Essential hypertension

**********

Blood pressure (BP) is essentially variable; it fluctuates continuously because of the influence of spontaneous rhythmical variations associated with the functioning of body systems (eg, respiration, sleeping-waking cycle, seasonal variations) and the effects of superimposed physical and mental activity and of a large number of other behavioral and environmental factors (eg, posture, exercise, mood, ingestion of food and drink, smoking, talking). Thus, BP fluctuations may be of very short duration, from seconds to minutes, or of longer duration, and these variations differ from 1 subject to another. (1)

The hypothesis that individual differences in BP variability may play a role in the development of hypertension is not new, (2) but has generated great interest in the past 15 to 20 years because of the development and popularization of devices and techniques for measuring BP outside the physician's clinic, namely self-monitoring and, especially, ambulatory BP monitoring techniques. In comparison with clinical BP measurement, both techniques allow for a larger number of BP readings in more varied circumstances. In addition, these circumstances are more representative of the natural environment in which patients spend a majority of their day and, consequently, permit the measurement of several components of BP variability. Thus, long-term BP variability is now frequently evaluated by ambulatory monitoring or self-monitoring techniques and expressed, for example, as the standard deviation of BP readings taken every 10-30 min during 24-hour or daytime ambulatory monitoring (3-5) or as the standard deviation of BP readings taken every 8-12 hours during 2-4 week self-monitoring. (6)

Research on BP variability has evolved following 2 lines. The first line examined the hypothesis that individuals with increased BP variability would suffer more vascular damage. Mechanical factors may contribute to the cardiovascular damage associated with hypertension (eg, by causing blood vessels to burst in cerebral hemorrhage) and blood vessels may be relatively resistant to continuous levels of stress, but are more susceptible to intermittent stress. (7) Findings from several cross-sectional studies in essential hypertensive patients suggest a relationship between BP variability and target-organ damage in hypertension. (3,5,8-11) For instance, Parati et al. subdivided 108 hypertensive patients into 5 groups with progressively higher 24-hour mean BP. (3) For any given mean BP level, patients with greater BP variability (expressed as the 24-hour standard deviation of mean BP) displayed more severe target-organ damage than patients with comparable mean BP levels who had less variability. Confirming that result, Palatini et al. found an independent relationship between target-organ damage and the standard deviation of daytime ambulatory BE after controlling for differences in mean daytime BP levels. (5)

Given that patients with more advanced hypertension and more target-organ damage are likely to have a diminished baroreflex sensitivity and, therefore, increased variability, (1) results from cross-sectional studies do not allow drawing firm conclusions on whether BP variability plays a causal role in the cardiovascular morbidity associated with hypertension or whether it is simply its effect. However, at least 3 longitudinal studies seem to confirm the prognostic value of BP variability for cardiovascular complications or mortality associated with essential hypertension. Frattola et al. assessed 73 essential hypertensive patients at 2 occasions separated by more than 7 years. (4) At comparable 24-hour ambulatory mean BP levels, patients with a higher degree of 24-hour BP variability (expressed as the 30-min standard deviation of mean BP) at the initial evaluation displayed a greater severity of target-organ damage at the follow-up visit, both when target-organ damage was assessed as an overall score and when it was assessed as an increased left ventricular mass index. Consistently, after following 729 patients with mild hypertension for 5 years. Pickering et al. found that diastolic BP (DBP) variability (measured as the standard deviation of the daytime ambulatory readings) was an independent predictor of risk for cardiovascular morbidity. (12) Finally, Kikuya et al. carried out a long-term prospective study with 1,542 subjects from the general population and estimated BP variability as a standard deviation of BP measured every 30 minutes by ambulatory monitoring. (13) They found that daytime ambulatory systolic BP (SBP) variability was an independent predictor for cardiovascular mortality at the follow-up period (m = 8.5 years).

The second research line examined BP variability in the context of the reactivity hypothesis. This hypothesis states that individuals who show increased cardiovascular reactivity to psychologically stressful stimuli are at increased risk of developing hypertension. Thus, it has been suggested that situations assessed by an individual as stressful events initially produce transient elevations in BP by neurohormonal mechanisms, and that these elevations may induce structural changes in the vascular wall, which eventually result in a sustained increase in peripheral resistance and BE. (14) Studies on the reactivity hypothesis have largely focused on the assessment of BP changes induced by stressful tasks administered in laboratory settings (eg, cold pressor test or diverse mental arithmetic and reaction time tasks). However, ambulatory monitoring and self-monitoring of BP seem to be ideally suited to study the effects of the stresses of everyday life on BP. Several studies have examined the effects of potential sources of daily stress such as working, being married, and having children on BP with those techniques. (1)

In some of its forms, the reactivity hypothesis provides the rationale for most behavioral treatments for essential hypertension labeled stress management. This label encompasses many different interventions, from relaxation to other, varied components, including health education, various biofeedback modalities, assertiveness training, cognitive restructuring, stress-inoculation training, and so on. Various reviews and meta-analyses have concluded that stress management training is the most effective psychological treatment for essential hypertension to date. (15-17) All stress management training methods seem to share the goal of modifying patients" stress responses, especially their emotional stress responses. Therefore, it would be expected that stress management interventions could reduce long-term BP variability by decreasing or preventing stress responses and could eventually reduce the cardiovascular risk for hypertensive patients.

Our aim in this study was to examine the effects of stress management training on BP variability. To do this, we analyzed data from a controlled study we conducted earlier on stress management training for essential hypertension. (18) In that study, we treated 22 patients with essential hypertension in a stress management program and assigned another 21 patients to a waiting list control group. We took clinical and self-measured readings of SBP and DBP at pretreatment and posttreatment and took self-measured readings of SBP/DBP at the 4-month follow-up. At posttreatment, mean reductions of clinical BP (17/13 mm Hg vs 6.9/4.7 mm Hg for SBP/DBP), percentages of subjects who achieved at least a 5-mm Hg reduction (86/86% vs 48/48% for SBP/DBP), and percentages of subjects who additionally achieved a normotensive level (59/68% vs 29/14% for SBP/DBP) were significantly higher in the treated group than in the control group. Although the efficacy of the stress management training was not so considerable for self-measured BP (mean reductions of 3.6/2.4 mm Hg and percentages of subjects who achieved a 5-mm Hg reduction of 52/38% for SBP/DBP), it was (1) significant, (2) maintained in a 4-month follow-up assessment (mean reductions of 3.8/2.5 mm Hg and percentages of subjects who achieved a 5-mm Hg reduction of 48/33% for SBP/DBP), and (3) statistically greater than that of the waiting-list condition (mean reductions of .1/0 mm Hg and percentages of subjects who achieved a 5-mm Hg reduction of 17/6% for SBP/DBP).

In the present study, we were interested in determining whether this stress management training that seems to be efficacious in reducing BP levels is also able to reduce long-term or day-to-day variability assessed by BP self-monitoring. To differentiate the changes induced by treatment, that is, reductions in BP levels from reductions in BP variability, we measured this last parameter both in absolute terms by means of the standard deviation of BP self-measurements and in relative terms by means of the coefficient of variation of BP self-measurements (ie, normalized for the level of BP).

METHOD

Subjects

Over a 3-year period 65 male patients, diagnosed by their family physicians with essential hypertension, volunteered to participate in a research study on the behavioral treatment of essential hypertension after being referred by their primary care physicians for an exhaustive evaluation of their BE Physicians referred subjects if their BP was not well controlled (ie, DBP [greater than or equal to] 90 mm Hg or SBP [greater than or equal to]140 mm Hg) despite previous pharmacological treatments or despite currently taking 1 or 2 medications. We recruited all subjects from 2 health centers near Madrid. We considered potential patients for this research if (1) the diagnosis was confirmed by 3 measurements of either DBP > 90 mm Hg or SBP > 140 during 3 consecutive casual BP measurement sessions occurring over a 2- or 3-month period in the clinic, (2) the patient's physician agreed not to change the patient's dose of medication or patient's usual diet throughout the duration of the investigation, and (3) the patient's recent medical history elaborated by their primary care physician did not include a diagnosis of mental disorder according to the ICD-9 (International Classification of Diseases, 9th revision) system. (19) We asked all patients who met these criteria and who were referred by their family physicians to participate in this study. All of them provided consent. Nevertheless, we excluded 22 patients from the study: we excluded 17 patients because their family physician changed their medication in some point during the pretreatment assessment or during the treatment, 4 patients discontinued the study because they moved out of town, and 1 patient abandoned the study because of unknown reasons.

The final sample of this study included 43 patients. Table 1 shows a breakdown of demographic and clinical characteristics of the sample for the 2 groups (stress management and control). Thirty-two patients were on antihypertensive medications at the beginning of the study, but there were no significant differences in BP levels or BP variability between medicated and unmedicated patients at pretreatment (medicated patients: 151/100 and 130/83 mm Hg for clinical and self-measured SBP/DBP levels, respectively, and 8.6/6.3 mm Hg and 6.7/7.7% for SBP/DBP variability expressed as the standard deviation and the coefficient of variation, respectively; unmedicated patients: 146/98 and 130/86 mm Hg for clinical and self-measured SBP/DBP, respectively, and 7.1/6.4 mm Hg and 5.4/7.5% for SBP/DBP variability expressed as the standard deviation and coefficient of variation, respectively; all Fs ns). Table 2 displays medication status for the 2 groups. At follow-up assessment, we dropped 4 patients from the data analyses because of medication change during follow-up period, failure to observe self-measurement requirements, or unknown whereabouts. Of these patients, 1 was in the stress-management group and 3 were in the control group.

Procedure

The investigation consisted of 4 phases: (1) a pretreatment assessment of BP and of physiological and psychosocial hypertension-related variables; (2) a stress management training phase for the treated group and a waiting list for the control group; (3) a posttreatment assessment identical to the pretreatment assessment; and (4) a follow-up assessment of BR We obtained written informed consent for each phase.

Pretreatment, Posttreatment, and Follow-up Assessments

Before a patient was referred to the study and during a 2- or 3-month period, a nurse in thee patient's regular health ceter took 3 clinical measurements of BP (each measurement being the average of 2 or 3 clinic BP readings). After being referred by their physicians, we invited patients to an individual assessment of their hypertension, composed of 2 60-minute sessions. The first session was held in the patient's health center. Patients completed an interview that assessed several variables related to their hypertension problem (stressful events, duration of hypertension, adherence to medication, etc). Then, we carefully instructed patients on how to self-measure and self-record BP readings correctly and carried out several practice trials. We asked patients to self-measure their BP on 3 occasions per day for 16 days: 2 times at home (getting up in the morning and before bedtime) and 1 at work (see our previous study (18) for more details). We held the second session of the pretreatment assessment 8 days alter the first one, in the laboratories of the Universidad Complutense de Madrid. After completing a physiological assessment, we reviewed instructions for self-measurement of BR Two weeks after the end of stress management training, patients underwent a posttreatment assessment identical to that mentioned above. Patients visited the health center another 3 times over a 2- or 3-month period for the purpose of clinical BP readings and self-recorded their BP at 3 occasions per day for 16 days. Four months after the end of stress management training (follow-up assessment), we asked patients to serf-monitor their BP on 3 occasions per day for 8 days.

Compliance to the request for BP self-monitoring was high. We expressed this compliance, alter counting BP readings, as a percentage of the BP readings required for each period of self-monitoring. Patients' mean percentages of compliance were 99.5% for the pretreatment assessment, 99.3% for the posttreatment assessment, and 99.1% for the follow-up assessment (excluding the 4 patients dropped from the data analyses at follow-up).

Stress Management Training

After pretreatment assessment, we randomly divided all patients into 2 groups: 22 to the stress management group and 21 to the control group. The control group did not receive any intervention during the 2 months that the stress management training lasted. Stress management training consisted of 7 individual sessions on a once weekly basis, except for the last 2 sessions that were 2 weeks apart. Each session lasted 60 to 90 minutes and took place in the health centers. The stress management training involved 3 basic components: (1) information about the hypertension problem based mainly on a self-help booklet; (2) relaxation training closely following Bernstein and Borkovec's adaptation of Jacobson's progressive muscle relaxation, and (20) 3) problem-solving therapy following D'Zurilla's procedure. (21,22) Details of the stress management training can be found in previous studies. (18)

Materials

We gave each patient an OMRON HEM-403C digital BP monitor (Omron Healthcare Co., Ltd., Kyoto, Japan) to self-record his or her BE This semiautomatic device uses an oscillometric method to measure BP with an accuracy of [+ or -] 3 mm Hg and generates a digital display giving SBP, DBP, and pulse rate. There is a wealth of data supporting the accuracy of the OMRON HEM-403C devices for BP measurement when compared with simultaneously determined auscultatory values obtained with mercury sphygmomanometers. Imai et a1 (23) reported that the mean difference between the BP values obtained with a standard mercury sphygmomanometer and the OMRON HEM-401C device was 1.6 [+ or -] 6.7 mm Hg in SBP and 2.4 [+ or -] 6.1 mm Hg in DBP; thus it met the Association for the Advancement of Medical Instrument (AAMI) criteria. The OMRON HEM-403C device is the same device as the OMRON HEM-401C. O'Brien et al and Foster et al confirmed that the OMRON HEM-705CP and OMRON HEM-706 devices, respectively, whose measuring methods and algorithms for obtaining BP values are similar to those used with the HEM-403C device, met the A rank criteria of the British Hypertension Society as well as the AAMI criteria. (24,25) Before each assessment phase, we rechecked monitors for accuracy and, if necessary, they were recalibrated by the technical service representatives of manufacturers of the digital blood pressure monitors in Spain.

As part of their treatment, we gave subjects in the stress management training group the following additional material: (1) sell-help booklet mentioned earlier, (2) self-recording sheets to keep a record of medication and stressful events, (3) self-recording sheets to keep a record of recommended relaxation exercises, (4) audiocassettes with the relaxation instructions recorded with each individual patient during sessions with the therapist, and (5) self recording sheets and checklists for the problem-solving process for use by patients as reminders when a stressful event occurred in their lives.

STATISTICAL ANALYSES

We calculated the mean of self-measurements of SBP and DBP for each patient and for each evaluation period (pretreatment, posttreatment, and follow-up). Thus, at pretreatment and posttreatment, we defined SBP and DBP levels as the average of 48 home and work readings (3 daily readings for 16 days) and, at follow-up, we defined SBP and DBP levels as the average of 24 home and work readings (3 daily readings for 8 days). We considered the standard deviation and the coefficient of variation of the SBP/DBP readings for each individual and for each evaluation period to represent, respectively, the absolute and relative long-term variability of SBP/DBP, Thus, at pretreatment and posttreatment the intra-individual long-term variability was based on 48 home and work BP readings taken over 16 days, and at follow-up it was based on 24 home and work BP readings taken over 8 days. The available evidence suggests that these definitions provide reliable estimates of long-term (or day-to-day) BP variability. Glasgow et al obtained multiple daily, self-measured BPs from 254 hypertensive patients for 1 month (252 readings per subject) and found that the standard deviation of 12 BP measures taken over different days provides a reliable estimate of the standard deviation for an entire month. (26)

To evaluate differences between groups at pretreatment, we carried out analyses of variance (ANOVA) for each demographic and clinical variable assessed at pretreatment with groups of patients (stress management vs. control) as the independent variable. To analyze changes in BP variability from pretreatment to posttreatment, we carried out ANOVAs with repeated measurements for each dependent variable (the standard deviation and the coefficient of variation of SBP and DBP) with group of patients or period of the study (pretreatment vs. posttreatment) as independent variables. Similarly, to analyze changes in BP from pretreatment to follow-up, we carried out ANOVAs with repeated measurements for each dependent variable with group of patients or period of the study (pretreatment vs. follow-up) as independent variables. We considered p < .05 as statistically significant.

RESULTS

Demographic and Clinical Data

As shown in Table 1, ANOVAs carried out on each demographic and clinical variable assessed at pretreatment did not reveal any statistically significant differences on these variables between the stress management and control groups.

Changes in BP Variability at Posttreatment

Table 3 shows BP variability in absolute terms by the standard deviation and in relative terms by the coefficient of variation, for the stress management and control groups at the pretreatment, posttreatment, and follow-up. Mean BP variability of both groups were similar at pretreatment, F(1, 41) = .98, .02, 1.54, and .01 for SBP and DBP standard deviation, and SBP and DBP coefficient of variation, respectively: all ns.

SBP Variability. Mean decreases in the standard deviation of SBP were 1.77 and .72 mm Hg for the stress management and control groups, respectively. There was a trend for a significant decrease for the stress management group, F(1, 21) = 3.68, p < .069, whereas the decrease was not significant for the control group, F(1, 20) = 1.79, ns. However, a 2 (Group) x 2 (Period) ANOVA did not reveal a significant interaction between period and group, F(1, 41) = .97, ns, indicating that the decrease for the stress management group was not significantly greater than that showed by the control group. Likewise, there was a trend for a significant decrease (1.1%) in the coefficient variation of SBP for the stress management group, F(1, 21) = 2.79, p <. 10, whereas the decrease (.5%) was not significant for the control group, F(1, 20) = 1.27, ns. Again the ANOVA did not yield a significant interaction between period and group, F(1, 41) = .61, ns.

DBP Variability. The stress management group showed a trend for a significant decrease in the standard deviation of DBP, with a mean reduction of .92 mm Hg, F(1, 21) = 3.42, p < .079. By contrast, the control group showed a mean decrease in the standard deviation of DBP of .88 mm Hg that was not statistically significant, F(1, 20) = 2.40, ns. Nevertheless, the decrease was not greater for the stress management group than for the control group as revealed by a nonsignificant interaction between group and period, F(1, 41) = .01, ns. Concerning the coefficient variation of DBP, the stress management group showed a mean reduction of .84%, whereas the control group showed a mean reduction of 1.09%. The decrease was not significant for both groups: F(1, 21) = 2.10, ns for the stress-management group and F(1, 20) = 2.62, ns for the control group.

Changes in BP Variability at Follow-up

SBP Variability. Mean decreases in the standard deviation of SBP were 2.6 and .6 mm Hg for the stress management and control groups, respectively. The decrease was significant for the stress management group, F(1, 20) = 9.18, p < .007, but it was not significant for the control group, F(1, 17) = .13, ns. This pattern of results was statistically confirmed by a significant interaction between group and period, F(I, 37) = 3.98, p < .05, lending support to a greater reduction in SBP variability for the stress management group in comparison with the control group. We found comparable results when the SBP variability was expressed as the variation coefficients (ie, normalized for the level of SBP). The stress management group showed a statistically significant reduction of 1.8%, F(1, 20) = 8.41, p < .009, whereas the control group showed a nonsignificant decrease of .3%, F(1, 17) = .07, ns. Furthermore, the decrease was greater for the stress management group than for the control group as revealed by a significant interaction between group and period, F(I, 37) = 5.26, p < .03.

DBP Variability. The stress management group showed a mean reduction in the standard deviation of the DBP of 1.5 mmHg that was statistically significant, F(1, 20) = 5.61, p < .028. In contrast, the control group showed a nonsignificant decrease of 1.21 mm Hg, F(1, 17) = 2.10, ns. However, the interaction between period and group was not significant, F(1, 37) = .20, ns, indicating that the decrease for the stress management group was not significantly greater than that showed by the control group. Consistently, the stress management group displayed a mean reduction in the variation coefficient of DBP of 1.59%, a decrease that was statistically significant, F(1, 20) = 5.06, p < .036, whereas the control group displayed a nonsignificant decrease of 1.46%, F(1, 17) = 2.49, ns. Again, the interaction between period and group was not significant, F(1, 37) = .04, ns.

Because 48 BP sell-measurements defined the pretreatment and post-treatment BP variability, whereas the 24 BP self-measurements defined the follow-up, it might be that this difference could explain the significant results found when we carried out comparisons of BP variability between pretreatment and follow-up evaluation. Table 3 also shows BP variability estimates at pretreatment and post-treatment based on 24 BP self-measurements (ie, taking into account the BP readings recorded for the first 8 days). When we reanalyzed the data with these new definitions of pretreatment and posttreatment BP variability, we found results were similar to those mentioned above and we replicated the main findings. Thus, for the stress-management group, mean decreases in the standard deviation of SBP/DBP and in the variation coefficient of SBP/DBP were statistically significant, F(1, 20) = 7.18, 5.10, 5.76, and 3.92, respectively; p < .014, .035, .026, and .061, respectively, whereas such decreases were not significant for the control group, F(1, 17) = .41, 1.08, .10, and 1.30, respectively; all ns. Likewise, the interaction between group and period (pretreatment vs follow-up) was significant for the standard deviation and variation coefficient of SBP, F(1, 37) = 4.28 and 4.02, respectively; p < .046 and .05, respectively, but not for those of DBP, F(1, 37) = .10 and .01, respectively; all ns, indicating that for SBP but not for DBP the decrease in BP variability showed by the stress management group was significantly greater than that showed by the control group.

DISCUSSION

The results of the present study reveal that patients in the stress management group lowered their day-to-day SBP/DBP variability significantly from pretreatment to the 4-month follow-up, showing a mean reduction in the standard deviation of SBP/DBP of 2.6/1.5 mm Hg. Given that, in a previous study, (18) we reported that patients in the stress management group also lowered their clinical and self-measured SBP/DBP levels significantly, it might be thought that the BP variability reductions could be proportional to the reductions in self-measured BP levels, so that when we took into account changes in BP levels induced by treatment, the resulting variability at follow-up would be similar to that assessed before treatment. It is interestingly that the reduction was also evident when we expressed the variability as the variation coefficient, that is, when absolute variability was normalized for the level of pressure. Effectively, the stress management group also showed a significant decrease of 1.84/1.59% from pretreatment to the 4-month follow-up in the coefficient of variation of SBP/DBP, Therefore, this last effect could be regarded as a true reduction in long-term BP variability as a result of treatment. Furthermore, the results of this study show that patients in the stress management group lowered their SBP variability from pretreatment to 4-month follow-up more than those in the control group. Although the use of a waiting list control is not the most satisfactory way of strengthening the internal validity of our conclusions, that significant difference suggests that stress management training is efficacious in reducing day-to-day SBP variability in the long-term, beyond the effects of repeated BP self-measurement or spontaneous declines in BP variability over time.

We also found that patients in the stress management group lowered their BP variability from pretreatment to posttreatment (measured both as the standard deviation of BP and the coefficient of variation of BP), but these reductions were not statistically significant. The fact that the reductions in BP variability were significant at the 4-month follow-up, but not at the posttreatment, is consistent with the presumably progressive effects induced by a treatment based on the learning of relaxation and problem-solving skills. As a patient progressively uses and consolidates those skills to deal with daily stress, we would expect that their effects on reduction or prevention of stress responses were greater and, therefore, we would also expect that their effects on BP variability were more evident at the short-term follow-up than at posttreatment.

As far as we know, the present results constitute the first evidence from a controlled study that psychological therapy may reduce the degree of day-to-day BP variability controlling for changes in BP levels induced by treatment. Inasmuch as BP variability contributes independently from BP level to the cardiovascular risk in hypertension, (10,12-13) optimal antihypertensive treatment should aim at reducing not only BP level, but also daily life BP fluctuations. Given that all studies carried out to date indicate that most antihypertensive drugs have little effect on long-term BP variability, (1) the present findings coupled with previous results supporting the efficacy in reducing DBP/SBP level (18,27) suggest that stress-management training is efficacious in the treatment of essential hypertension, without known secondary effects, and should therefore be regarded as a therapeutic alternative, especially for mild hypertensives, for whom pharmacological therapeutic decisionmaking is more difficult, as it is unclear whether the benefits of medication exceed the costs. (28)

Caution is needed in attributing the reductions in BP variability found in the stress management group to stress management training alone. Part of its effects on BP variability may be attributable to its interaction with the pharmacological therapy that most patients received. As far as we know. no published study has examined the effects on BP variability of a combination of pharmacological and psychological interventions. Nonetheless, apart from interaction effects, it is not very plausible that the results of this study were only due to pharmacological therapy because, as mentioned above, most evidence suggests that current antihypertensive drugs fail to alter long-term BP variability. On the other hand, reductions in SBP variability from pretreatment to follow-up were significantly greater for the stress management group than for the control group, despite the fact that most patients of the control group also received medication. It should be emphasized that medication was constant throughout the study and that, in both groups, the most medicated patients received diuretics, angiotensin converting enzyme inhibitors, calcium-channel antagonists, or their combinations (see Table 2). As Pickering (1) has pointed out, these antihypertensive agents act independently of the autonomic nervous system, but it is supposed that long-term BP fluctuations associated with changes in physical and mental activity depend on the functioning of the sympathetic nervous system. Therefore, it might be expected that in most medicated patients of the present study, the antihypertensive drugs would not reduce BP variability. (1) Further research using at least 3 different conditions in 3 large groups (medication, stress management training, and medication + stress management training) could clarify the interaction between pharmacological and stress-management therapies concerning their effects on long-term BP variability.

Considering SBP variability reductions in the control group, we could approximate the efficacy of stress management training after excluding the effects of repeated BP self-measurements or spontaneous declines in SBP variability over time. Thus, after subtracting effects controlled by the waiting list condition, the reductions in day-to-day SBP variability for stress-management training alone could be estimated at 2 mm Hg when BP variability is expressed as the standard deviation of SBP, and at 1.5% when BP variability is expressed as the coefficient of variation of SBP. Tozawa et al (29) have prospectively studied the association between SBP variability and cardiovascular mortality and mortality from all causes in end-stage renal disease patients. (28) They found that the hazard ratio for death from all causes increased 1.63 times per 1% increase in the coefficient of variation in SBP. Although prospective studies with hypertensive patients are required to determine accurately the prognostic value of BP variability for cardiovascular mortality in hypertensive patients, Tozawa et al's (29) findings tentatively would support the clinical significance of BP variability reductions induced by stress-management training in this study. In this line, it should be pointed out also that a decrease of 2 mm Hg in SBP variability represents a reduction of 22.2% over the standard deviation of SBP shown by the stress management group at pretreatment. Moreover, a reduction of 22.2% over the standard deviation of SBP at pretreatment is at least similar to the reductions found in the few studies showing that some antihypertensive drugs can reduce long-term BP variability. For example, in diabetic hypertensive patients, lacidipine, a long-acting calcium antagonist, reduced 24-hour, daytime, and nighttime BP standard deviation by 19.6%, 14.4%, and 24.0%, respectively, (30) whereas in essential mild-to-moderate hypertensive patients, felodipine extended release reduced 24-hour SBP/DBP standard deviation by 17.8/17.5%. (31)

In sum, our results indicate that stress management training is an efficacious antihypertensive treatment for reducing day-to-day BP variability, especially SBP variability, and that its effects are evident at least for 4 months providing, in principle, an additional reduction in cardiovascular risk for hypertensive patients. Since most antihypertensive drugs are frequently unable to reduce long-term BP variability, replication of the present findings is an important issue for future studies.

ENDNOTES

(i.) Given that the stress management group kept records of medication, if further studies could prove the effectiveness of antihypertensive drugs in reducing BP variability, an alternative explanation would be that patients in the latter adhered to medication better or patients in the control group decided to reduce or eliminate medications of their own accord because they were getting "normal" readings at work and home. Although we only have medication compliance data from interviews carried out at pretreatment and posttreatment, the alternative explanations do not seem to be plausible. At posttreatment, there were no significant group differences in the medication adherence figures estimated by patients [92.1 [+ or -] 24.3 for the stress management group vs. 97.27 [+ or -] 4.7 for the control group; F(1, 26) = .48, ns]. In addition, a 2 (group) x 2 (period) ANOVA revealed no main or interaction effect for period factor, F(1,26) = .39 and .91, respectively, both ns; that is, no group changed its medication compliance from pretreatment to posttreatment (because of errors in the recording procedure of the posttreatment interview, we lost medication compliance data from 4 subjects: 3 subjects from the control group and 1 subject from the stress management group).

NOTE

This article was based on the doctoral dissertation carried out by the first author (MPG-V) under the direction of the third author (FJL) and with the support provided by a Formacion del Personal Investigador (FPI) Grant from the Comunidad Autonoma de Madrid. We are very grateful to Dr. Thomas Pickering from the New York Hospital-Cornell Medical Center and the late Dr. Richard Friedman from the University of New York at Stony Brook for their suggestions on the design of this study. We are indebted to Jose Maria Arribas Blanco, family physician from the Pozuelo de Alarcon Health Center, for his assistance with subject recruitment.

For comment and further information, please address correspondence to Maria Paz Garcia-Vera, Departamento de Personalidad, Evaluacion y Psicologia Clinica, Universidad Complutense de Madrid, Campus de Somosaguas, 28223 Madrid, Spain (e-mail: mpgvera@psi.ucm.es).

REFERENCES

(1.) Pickering, TG. Ambulatory Monitoring and Blood Pressure Variability. London: Science Press, 1991.

(2.) Hines EA, Brown GE. A standard test for measuring the variability of blood pressure: its significance as an index of the prehypertensive state. Ann Intern Med. 1933;7:209-217.

(3.) Parati G, Pomidossi G, Albini F, Malaspina D, Mancia G. Relationship of 24-hour blood pressure mean and variability to severity of target-organ damage in hypertension. J Hypertens. 1987;5:93 98.

(4.) Frattola A, Parati G, Cuspidi C, Albini F, Mancia G. Prognostic value of 24-hour blood pressure variability. J Hypertens. 1993:11:1133-1137.

(5.) Palatini P, Penzo M, Raccioppa A, Zugno E, Guzzardi G, Anaclerio M. Clinical relevance of night-time blood pressure and day-time blood pressure variability. Arch Intern Med. 1992; 152:1855-1860.

(6.) Imai Y, Nishiyama A, Sekino M, et al. Characteristics of blood pressure measured at home in the morning and in the evening: the Ohasama study. J Hypertens. 1999;17:889-898.

(7.) O'Rourke ME Basic concepts for the understanding of large arteries in hypertension. J Cardiovasc Pharmacol. 1985;7:S14-$21.

(8.) Mancia G, Parati G, Hennig M, et al. Relation between blood pressure variability and carotid artery damage in hypertension: baseline data from the European Lacidipine Study on Atherosclerosis (ELSA). J Hypertens. 2001 ; 19:1981-1989.

(9.) Roman MJ, Pickering TG, Schwartz JE, Pini, R, Devereux RB. Relation of blood pressure variability to carotid atherosclerosis and carotid artery and left ventricular hypertrophy. Arterioscler Thromb Vase Biol. 2001:21:1507-1511.

(10.) Pessina AC, Palatini P, Sperti G, Cardone L, Libardoni L, Mos L. Evaluation of hypertension and related organ damage by average day-time blood pressure. Clin Exp Hypertens. 1985:A7:267-271.

(11.) Gomez-Angelats E, de La Sierra A, Sierra C, Parati G, Mancia G, Coca A. Blood pressure variability and silent cerebral damage in essential hypertension. Am J Hypertens. 2004;17: 696-700.

(12.) Pickering TG, James GD. Ambulatory blood pressure and prognosis. J Hypertens Suppl. 1994:12:$29-33.

(13.) Kikuya M, Hozawa A, Ohokubo T, et al. Prognostic significance of blood pressure and heart rate variabilities: the Ohasama study. Hypertension. 2000:36:901-906.

(14.) Folkow B. Integration of hypertension research in the era of molecular biology: G.W. Picketing Memorial Lecture (Dublin 1994). J Hypertens 1995; 13:5-18.

(15.) Jacob RG, Chesney MA, Williams DM, Yijun Ding BS, Shapiro AR Relaxation therapy for hypertension: design effects and treatment effects. Ann Behav Med. 1991:13:5-17.

(16.) Linden W, Chambers L. Clinical effectiveness of non-drug treatment for hypertension: a meta-analysis. Ann Behav Med. 1994;16:35-45.

(17.) Johnston DW: The behavioral control of high blood pressure. Curr Psychol Res Rev. 1987;6:99-114.

(18.) Garcia-Vera MR Labrador F, Sanz J. Stress management training for essential hypertension: a controlled study. Appl Psychophysiol Biofeedback. 1997;22:261-283.

(19.) World Health Organization. International Classification of Diseases, 9th revision. Geneva, Switzerland: World Health Organization: 1998.

(20.) Bernstein DA, Borkovec TD. Progressive Relaxation Training. Champaign, IL: Research Press, 1973.

(21.) D'Zurilla TJ. Problem-solving training for effective stress management and prevention. J Cognit Psychother Int Q. 1990;4:327-353.

(22.) D'Zurilla TJ. Problem-solving Therapy. New York: Springer, 1986.

(23.) Imai Y, Abe K, Sasaki S, Minami N, Munakata M, Sakuma H. Clinical evaluation of semiautomatic and automatic devices for home blood pressure measurement: comparison between cuff-oscillometric and microphone methods. J Hypertens. 1989:7:983-990.

(24.) O'Brien E, Mee F, Atkins N. Thomas M. Evaluation of three devices for self-measurement of blood pressure according to the revised British Hypertension Society Protocol: the Omron HEM-705CR Philips HP5332, and Nissei DS-175. Blood Press Monit. 1996:1:55-61.

(25.) Foster C, McKinlay S, Cruickshank JM, Coats AJS. Accuracy of the OMRON HEM 706 portable monitor for home measurement of blood pressure. J Hum Hypertens. 1994:8:-664.

(26.) Glasgow MS, Engel BT, D'Lugoff BC. The use of successive blood-pressure measurements to estimate blood-pressure variability. J Behav Med. 1988;11:435-446.

(27.) Garcia-Vera MR Sanz J, Labrador FJ. Psychological changes accompanying and mediating a stress-management training for essential hypertension. Appl Psychophysiol Biofeedback. 1998:23:159-178.

(28.) Schechter CB. Sequential decision making with continuous disease states and measurements. 11. Applications to diastolic pressure. Med Decis Making. 1990:10:256-265.

(29.) Tozawa M, Iseki K. Yoshi S, Fukiyama K: Blood pressure variability as an adverse prognostic risk factor in end-stage renal disease. Nephrol Dial Transplant. 1999:14:1976-1981.

(30.) Frattola A, Parati G, Castiglioni R et al. Lacidipine and blood pressure variability in diabetic hypertensive patients. Hypertension. 2000:36:622-628.

(31.) Fariello R, Boni E, Corda L, et al. Extended release felodipine in essential hypertension. Variations in blood pressure during whole-day continuous ambulatory recording. Am J Hypertens. 1991:4:27-33.

Drs Garcia-Vera, Sanz, and Labrador are with the Universidad Complutense de Madrid in Madrid, Spain. Drs Garcia-Vera and Sanz are associate professors, and Dr Labrador is a full professor.

COPYRIGHT 2004 Heldref Publications
COPYRIGHT 2005 Gale Group