mental technique on BP. From the perspective of the Maharishi Vedic Approach to Health, of which the TM technique is a key part, stress itself arises from a lack of integration of major physiological systems with the holistic "inner intelligence" of the body.20 This lack of integration from a contemporary perspective would appear to correspond to alterations of homeostatic mechanisms that have been related to hypertension.52 54 A growing body of research suggests that such alterations may be corrected or prevented through the TM technique, as indicated by decreased sympathetic activation,54 60 decreased hypothalamic-pituitary-adrenocortical activation,52 54 enhanced neurophysiological function,61 62 63 and enhanced serotonin metabolism.53 54 Such global physiological changes may have contributed to the reduction in BP seen across all sex and risk factor subgroups. Alternatively, there may be a different physiological mechanism specific to each risk subgroup. Further research to investigate the mechanisms for BP reduction in such diverse subpopulations is warranted.
Future studies also will be necessary to confirm and extend the current findings on the generalizability of these stress-reduction approaches, especially the TM technique, for the treatment of hypertension in diverse risk subgroups. Design features to incorporate in future research include longer-term follow-up of at least 1 year, larger subgroup sample sizes, assessment of additional objective risk factors (eg, sodium and potassium excretion, family history of hypertension, renin subtyping), stratification on key demographics before randomization, comparison with other lifestyle-modification approaches, and inclusion of other ethnic groups.
Selected Abbreviations and Acronyms
BMI = body mass index
BP = blood pressure
DBP = diastolic blood pressure
EC = lifestyle education control
PMR = progressive muscle relaxation
SBP = systolic blood pressure
TM = Transcendental Meditation
This work was supported by grants from the Retirement Research Foundation (Nos. 88-95 and 88-96), Chicago, Ill, and the Lancaster Foundation, Bethesda, Md. Preparation of this manuscript was supported in part by National Institutes of Health grant 5R01HL-48107. Transcendental Meditation and TM are service marks registered in the US Patent and Trademark Office licensed to Maharishi Vedic Education Development Corp and used under sublicense. The authors are grateful to the following individuals for their helpful advice and/or editorial comments: Brian Hofland, PhD; R. Keith Wallace, PhD; David Orme-Johnson, PhD; Robert Cooper, MD; Carolyn Gaylord King, PhD; and for technical assistance, Linda Heaton.
Reprint requests to Charles N. Alexander, PhD, Center for Health and Aging Studies, Maharishi University of Management, FB 1028, Fairfield, IA 52557-1028.
Preliminary portions of this study were presented at the Society of Behavioral Medicine, San Diego, Calif, March 22-25, 1995, and at the Centennial Conference of the National Medical Association, Atlanta, Ga, July 29 to August 3, 1995.
Received October 3, 1995; first decision October 27, 1995; first decision February 21, 1996;
Top Abstract Introduction Methods Results Discussion References
1. Working group on research in coronary heart disease in blacks. Report of the Working Group on Research in Coronary Heart Disease in Blacks. Washington, DC: National Institutes of Health, National Heart, Lung, and Blood Institute; 1994.
2. Gillum RF. Cardiovascular disease in the United States: an epidemiological overview. In: Saunders E, ed. Cardiovascular Diseases in Blacks. Philadelphia, Pa: FA Davis; 1991:3-14.
3. US Department of Health and Human Services. Report of the Task Force on Black and Minority Health. Volume IV: Cardiovascular and Cerebrovascular Disease, Part 2. Washington, DC: US Government Printing Office; 1986:229-284.
4. National Cholesterol Education Program. Classification, prevalence, detection, evaluation. In: Second Report of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Washington, DC: National Institutes of Health, National Heart, Lung, and Blood Institute; 1993:I-1-I-21.
5. Guralnick JM, Farmer ME, Blazer DG, Cohen HJ. Chronic conditions. In: Established Populations for Epidemiologic Studies of the Elderly. Washington, DC: National Institute of Aging; 1992;2:51-69.
6. Kumanyika SK. Obesity in black women. Epidemiol Rev. 1987;9:31-50.[Medline]
7. Keil JE, Sutherland SE, Knapp RH, Lackland DT, Gazes PH, Tyroler HA. Mortality rates and risk factors for coronary disease in black as compared with white men and women. N Engl J Med. 1993;329:73-78.[Abstract/Free Full Text]
8. Dunbar-Jacob J, Dwyer K, Dunning EJ. Compliance with antihypertensive regimen: a review of the research in the 1980's. Ann Behav Med. 1991;13:31-39.
9. Harburg E, Blakelock EH Jr, Roeper PJ. Resentful and reflective coping with arbitrary authority and blood pressure: Detroit. Psychosom Med. 1979;41:189-202.[Abstract]
10. Johnson EH, Schork NF, Spielberger CD. Emotional and familial determinants of elevated blood pressure in black and white adolescent females. J Psychosom Res. 1987;31:731-741.[Medline]
11. Anderson NB, Myers HF, Pickering T, Jackson JS. Hypertension in blacks: psychosocial and biological perspectives. J Hypertens. 1989;7:161-172.[Medline]
12. Anderson NB, McNeilly M, Myers HF. Autonomic reactivity and hypertension in blacks: a review and proposed model. Ethn Dis. 1991;1:154-170.[Medline]
13. Markowitz JH, Matthews KA, Kannel WB, Cobb JL, D'Agostino RB. Psychological predictors of hypertension in the Framingham Study. JAMA. 1993;270:2439-2443, 2494.[Abstract]
14. Schneider RH, Staggers F, Alexander C, Sheppard W, Rainforth M, Kondwani K, Smith S, King CG. A randomized controlled trial of stress reduction for hypertension in older African Americans. Hypertension. 1995;26:820-827.[Abstract/Free Full Text]
15. Magnus MH. Cardiovascular health among African-Americans: a review of the health status, risk reduction, and intervention strategies. Am J Health Promotion. 1991;5:282-290.
16. Dillbeck MC, Orme-Johnson DW. Physiological differences between Transcendental Meditation and rest. Am Psychol. 1987;42:879-881.
17. Alexander CN, Robinson P, Rainforth M. Treatment and prevention of drug addiction through Transcendental Meditation: an overview and statistical meta-analysis. Alcoholism Treatment Quarterly. 1994;11:11-84.
18. Eppley K, Abrams AI, Shear J. Differential effects of relaxation techniques on trait anxiety: a meta-analysis. J Clin Psychol. 1989;45:957-974.[Medline]
19. Alexander CN, Rainforth M, Gelderloos P. Transcendental meditation, self-actualization and psychological health: a conceptual overview and statistical meta-analysis. J Soc Behav Personal. 1991;6:189-247.
20. Nader T. Human Physiology: Expression of Veda and the Vedic Literature. Vlodrop, Holland: Maharishi Vedic University Press; 1994.
21. Maharishi Mahesh Yogi. On the Bhagavad-Gita: A New Translation and Commentary. Baltimore, Md: Penguin/Arkana; 1967: chaps 1-6.
22. Roth R. Maharishi Mahesh Yogi's Transcendental Meditation. Washington, DC: Primus; 1994:90-102.
23. Bernstein DA, Borkovec TD. Progressive Relaxation Training: A Manual for the Helping Professions. Chicago, Ill: Illinois University Press; 1973.
24. Jacobson E. Progressive Relaxation. Chicago, Ill: University of Chicago Press; 1992:28-100.
25. Joint National Committee. Fifth Report of the Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure (JNC V). Arch Intern Med. 1993;153:154-186.[Medline]
26. Multiple Risk Factor Intervention Trial Research Group. The Multiple Risk Factor Intervention Trial: risk factor changes and mortality results. JAMA. 1982;248:1465-1477.[Abstract]
27. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol. 1986;124:453-469.[Abstract]
28. Kaplan NM. Clinical Hypertension. Baltimore, Md: Williams & Wilkins; 1994.
29. Jackson J, Gurien G, Bowman P, Hatchett S. National Survey of Black Americans: A Study of Black American Life. Ann Arbor, Mich: Survey Research Center, University of Michigan; 1983:1085.
30. Spielberger CD, Jacobs G, Barker L. Preliminary Manual for the State-Trait Personality Inventory (STPI). Tampa, Fla: Center for Research in Behavioral Medicine and Community Psychology, University of South Florida; 1979.
31. Bradburn N. The Structure of Well-Being. Chicago, Ill: Aldine; 1969:1-120.
32. Hunt SM, McKenna SP, McEwen J. The Nottingham Health Profile: subjective health status and medical consultations. Soc Sci Med. 1981;15A:221-229.
33. Dixon WJ, Brown MB, Engelman L, Hill MA, Jennrich RI. BMDP Statistical Software Manual. Berkeley, Calif: University of California Press; 1990;2:1207-1244.
34. Schneider RH, Egan BM, Johnson EH, Drobny H, Julius S. Anger and anxiety in borderline hypertension. Psychosom Med. 1986;48:242-248.[Abstract]
35. Johnson EH, Nazzaro P, Gilbert DC, Weder A, Jamerson K. Similarities in cardiovascular reactivity to behavioral stressors in African-American and white males. Ethn Dis. 1992;2:232-245.[Medline]
36. Sherbourne CD, Stewart AL. The MOS Social Support Survey. Soc Sci Med. 1991;32:705-714.[Medline]
37. Seeman TS, Symes A. Social networks and coronary artery disease: a comparison of the structure and function of social relationships as predictors of disease. Psychosom Med. 1987;49:341-354.[Abstract]
38. Alexander CN. Transcendental meditation. In: Corsini RJ, ed. Encyclopedia of Psychology. 2nd ed. New York, NY: John Wiley & Sons; 1994;3:545-546.
39. Alexander CN, Chandler HM, Langer EJ, Newman RI, Davies JL. Transcendental meditation, mindfulness, and longevity: an experimental study with the elderly. J Pers Soc Psychol. 1989;57:950-964.[Medline]
40. Saunders E. Hypertension in blacks. Cardiovasc Clin. 1991;18:607-622.
41. Saunders E. Cardiovascular Diseases in Blacks. Philadelphia, Pa: FA Davis; 1991.
42. Hypertension Prevention Trial Research Group. The hypertension prevention trial: three-year effects of dietary changes on blood pressure. Arch Intern Med. 1990;150:153-162.[Abstract]
43. Kokkinos PF, Narayan P, Papademetriou V. Exercise training and hypertension in adult patients. Cardiology in the Elderly. 1994;2:433-438.
44. Kokkinos PF, Narayan P, Colleran JA, Pitteras A, Notargiacomo A, Reda D, Papademetriou V. Effects of regular exercise on blood pressure and left ventricular hypertrophy in African-American men with severe hypertension. N Engl J Med. 1995;333:1462-1467.[Abstract/Free Full Text]
45. Collins R, Peto R, MacMahon S, Fiebach HP, Eberlein KA, Godwin J, Qizilbash N, Taylor JO, Hennekens CH. Blood pressure, stroke in coronary heart disease, part II: short term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet. 1990;335:827.[Medline]
46. Thijs L, Fagard R, Lijnen P, Staessen J, Van Hoof R, Amery A. A meta-analysis of outcome trials in elderly hypertensives. J Hypertens. 1992;10:1103-1109.[Medline]
47. Alexander C, Barnes V, Schneider R, Langer E, Newman R, Chandler H, Davies J, Rainforth M. A randomized controlled trial of stress reduction on cardiovascular and all-cause mortality in the elderly: results of 8-year and 15-year follow-ups. Circulation. 1996;93:19. Abstract.
48. MacMahon S, Cutler J, Brittain E, Higgins M. Obesity and hypertension: epidemiological and clinical issues. Eur Heart J. 1987;8(suppl B):57-70.
49. Cutler JA, Follmann D, Elliot P, Suh I. An overview of randomized trials of sodium reduction and blood pressure. Hypertension. 1991;17(suppl I):I-27-I-33.
50. World Hypertension League. Physical exercise in the management of hypertension: a consensus statement by the world hypertension league. J Hypertens. 1991;9:283-287.[Medline]
51. Linden W, Chambers L. Clinical effectiveness of non-drug treatment for hypertension: a meta-analysis. Ann Behav Med. 1994;16:35-45.
52. Alexander CN, Robinson P, Orme-Johnson DW, Schneider RH, Walton KG. The effects of Transcendental Meditation compared to other methods of relaxation and meditation in reducing risk factors, morbidity, and mortality. Homeostasis. 1994;35:243-264.
53. Walton KG, Levitsky D. A neuroendocrine mechanism for the reduction of drug use and addictions by Transcendental Meditation. Alcoholism Treatment Quarterly. 1994;11:89-117.
54. Walton KG, Pugh NDC, Gelderloos P, Macrae P. Stress reduction and preventing hypertension: preliminary support for a psychoneuroendocrine mechanism. Journal of Alternative and Complementary Medicine. 1995;1:263-283.
55. Trials of Hypertension Prevention Collaborative Research Group. The effects of nonpharmacological interventions on blood pressure of persons with high normal levels: results of the Trials of Hypertension Prevention, phase I. JAMA. 1992;267:1213-1220.[Abstract]
56. van Montfrans GA, Karemaker JM, Wieling W, Dunning AJ. Relaxation therapy and continuous ambulatory blood pressure in mild hypertension: a controlled study. Br Med J. 1990;300:1368-1372.[Medline]
57. Alexander CN, Schneider RH, Rainforth M, King CG, Kondwani K, Staggers F, Sheppard W, Smith S. A randomized controlled trial of stress reduction for the treatment of hypertension in older African Americans: quality of life results. In: Proceedings of the Tenth International Interdisciplinary Conference on Hypertension in Blacks; St Thomas, US Virgin Islands; June 24-29, 1995. Abstract.
58. Orth-Gomer K, Unden AL. Type A behavior, social support, and coronary risk: interaction and significance for mortality in cardiac patients. Psychosom Med. 1990;52:59-72.[Abstract]
59. Kaplan GA. Where do shared pathways lead? Some reflections on a research agenda. Psychosom Med. 1993;57:208-212.[Abstract]
60. Mills PJ, Schneider RH, Hill D, Walton KG, Wallace RK. Beta-adrenergic receptor sensitivity in subjects practicing Transcendental Meditation. J Psychosom Res. 1990;34:29-33.[Medline]
61. Jevning R, Wallace RK, Beidebach M. The physiology of meditation: a review—a wakeful hypometabolic integrated response. Neurosci Bio Behav Rev. 1992;16:415-424.[Medline]
62. Cranson RW, Orme-Johnson DW, Gackenbach J, Dillbeck MC, Jones CH, Alexander CN. Transcendental Meditation and improved performance on intelligence-related measures: a longitudinal study. Personality and Individual Differences. 1991;12:1105-1116.
63. Elias AN, Wilson AF. Serum hormonal concentrations following Transcendental Meditation—potential role of gamma aminobutyric acid. Med Hypotheses. 1995;44:287-291.[Medline]
Institution: CTRO HOS COIMBRA HOSP GRAL Sign In as Member/Individual (Non-Member) Subscription Activation
Abstract of this Article ( )
Reprint (PDF) Version of this Article
Email this article to a friend
Similar articles found in: Hypertension Online PubMed
Search PubMed for articles by: Genest, J.
Alert me when: new articles cite this article
Download to Citation Manager
© 2001 American Heart Association, Inc.
Hypertension Electronic Pages
Progress in Hypertension Research
From the Clinical Research Institute of Montréal, Montréal, Quebec, Canada.
Correspondence to Jacques Genest, CC, GOQ, MD, Institut de Recherches Cliniques de Montréal, Clinical Research Institute of Montréal, 110 ave des Pins Quest, Montréal, Quebec H2W 1R7, Canada.
Abstract—— The author reviews the various factors (sodium, aldosterone, renin-angiotensin system, and norepinephrine; each of these factors being influenced by others) involved in the mechanism of human hypertension. A coherent picture is emerging, with the final pathway of these mechanisms converging on the renin-angiotensin system in the presence of a positive sodium balance and responsible for arteriolar resistance and responsiveness to pressor agents. This would correspond to the labile phase of hypertension, which leads with time to arteriolar restructuring and the increased media/lumen ratio, as demonstrated by Schiffrin and coworkers, and which can revert to normal structure with the administration of antihypertensive drugs such as converting-enzyme inhibitors, calcium-blocking drugs, and antagonists of the angiotensin II type 1 receptor. The author also presents the experience obtained in the Hypertension Clinic of the Clinical Research Institute of Montréal, which has been in existence since 1953; this experience is based on the observation of senior observers (clinical scientists, clinicians, and nurses) that the blood pressure of hypertensive patients can be controlled to normal levels in almost all cases for years and decades with a proper combination of the present antihypertensive drugs.
Key Words: sodium • aldosterone • renin-angiotensin system • norepinephrine
The extensive financial support given to the study of hypertension from the point of view of its mechanisms and management is based on its high incidence ( 25% of the population has blood pressure [BP] >140/90 mm Hg); the markedly decreased life expectancy, because of severe complications, if it is not treated; and the fact it is one of the largest healthcare expenditures.
In this presentation, we will deal with (1) the physiopathological molecular biology and genetic mechanism hypertension and (2) its modern management, which is one of the major advances in modern therapeutics.
Mechanisms of Hypertension
In our view of the mechanisms of essential hypertension (EH), there are 3 major landmarks. First in 1904, Ambard in France reported the BP lowering effect of a low-salt diet in hypertensive patients. Allen confirmed this in the period of 1918 to 1922. Later in the mid 1940s, similar observations were reported in patients on the rice-fruit diet of Kempner, who found this effect by serendipity. Dole, Dahl, and Cotzias showed conclusively in 1951 that this diet had a low-sodium content and that BP in patients with EH was often a function of the amount of salt ingested. Many others confirmed that hypertension could be corrected by a diet containing <230 mg of sodium per day to levels 140/90 mm Hg in 35% to 50% of patients with EH.
Second in 1934, Goldblatt and coworkers produced hypertension in dogs by narrowing 1 or 2 renal arteries by means of silver clamps. They attributed this effect to a substance that Tigersted and Bergman in 1898 called renin and to the pressor effects of crude renal extracts. This led to the renal experimental model of hypertension and a flurry of work in this field. It was quickly established that purified renin had no pressor effect by itself and acted on a plasma globulin called angiotensinogen (452 amino acids) to liberate the first 10 amino acids, thus yielding angiotensin (Ang) I, again an inactive substance. Renin is present principally in the juxtaglomerular cells of kidneys and, as demonstrated by Ganten and others, in the adrenals, brain, and many extrarenal tissues. The work of Skeggs demonstrated that the active pressor substance Ang II (an 8AA peptide) is liberated from Ang I by a converting enzyme, which released His 9–Leu 10 from Ang I in the presence of chloride ions. Ang III, a des-aspartyl heptapeptide of Ang II, has only 25% of the pressor activity of Ang II, but the major route goes directly from renin to Ang II.
Since the 1940s, this enzymatic cascade became known as the renin-angiotensin system, with Ang II being the most important pressor substance. The kidney was thus probably implicated in the pathogenesis of hypertension.
These findings formed the basis and starting point of our working hypothesis in 1945 that there were in EH at least 2 important factors involved: the renin-angiotensin system from the kidney and a dysregulation of sodium. It was generally believed that the key factor involved in the renal regulation of sodium was most probably of adrenal origin. In this search, Tait and Simpson succeeded in 1952 in isolating aldosterone from the amorphous fraction of adrenal extracts; Wettstein and associates at Ciba in Basle chemically characterized aldosterone in 1953.
Since these major contributions, many other pieces of evidence linked sodium to hypertension. Among them, some of the most important are (1) the effectiveness of a low-sodium (<230 mg/d) or rice-fruit diet in 35% to 50% of hypertensive patients with a decreased arteriolar responsiveness to norepinephrine and Ang II; (2) the effectiveness of natriuretic agents alone or in potentiating the effects of other anti-hypertensive drugs; (3) the absence of hypertension in populations on low-sodium or very low-sodium diets; (4) the prolonged pressure response to cross transfusion of blood from a renal hypertensive rabbit to a high salt-fed rabbit in contrast to the absence of any effect in rabbits fed a normal diet; (5) the increase of sodium concentration in red blood cells, leukocytes, and lymphocytes in patients with EH; and (6) the production of hypertension in a strain of rats on high sodium intake (Dahl rats).
Many observations showed a relation between aldosterone and hypertension: (1) a mean increase in urinary excretion of the 3-oxo conjugate metabolite in patients with EH, and (2) an increase in plasma aldosterone in 2-month-old spontaneously hypertensive rats (SHR) and a significantly enhanced response to adrenocorticotropic hormone compared with that of control rats (this latter finding was also observed in hypertensive patients). In human hypertensive patients compared with control subjects, there is (3) a blunted aldosterone response to severe sodium restriction or depletion; (4) a lesser suppression of aldosterone secretion rate, urinary secretion, and plasma levels on a high sodium intake (300 mmol/L per day); (4) an exaggerated plasma aldosterone increase in response to Ang II infusions; and (5) a marked increase in digital vascular reactivity in hypertensive patients when pretreated with aldosterone (1 mg IM for 3 days).
Third, with the finding that aldosterone was the key adrenal steroid involved in the renal regulation of sodium excretion, the search began for a connection between pressor substances (norepinephrine, Ang II, and others) and aldosterone secretion and salt. It was here that in 1959, our group made the key contribution that Ang II is the major stimulant of aldosterone production, whereas other pressor agents such as norepinephrine and phenylephrine, for example, were without any effect at infusion rates inducing similar increases in BP. These findings first reported in October 1959, and in early 1960 they were confirmed by Laragh and by the groups of Davis, Ganong, and Mulrow.1,2
Now, at last, we could integrate the renin-angiotensin system with the adrenal secretion of aldosterone and the renal regulation of sodium, which is now expressed as the renin-angiotensin-aldosterone-sodium (RAAS) system, to which so many researchers have devoted long and intensive efforts to clarify its various aspects and their relationships with other factors and to establish its fundamental role in human hypertension.
The role of the nervous system is mainly linked to the vasoconstrictor activity of norepinephrine, a major pressor substance liberated by the sympathetic nervous system and shown to be dependent on the sodium balance, being much less vasoactive under severe restriction or depletion and enhanced by high salt intake and aldosterone. The same vasoactivity dependence on salt balance was shown for Ang II. The norepinephrine plasma concentration in patients with EH is similar in most studies to that of normal subjects in the same age group. In addition, renin secretion is increased by infusions of norepinephrine, so that a relationship could be established between the sympathetic nervous system and the RAAS system.
Clinical experience during the past 50 years has taught me that the role of the nervous system (ie, of stress) is the single most frequent aggravating factor in patients with established EH. But it is not the causal factor, except for brief periods and under conditions of acute and severe stress. Such stress can provoke a marked rise of BP for days, weeks, and even months, sometimes culminating in complications of a cerebrovascular or coronary nature. The resolution of such stresses is accompanied by a return of the BP to normal levels or, in already hypertensive patients, to more a mildly elevated state. Experimentally, normalization of BP in normotensive subjects or animals occurs rapidly after cessation of prolonged pressor infusions of norepinephrine.
In the past 50 years, there has been much debate among researchers as to the respective part exercised by the components in a closed system made of a pump (heart), arteries of progressively decreased diameter, and blood volume. It was demonstrated that at times and in special conditions, an increased cardiac output (eg, thyrotoxicosis) or increased blood volume (eg, polycythemia) could be incriminated in these forms of hypertension. But it is now generally accepted that the key factor involved in EH is the increased peripheral resistance of the precapillary arterioles to the flow of blood ejected by the heart. It is at this level that the RAAS system plays a major role.
A major concern of many researchers has been to determine whether the structural changes in the precapillary arterioles were the cause of the increased BP or a consequence of it. It has been our view that the primary event in the early and labile phase of hypertension is an increased tonicity or stiffness and responsiveness of these small arterioles to the pressor activity of Ang II and norepinephrine as a consequence of an increased renal tubular reabsorption of sodium because of excessive sodium intake or of genetic mutation. This leads in the long term to structural changes (increased media width and media/lumen ratio) and a state of established hypertension. (This aspect was examined in great details in Genest3).
Many factors counteract the elevation of BP and the impaired sodium excretion by the kidney. They are the atrial natriuretic factor, vasopressin, bradykinin, NO, kallikrein, medullipin, and others. The role of endothelin, a potent vasoconstrictor produced by vascular endothelial cells, remains to be elucidated, as well as the interplay of all these factors at the level of the precapillary arterioles—the site of the increased peripheral resistance (Figure 1). But it is now more widely accepted that the hypertensive process expresses itself through the common final pathway of the RAAS system. A confirmation was recently provided by the work of Schiffrin and his group on the regression of the structural arteriolar changes in human EH after the chronic administration of antihypertensive drugs interfering with this system (see Management of Human Hypertension below).
View larger version (45K):[in this window][in a new window] Figure 1. View of the mechanism of EH, showing the relationship between sodium, the adrenals, the kidneys, and the nervous system. The major vasoconstricting influences are Ang II, which stimulates the aldosterone production, and norepinephrine. Their combined vasoactive effect is enhanced in the presence of a positive sodium balance due to increased renal reabsorption of sodium caused either by an excessive intake or by mutations of the genes controlling the mineralocorticoid receptor or those responsible for the sodium transport through the renal tubule. This would lead to the state of increased stiffness and responsiveness of the precapillary arterioles and the labile state of hypertension and would be followed with time by the "restructuring" of these arterioles and a more stable type of hypertension.
Now, after what might be called the physiopathological era, many researchers in the past few years have concentrated their efforts to elucidate the genetic factors involved in hypertension. In this field, Lifton’s group at Yale has been at the forefront of research in elucidating some of the genetic disorders in the rare mendelian, monogenetic forms of hypertension. As recently stated by Lifton, "molecular genetic studies have now identified mutations in 8 genes that cause mendelian forms of hypertension and 9 genes that cause mendelian forms of hypotension in humans (such as in Bartter syndrome). The mutated gene products in all cases act by altering the tubular sodium reabsorption".4 Some examples include the following.
Syndrome of Glucocorticoid-Remediable Aldosteronism
The syndrome of glucocorticoid-remediable aldosteronism was described in 1966 by the group of Laidlaw, who showed high aldosterone levels suppressed renin activity and hypokalemia and significantly increased urinary excretion of 18-oxo-cortisol and 18-OH-cortisol. Ectopic aldosterone secretion occurs in these cases in the adrenal fasciculara zone under the stimulation of adrenocorticotropic hormone (ACTH). Lifton and his group have shown that it is due to a chimeric gene, a crossover of aldosterone synthase, and 11-ß-hydroxylase genes and that it encodes a protein with aldosterone synthase activity regulated by ACTH. Administration of cortisol suppressed the secretion of ACTH and consequently of aldosterone, and spironolactone corrects the excessive renal sodium reabsorption and the hypertension. It has been strongly recommended that hypertensive children with a family history of hypertension and suppressed plasma renin levels should be screened for GRA.
Syndrome of Mineralocorticoid Excess
The hypertension syndrome of "mineralocorticoid excess" with low levels of renin and aldosterone and with hypokalemia is associated with an absence in 11-ß-hydroxysteroid dehydrogenase, which normally converts cortisol to cortisone, a steroid incapable of activating the mineralocorticoid receptor. This deficiency is responsible for the excessive unmetabolized cortisol, which exerts its mineralocorticoid activity and subsequently increases renal sodium reabsorption. A similar 11-ß-hydroxysteroid dehydrogenase deficiency is seen in patients with hypertension caused by high licorice ingestion, because of its content in glycyrrhetinic acid, and possibly with patients with Cushing’s syndrome.
Suppression of Plasma Renin Activity and Aldosterone
Another extremely rare hypertensive condition, first described by Liddle, consists of suppression of plasma renin activity and aldosterone, and responds to the administration of amiloride and triamterene, 2 inhibitors of the epithelial sodium channel (ENaC). Mutations in the ß- or -subunits of ENaC gene (on chromosome 16) were found to be responsible for the increased renal sodium reabsorption.
Such studies strongly suggest that the renin-angiotensin-aldosterone-sodium system and the renal mineralocorticoid receptor are the final pathways of hypertension process (Figure 2). It is a confirmation of the starting hypothesis of our own research work back in 1948 that sodium and the kidneys through its renin-angiotensin system were intimately involved in the basic mechanism of hypertension.
View larger version (62K):[in this window][in a new window] Figure 2. Using mostly date from Lifton’s group, this illustrates the renal reabsorption and the localization of the genetic defects leading to increased renal reabsorption at the level of the distal collecting duct and the mineralocorticoid receptor activated either by aldosterone or by cortisol, which has 5% the activity of aldosterone.
Despite these impressive advances in understanding the molecular basis for some of the rare monogenic forms, study of populations of patients with EH in whom the disease is known to result from the actions of multiple genes has been frustrating. As a result, there does not exist a predictive genetic test for hypertension risk in humans. Many excellent scientists are studying the genetics of hypertension and end-organ damage in animal models, and this work has the potential to identify new genes and new intervention targets in the future. With the advent of the sequence of the entire human genome, we can expect the pace of research on the genetics of EH to pick up dramatically over the next few years.
Efforts are being made in several laboratories throughout the world to unravel the genes responsible for EH, using predominantly linkage and association studies for testing for candidate genes. Similar efforts are made for the experimental types of hypertension, such as the Dahl rats and the SHR, in which some suspect overexpression of the renin gene.
Despite the great advances made in the past 50 to 60 years, fundamental questions remain. First, how does the RAAS system exert its effect at the arteriolar level and how do we unravel the complex interplay of the many factors (including neurogenic) influencing myogenic tone and contractile response to vasoconstricting and vasodilating agents in the ionic (sodium, potassium, magnesium, and calcium) environment of the precapillary arteriolar cells. Second, the fine tuning of the final sodium excretion in EH must be elucidated. The integration of the key factors involved (1) the mineralocorticoid receptors, their genetic mutations, and the sodium exchange and transport systems; (2) the major hormonal factor, ie, the RAAS system and the interplay of others such as atrial natriuretic factor, endothelin, NO, local prostaglandins, and kinins; (3) the reflex pressor sensors in the atria and pulmonary renins in the carotid sinus and the aorta arch and in the glomerular afferent artery, especially in relation to the role of the sympathetic nervous system and the release of norepinephrine; and (4) blood flow and renal perfusion pressure, including the glomerular filtration rate, the medullary blood flow, and the peritubular capillary hydrostatic pressure.
Management of Human Hypertension
The advances made in the management of human hypertension are even more impressive because of their importance in the lives of patients, their productivity, and the significantly decreased incidence of hypertensive complications and reduction of expenditures in the healthcare system.
Despite severe criticisms and shortcomings of the large clinical trials,5 since the first one done by Freis in 1967 and 1970 (there have been 15 more such large trials since), on the effects of specific antihypertensive drugs (several now used infrequently or not at all) on the mortality and morbidity of hypertension, they nevertheless demonstrated conclusively that even a small reduction of BP resulted in a significant decrease in the incidence of the severe complications such as cerebral hemorrhage, atherothrombotic stroke, myocardial infarction, congestive heart failure, and renal failure.2 But the major aim of these large trials was not the management of the individual patients with EH and the control of their BP to normal. This aspect has been left to the practicing physician.
Despite all the efforts and the extensive educational programs and the fact that there is a 25% incidence of Eh in North Americans, it remains a fact that only 16% (Canada) to 27% (USA) of patients are well controlled with BP <140/90 mm Hg. These findings contrast greatly when they are compared with the successful management of almost all hypertensive patients in specialized clinics with the antihypertensive drugs now available (see below). The success seen in these clinics indicates the importance for patients with moderate to severe EH to be managed by clinicians with a long experience in the field and who are best equipped because of their knowledge of the disease and of the pharmacology of these drugs. Numerous guidelines have been proposed for a more effective management by the practicing physicians.
There are many such hypertension clinics in university hospitals and in medical centers such as Mayo Clinic, Cleveland Clinic, and Oschner Clinic. But I would like to describe briefly the organization and results obtained at our own clinic at the Clinical Research Institute of Montreal and at the University of Montreal Hôtel-Dieu Hospital since 1953.
Our clinic was held once a week and then twice a week for several years, with 6 to 8 senior physicians and clinical scientists with a large experience in the field and with 4 or 5 specialized nurses who were trained in checking compliance to drugs and diet and in inquiring about their emotional and social aspects and were available at all times during the day for calls from patients. The clinical group also included a dietician and the active collaboration of cardiologists, nephrologists, endocrinologists, ophthalmologists, a vascular surgeon, a radiologist, urologists, and a pathologist. This clinic was a division of our Hypertension Group, which also had the control of a large section of the hospital clinical investigation unit, where patients could be investigated in full, and of research laboratories, which were under the same scientific direction. These laboratories provided the biochemical support for the determination of plasma renin activity, Ang II, steroids (aldosterone, cortisol, and others and their metabolites, progesterone, dehydroepiandrosterone) and catecholamine (norepinephrine, epinephrine, dopamine, and their metabolites). The whole constituted the Multidisciplinary Research Group on Hypertension supported for the past 3 decades by a Group Grant from the Medical Research Council of Canada (now the Canadian Institutes for Health Research). At the clinic, there was an average of 350 to 400 new patients referred per year and 700 to 1100 patients followed every year, for 3500 to 7000 patients visits per year.
I am firmly convinced on the basis of our experience that in "using what we know" about hypertension physiopathology (Lenfant6), >95% of patients with EH—whatever the initial level of their BP—can be controlled to normal levels for years and for more than a decade with the existing drugs with no or minimal side effects and with greatly improved quality of life provided that (1) these patients were thoroughly evaluated and secondary causes of hypertension were eliminated (renovascular, primary reninism, primary aldosteronism, unilateral renal atrophy glucocorticoid remediable aldosteronism [all associated with an overactive RAAS system], Cushing syndrome, pheochromocytoma, coarctation of the aorta, contraceptive pills, licorice), and (2) a judicious combination of antihypertensive drugs is used. Such combinations used since the 1960s permit a lower dosage of each drug and decrease the incidence of side effects.
One quite important aspect of such a clinic, besides the pharmacological expertise and clinical competence of the physicians, is the atmosphere provided by a personnel quick to smile and provide help, especially a staff of highly intelligent nurses actively participating and interested in their work. A consequence is a harmonious atmosphere and a close rapport with patients, who in return easily confide to the nurses about worries (such as family or work) and their emotional stresses. Compliance is easier to verify, and the patient becomes a friend and does not feel like he is a number!
The 5 groups of drugs most commonly prescribed at the present for the control of hypertension, are (1) the thiazide diuretics (1958), which occupy a central place in the modern treatment of hypertension; (2) the ß-blockers (1970s), which block norepinephrine release and decrease renin production; (3) the calcium-channel blockers (1979), which decrease the entry of calcium in arterioles; (4) the inhibitors of the converting enzyme (1977), which block the conversion of Ang I into Ang II, the effector component of the renin cascade; and (5) recently in the 1990s, the Ang II type 1 receptor antagonists. Other drugs such as -methyldopa, clonidine, hydralazine, and spironolactone can be added and may be quite beneficial in certain cases of hypertension. A new group of vasopeptidase inhibitors (specifically of neutral endopeptidases and ACE) is under intensive study at present.
Since 1953, we have made a systematic recommendation to all our patients to follow a diet without added salt (at the table or of salted food such as chips, salted peanuts, salted popcorn, soy sauce, bouillons, and frozen dinners), a diet restricted to about 3 to 4 grams of salt per day and with low animal fat (little or no butter, cream, mayonnaise, egg yolk, or cheese and elimination of meat fat and rich desserts). Such a diet was easy to follow and was recommended at the time on the basis of the work of Ahrens, Beveridge, Kinsell, and Mustard and their groups, who demonstrated in the 1950s the dietary link of animal fats to plasma cholesterol and platelet aggregation. It was also in line with the old observations of the effects of salt restriction by Ambard (1904), Allen (1918–1922), Kempner (1945), Dole (1946), and many others. It is important to recall that >60% of hypertensive people are salt sensitive and that dietary sodium restriction to 10 mmol/d was shown to lower BP to normal levels in 35% to 50% of hypertensive patients.
The effective management of most hypertensive patients implies more than the mere prescription of antihypertensive drugs and often requires important changes in lifestyle. Factors other than the genetic and physiopathological are frequently involved in the severity of hypertension and must be corrected and prevented. These factors are related to overeating (obesity); to excessive intake of salt, of "rich" foods, and of alcohol; and to the innumerable stresses of modern daily life. Their "management" by the patients requires self-discipline (or self-control), willpower, moderation, and a better adaptation or attitude toward various stresses—all qualities judged essential for a right way of life by the great Greek and Roman scholars of the Stoic School of Cicero and Seneca and by the great religions. It also implies (1) the acquisition of a greater degree of serenity, for which transcendental meditation can also be of help in decreasing the level of mental tension under which we live; and (2) daily moderate exercise such as walking, swimming, and cycling.
Finally, I must point to an additional advantage for our clinic in the management of our patients. We were greatly helped in our clinic by the healthcare system in Canada, which is free, ie, no direct cost to the patients for doctor visits, consultations, and tests (biochemical, hormonal, radiological, imaging, nuclear medicine, and others). In addition, at least in Quebec, a compulsory government drug insurance plan provides drugs at a greatly reduced cost or for free above a maximal contribution of $750.00 (Can).
Many studies in the past 20 years and the data from the National Center for Health Statistics (1999) have confirmed the marked decrease in the mortality rate from stroke and coronary heart disease by >65% and 53%, respectively. In our clinic, we have encountered in the past 15 years no case of cerebral hemorrhage, malignant hypertension, or advanced renal failure, thanks to the various education programs and the efficacy of new drugs. Severe complications, when they occurred, were the consequence not of the hypertension itself but of atherosclerosis (atherothrombotic stroke and coronary thrombosis). Is it necessary also to emphasize how much quality of life is improved in patients with BP controlled to normal levels who are free of the anxieties of possible and severe hypertensive complications?
The recent work of Schiffrin and his group is of special interest. Following the observations of Mulvany on arterioles and vascular remodeling, Schiffrin has demonstrated, in patients with mild EH and with BP controlled to normal levels by antihypertensive drugs for 1 or 2 years, a regression of the structural changes, ie, a decrease of the media thickness and media/lumen ratio to normal and of the excess collagen deposition. This structural normalization occurs after the administration of Ang II type 1 receptor antagonists, inhibitors of ACE, and calcium-channels blockers but not with ß-blockers, despite a similar decrease of BP to normal levels (Figure 3). These results suggest that the arteriolar structural changes are the result of the increased BP and not its cause.
View larger version (78K):[in this window][in a new window] Figure 3. Summary of the results obtained by Schriffin et al on the structural remodeling of the precapillary arterioles in hypertension. All the groups studied were composed of patients with mild EH, in whom the mean media/lumen ratio was 8.3% and significantly higher than those of normal subjects, whose mean media/lumen ratio was 5.9%. Under treatment for 1 or 2 years with cilazapril or perindopril (converting-enzyme inhibitors), losartan (antagonists of the Ang II type 1 receptor), and nifedipine Gits (a calcium-channel blocker), all groups had their BP returned to normal levels, the media/lumen ratio decreased to a level not significant from that of the control subjects. On the other hand, patients given atenolol for 1 or 2 years with BP controlled to normal did not show any significant change compared with hypertensive subjects. Slide is courtesy of Dr Ernesto L. Schiffrin.
The very great and rapid advancement of our knowledge of the mechanisms of hypertension and the effective management of the individual hypertensive patient are in my view two of the great accomplishments of modern medicine and a major contribution to public health. It is the result of close collaboration between clinical scientists, expert clinicians, and researchers in universities, research institutes, and pharmaceutical industry.
I wish to express my sincere thanks to Drs Timothy Reudelhuber, Daniel Bichet, and Ernesto Schiffrin, who have provided wise comments and help; to Claudia Jones and Martine Lauzier of the Documentation Center at the Clinical Research Institute of Montreal; to Isabelle Blain for her excellent secretarial support; and to Christian Charbonneau, head of the audiovisual department, for the illustrations.
Presented in part as the International Academy of Cardiovascular Sciences Award Lecture at the 17th World Congress of the International Society for Heart Research, Winnipeg, Canada, July 6–11, 2001.
Received July 27, 2001; first decision August 9, 2001; accepted August 16, 2001.
1. Genest J.c Mechanisms regulating the secretion and metabolism of aldosterone in experimental secondary hyperaldosteronism. Recent Prog Horm Res. 1961; 17: 293.
2. Genest J, Koïw E, Kuchel O. Hypertension: Physiopathology and Treatment, 1st ed. New York: McGraw-Hill; 1977: 260–265.
3. Genest J. Personal views on the mechanisms of essential hypertension.In: Genest J, Kuchel O, Hamet P, Antin M. Hypertension: Physiopathology and Treatment, 2nd ed. New York; McGraw-Hill; 1983: 599–614.
4. Lifton UP, Gharavi AG, Geller DS. Molecular mechanisms of human hypertension. Cell. 2001; 104: 545–556.[Medline]
5. Sever PS, Poulter NR. Hypertension drug trials: past, present, and future. J Hum Hypertens. 2000; 14: 729–738.[Medline]
6. Lenfant C. Interview with the editor. World Hypertension League Newsletter. 2001; 75: 1.
Posted by Emanuel at 5:13 PM