High blood pressure (BP) is the leading risk factor for global morbidity and mortality. Clinical trials in hypertension have largely focused on the efficacy of oral medications to lower BP and reduce cardiovascular events. However, a variety of other nonpharmacologic treatments have been developed and studied with varying degrees of scientific rigor. The focus of this chapter is to review the BP-lowering efficacy of therapeutic approaches that are alternatives to medications and dietary (or herbal) interventions. The American Heart Association (AHA) recently published in 2013 a comprehensive scientific statement in this regard and highlighted the evidence for, or against, use of these approaches in clinical practice. This chapter summarizes the major conclusions reached by the AHA ( Table 29.1 ) and provides an updated review of randomized clinical trials (RCTs) of at least several weeks’ duration and that focus primarily on BP and meta-analyses subsequently published that could substantially impact this field of clinical medicine.
| Alternative Treatments | Level of Evidence a | Recommendation b | Meta-Analyses Published Since the 2013 Scientific Statement | Selected Trials Published Since the 2013 Scientific Statement |
|---|---|---|---|---|
| Behavioral Therapies | ||||
| Transcendental meditation | B | IIB | Refs | |
| Other meditation techniques | C | III (no benefit) | Ref | |
| Biofeedback approaches | B | IIB | ||
| Yoga | C | III (no benefit) | Refs | Refs |
| Other relaxation techniques | B | III (no benefit) | ||
| Noninvasive Procedures or Devices | ||||
| Acupuncture | B | III (no benefit) | Refs | |
| Device-guided breathing | B | IIA | Refs | Ref |
| Exercise-Based Regimens | ||||
| Dynamic aerobic exercise | A | I | ||
| Dynamic resistance exercise | B | IIA | ||
| Isometric handgrip exercise | C | IIB | Refs | |
a A, Data from multiple random controlled studies (RCTs) and/or meta-analyses; B, Data from a single RCT or observational studies; C, Case studies or standard of care.
b I, Treatment SHOULD be performed; IIA, It is REASONABLE to perform the treatment; IIB, Treatment can be considered; III, Treatment is not helpful (or harmful) and should not be performed
It is important to begin with a cautionary note that many (if not all) of the alternative approaches described in this chapter present several common challenges to researchers. One principal difficulty is the lack of consensus about the appropriate sham or placebo comparator. In principle, a sham or placebo should be similar to the active treatment in all respects (and not be discernable by the study participants), except for the “active ingredient” in the treatment. However, this ideal is not easily implemented or even imagined in many instances. In addition to randomization, gold standard comparators and blinding are crucial elements of robust clinical studies. These elements have not necessarily been successfully achieved in the large variety of studies related to alternative approaches for high BP, even though this review is limited to RCTs and does not involve observational studies. Numerous other potential biases (e.g., cointervention, Hawthorne effect) have often plagued many published trials to date and may thus limit the broader generalization of their findings. Finally, the BP-lowering efficacy of most of the alternative approaches has only been investigated over the short term (few weeks to months) and their benefits in regards to reducing hard cardiovascular events has rarely been evaluated.
Mirroring the AHA scientific statement, we have divided this chapter into four classes of alternative approaches to high BP: behavioral therapies, noninvasive procedures and devices, exercise-based regimens, and other additional noninvasive interventions. Further information and more detailed methodologic descriptions pertaining to the individual approaches can be found in the original AHA scientific statement.
Behavioral Therapies
Meditation
There is a large variety of meditation types, and an appropriate placebo control is difficult to identify for any of them. The “active ingredient(s)” among the various techniques is equally difficult to verify given that these ancient approaches were not specifically formulated with the intent of lowering BP. Thus, it is difficult to design a viable comparator placebo intervention that lacks only the active ingredient (i.e., without changing the entire experience dramatically). Blinding is also problematic, for example, in instructor-led meditation, because either the same instructor must teach two methods (one active and one “placebo”) or the instructors must vary. Therefore, these studies often suffer from numerous important limitations. Although acknowledging these inherent shortcomings, numerous meditation approaches have been investigated over the past several decades with respect to their ability to decrease BP.
Transcendental Meditation
One particular form of “mantra-based” meditation, transcendental meditation (TM), has been studied for its effects on a multitude of health-related measures, including high BP. Most studies have compared TM with health education, relaxation, wait-list control, or no treatment. Several attempts have been made to summarize the efficacy of TM in regards to lowering BP. In 2004, a meta-analysis suggested the available studies were of inadequate quality for any conclusion to be drawn. A 2007 review and synthesis of prior and subsequent studies concluded that TM lowered both systolic BP (SBP) and diastolic BP (DBP) compared with progressive muscle relaxation; however, it was not superior to health education.
A few subsequent meta-analyses have concluded that TM lowers systolic and diastolic BP compared with control interventions. An important caveat is that meta-analyses necessarily involve many decisions that can unintentionally influence the results, most notably decisions about which studies meet quality standards for inclusion. Thus, the risk of unintentional biases is high. Nevertheless, the most recently published meta-analysis (12 studies; n = 996) to evaluate the effect of TM found evidence for a modest but significant BP-lowering effect (−4.3/2.3 mm Hg) compared with controls. Contrarily, a recent Cochrane review concluded that only two studies could be included in their meta-analysis regarding BP and that excessive trial heterogeneity did not allow for the combining of further data. Therefore, the authors stated that evidence regarding the BP-lowering efficacy of TM should be viewed as “suggestive” at this point in time. Finally, in one of the few long-term trials beyond a few months’ duration, BP remained stable (i.e., significantly lower) in the TM group compared with that in participants randomized to health education, in whom systolic BP rose during the 5.4 years of average follow-up. As such, the overall evidence supporting the efficacy of TM in regards to controlling high BP and reducing cardiovascular risk is modest and requires further investigation before reaching firm conclusions (see Table 29.1 ).
Other Forms of Meditation
The potential BP-lowering efficacy of Zen meditation has also been investigated. A 2007 review and synthesis of data concluded that Zen meditation reduced DBP, but not SBP when compared with repeated BP checks. It should be noted that comparison to repeated BP checks is an appropriate way of controlling for “regression toward the mean.” However, it is not equivalent to comparing Zen meditation with a different effective treatment or to the effect of placebo or sham on BP. As such, weaker evidence exists in support of Zen meditation compared with TM for BP-lowering (see Table 29.1 ). We did not find any RCTs specifically regarding Zen meditation published since the AHA scientific statement.
In contrast to the ancient tradition of Zen meditation, Mindfulness-Based Stress Reduction (MBSR) is a program of contemplative meditation developed in more recent years. The recent HARMONY study was an RCT of MBSR versus wait-list control among 101 adults with untreated stage 1 hypertension. In this well-performed contemporary trial, MBSR did not significantly lower ambulatory BP levels. Moreover, Park et al performed a randomized, controlled crossover trial comparing mindfulness meditation, BP education, and controlled breathing in 15 African Americans with hypertension in the setting of chronic kidney disease. SBP, DBP, mean arterial pressure, heart rate, and muscle sympathetic nerve activity decreased more in the mindfulness meditation condition compared with the BP education control condition. Controlled breathing did not affect these parameters. We identified no other germane RCTs of meditation technique published since the AHA statement.
If one concludes from the preceding data that some forms of meditation lower BP, the question of mechanism of action becomes relevant. This question remains open at the present time; however, it is possible that reductions in sympathetic nervous system activity might be involved. In summary, the AHA scientific statement conferred to TM a Class IIB, level of evidence B recommendation for BP-lowering efficacy. They rated all other forms of meditation a Class III, no benefit, level of evidence C (see Table 29.1 ). Although some small studies have been published since, our review did not identify RCTs or meta-analyses regarding meditation techniques (other than TM) that alter these conclusions.
Biofeedback Techniques
Biofeedback prototypically involves monitoring of BP and/or one or more putative surrogates for BP or other closely linked cardiovascular parameter (e.g., galvanic skin response, heart rate variability), such that mental states favoring lower BP can be identified and, ideally, recalled and reproduced at will. As with meditation, it is hard to synthesize the trial results into a simple statement of efficacy. The same issues of heterogeneity in study design and biofeedback approaches used, lack of blinding, and lack of consensus about appropriate negative control interventions preclude a simple analysis.
A meta-analysis that included some practitioners of TM is one of two relatively recent meta-analyses reporting that biofeedback does not lower BP. Of note, two other systematic reviews reported different conclusions about whether biofeedback reduces BP. The review published in 2003 reported that biofeedback lowers BP more than does nonintervention. In contrast, a 2010 systematic review performed with stricter inclusion criteria reported no effect of biofeedback on hypertension compared with a variety of negative controls.
A few trials have been performed since these meta-analyses were published as outlined by the AHA scientific statement. In one of the most notable studies, 65 participants were randomized to behavioral neurocardiac training (heart rate variability biofeedback) and behavioral relaxation or repetitive visualizations as a control for two months. Daytime (−2.4 mm Hg) and 24-hour (2.1 mm Hg) SBP were reduced by biofeedback with no effect in the control group. Given the mixed results and the large degree of variability between the numerous different biofeedback methodologies, the authors of the AHA Scientific Statement on Alternative Approaches to Lowering Blood Pressure assigned biofeedback a Class IIB, level of evidence B recommendation for lowering BP (see Table 29.1 ). We did not identify any subsequently published RCTs or meta-analyses devoted to biofeedback that significantly alter this rating.
Yoga
Yoga is practiced in many forms, which may involve quiet contemplation or physically strenuous activity. Several systematic reviews of yoga for hypertension have recently been published since the AHA statement. The meta-analyses generally support that the available studies to date are of low quality or suffer from methodologic variations and limitations that do not allow for firm conclusions to be made in regards to the independent BP-lowering efficacy of yoga techniques at the present time. In analyzing 17 RCTs, Posadzki et al concluded that the evidence in favor of an effect on BP was “encouraging but inconclusive.” In analyzing seven RCTs with 452 participants, the authors judged the available evidence to be of low quality. With this caveat, they noted that compared with usual care, yoga lowered BP. However, compared with exercise, there was no effect of yoga on SBP or DBP. Another meta-analysis published in 2016 concluded that yoga nonsignificantly reduced SBP by −5.21 mm Hg (95% confidence interval [CI], −8.01 to 2.42) and DBP by −4.98 mm Hg (−7.17 to 2.80).
Unlike many of the other alternative approaches reviewed in this chapter, there has been a recent flurry of RCTs studying the effect of yoga on BP. Wolff randomly assigned primary care patients with hypertension to home-based Kundalini yoga (n = 96) or usual care (n = 95) for 12 weeks. The BP reduction in the yoga group was not different from in the control group. The recent LIMBS (LIfestyle Modification and Blood Pressure Study) trial randomized 137 participants to yoga, BP education, or both. At the end of the 24-week trial, there was no difference in BP lowering between the yoga and control arms. Among the 90 participants who completed the study, the 24-week BPs marginally favored the BP education program over yoga. There was no additive benefit of combining yoga and BP education.
Siu et al randomized 182 patients with metabolic syndrome to yoga or monthly telephone contact for 1 year. There was a trend to greater improvement in SBP in the yoga arm, although this difference was not statistically significant ( p = 0.07). In another recent RCT, 171 underactive adults with metabolic syndrome were randomized to yoga or a program of stretching, there was no difference in SBP between the two arms at 6 months or 12 months. Hagins et al randomized 84 participants with prehypertension or stage 1 hypertension to yoga or nonaerobic exercise. There was no significant difference in 24-hour SBP or DBP, daytime SBP or DBP, or nocturnal SBP. The nighttime DBP was marginally different ( p = 0.04). In view of the number of analyses performed, there is reason to be concerned that this difference arose by chance. The effect of nonaerobic exercise on hypertension is a relatively unexplored area, so its usefulness as a control intervention is limited. Mechanisms whereby yoga might lower BP are difficult to study rigorously, and we do not yet have satisfactory answers. Despite the variable and occasional positive trend from meta-analyses, numerous limitations exist in regards to the individual trials evaluating the effect of yoga on BP. The more recently published studies have also not provided encouraging findings in regards to BP lowering. Finally, similar to biofeedback, yoga cannot be described as a homogeneous practice. There are many methods involving a variety of different aspects and some practices may be effective, whereas others may not provide the necessary activity (e.g., exercise, breathing, mental state) required to lower BP. The AHA scientific statement concludes that the overall evidence does not support that yoga per se lowers BP (Class III, no benefit, level of evidence C). Based upon the subsequently published RCTs and meta-analysis, we did not find any persuasive evidence to support upgrading the efficacy of yoga at the present time (see Table 29.1 ). Given the occasionally positive results, further studies would be helpful, in particular if they help to identify the most effective aspects of any practice. The independent actions of yoga per se, beyond the exercise components, also require clarification.
Other Relaxation Techniques
Beyond meditation, biofeedback, and yoga, a broad range of relaxation methods have been studied to see whether they lower BP in the long term. The Hypertension Intervention Pooling Project analyzed the results of 12 RCTs. The authors concluded that relaxation methods reduce DBP by a small amount, but do not reduce SBP. A review in 1991 found that baseline characteristics of patients predicted response to relaxation methods; after controlling for these baseline characteristics, little evidence for a response remained. There was evidence that regression toward the mean, rather than an effect of relaxation, was responsible for decreases in BP. A 1993 review of the literature found that relaxation alone did not reduce BP compared with appropriate sham controls. However, BP appeared to be reduced when compared with no treatment, highlighting again the need for appropriate sham controls. Yet another review in 1994 reported that interventions including more than just stress reduction were more effective than those with stress reduction alone.
A 2008 Cochrane review found that in contrast to poor quality studies, higher quality studies showed relaxation to cause a smaller decrease in BP or even a possible increase in BP. For example, in studies that included a sham control, there was no significant reduction in BP. The authors were unable to determine whether any particular method of relaxation is effective for lowering BP in the long term. We will refrain from speculating about mechanisms because a long-term BP lowering effect of relaxation has not been convincingly demonstrated to exist in the first place. The AHA writing group viewed the data as consistent with a Class III, no benefit, level of evidence B recommendation (see Table 29.1 ). We found no recent evidence to support altering this conclusion.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
