Relaxation lowers blood pressure: a genetic mechanism?

​Studies have suggested that the Relaxation Response (RR) lowers blood pressure hypertensive patients. A new article assesses the potential mechanism, suggesting that relaxation-related effects on the innate immune system may well be associated with the antihypertensive effects of relaxation (see http://www.wbur.org/commonhealth/2018/04/06/harvard-study-relax-genes  ; for article: see bp genetic resp to relaxation jaltcompmed2018 in dropbox, or DOI: 10.1089/acm.2017.0053​​).

Details:
--8-week intervention on 24 non-medicated patients with BP 140-159/90-104, defined rigorously  as 3 weekly run-in visits: after patient seated at rest for 5 minutes, 3 BP measurements were taken at least 2 min apart, and if SBP varied >8mmHg, continue checking until values within this range, then average them
--baseline: 56 yo, HTN duration 5 yrs, 71% female, 50% married, 83% white, 75% employed, 63% completed college
--baseline BP: 143.7/90.9
--patients then had 8 weekly individual training sessions from an RR trainer (Relaxation Response), including RR diaphragmatic breathing, body scan, mantra repetition, and mindfulness meditation, while passively ignoring intrusive thoughts. and given 20-minute audio CD to listen to at home daily.
--RR responders defined rigidly as both SBP decrease of at least 10 mmHg and DBP decrease of at least 5mmHg
    --13 were "responders", 11 "non-responders"
--patients completed the Perceived Stress Scale (PSS), Beck Depression Inventory-II (BDI-II), and Beck Anxiety Inventory (BAI)
--blood drawn for genetic analysis (transcriptome analysis in peripheral blood mononuclear cells) at baseline and after the 8 weeks, looking at specific changes in those who were responders vs nonresponders

Results:
--Responders: mean BP change 143.9/90.7 to 126.9/75.4 (ie 17/15.3 mmHg)
--Nonresponders: mean BP change 143.5/91.2 to 136.6/85.8 (ie 6.9/5.4 mmHg)
--Psych tests:
    --BAI:
        --responders: decreased from 4.83 to 2.83, p=0.047, statistically significant, a moderate clinical response
        --nonresponders: decreased from 5.9 to 3.4, p=0.195, not statistically significant
    --BDI-II:
        --responders: decreased from 4.08 to 2.62, p=0.054, borderline statistically significant, a moderate clinical response
        --nonresponders: decreased from 7.36 to 5.91, p=0.195, not statistically significant
    --PSS:
        --responders: decreased from 12.9 to 10.5, p=0.107, not statistically significant
        --nonresponders: decreased from 16.91 to 13.45, p=0.124, not statistically significant​
        --but, for the group as a whole, this would be considered a moderate clinical response, with p=0.02 [as noted below, this study is limited statistically because of very few participants, but if the 2.8 difference between baseline and 8 weeks later held for a larger group, this would be considered to be moderately clinically effective]
--Transcriptome analysis: 1387 transcripts were significantly different between baseline and at 8 weeks after RR training, including overrepresentation of upregulated genes in pathways linked to cell proliferation and survival, T and B cell lymphocyte signaling and energy production.
    --more differences in transcript, comparing pre and post RR training, were noted in responders (1821) vs nonresponders (1280)
        --responders had  decreased inflammation both by upregulating anti-inflammatory pathways (ie glucocorticoid receptor signaling) and downregulating Nuclear Factor-kB (NF-kB, which affects both immunologic and inflammatory responses) and TNF pathways
        --nonresponders actually had significant upregulation of multiple immune and inflammation-related activation
        --responders also had transcripts associated with both lipid and carbohydrate metabolism as well as circadian rhythm and vascular endothelial growth factor

Commentary:
--there are many, many studies showing that lifestyle changes (weight loss, exercise, low salt diet) are about as effective as meds in treating hypertension, and some studies finding that mind-body interventions can be effective as well. for example, a study by researchers in the above study found that RR enabled the elimination of at least one BP med in 32% of participants
--in the above study the level of blood pressure lowering in the 1/2 of patients who responded to relaxation is on at least a par with medication response!!! and very few adverse effects with relaxation...
--overall, responders did have a clinically significant decrease in measures of depression and anxiety, though a nonsignificant decrease in perceived stress. I should point out that this was a very small study, so surprising that any of these results were statistically significant; both responders and nonresponders had decreases in all of these scores, all in the “moderately clinically significant range”; and that even nonresponders had some antihypertensive response.  These changes suggest an association (cannot prove causality) between decreases in these psychological metrics and blood pressure response
    ​--the above baseline depression and anxiety assessment in nonresponders was higher than in responders. not sure this is significant, but would be interesting to know if this were a true differentiator of responders vs nonresponders in a much larger study
--as a reference point, however, the mean depression score at baseline was around 5 (where score of  0-13 indicates minimal depression); and the mean anxiety score at baseline was 5​ (where score of 0-9 means normal to minimal anxiety). So, these were not particularly depressed or anxious patients at baseline.
--stress response involves activation of the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system. Acutely, these changes can lead to a hypertensive response, with some suggestive studies finding that chronic stress leads to chronic activation of these systems and chronic hypertension. Also, the increased cortisol from stress affects essentially every other hormone system in the body. 
--RR is associated with decreased oxygen consumption, decreased carbon dioxide elimination, decreased respiratory rate, increased brain cortical thickness, increased low-frequency heart rate oscillations, increased exhaled nitric oxide. And this study confirms potentially very important changes in gene expression
--one interesting sideline in the above transcriptomal analysis is that it provides some mechanistic rationale for the evident clinical association between hypertension and abnormal glucose tolerance, noting that those individuals who had a strong RR had significant changes in insulin/glucose metabolism, as well as inflammation, which may also be part of the pathophysiology of diabetes. several studies have suggested the systemic low-grade inflammation (which may be reflected in C-reactive protein and/or cytokine levels) may precede the development of hypertension. and several studies (including the ARIC study, Atherosclerosis Risk in Communities), have found a 2-fold increased risk of depression in diabetics
--there have been many studies linking depression with inflammation over the years (which might be a link between depression and heart disease). a recent study from the NHANES cohort found that 8% had depressive symptoms and their CRP levels were 31% higher and fractional exhaled nitric oxide (a negative marker of inflammation) was 11% lower than in those without depression (see Cepeda MS. J Clin Psychiatry 77 (12): 1666-71). and in fact other studies show that higher levels of inflammation are associated with higher rates of depression, those with depression have more neuroinflammation on their brain scans. And, higher levels of stress are associated with upregulation of the immune system involved in inflammation (which might be a link between stress and depression, see Slavich GM. Psychol Bull. 2014 May: 140 (3): 774)
--other studies have found consistent anti-inflammatory gene expression changes in those performing tai chi and qigong.
​--it should be emphasized that there was no placebo group in the above study, so the comparison between responders and nonresponders (many of whom had significant but sub-cutpoint decreases in blood pressure, such as having qualifying SBP decreases but suboptimal DBP changes) is likely to have muted the actual effect of RR
--and, of course, there needs to be a much larger study with more social and demographic diversity

so, this seems to make sense: those whose BP responded more to relaxation also had impressive decreases in their depression and anxiety. this study provided an interesting mechanism, providing the potential big picture scenario:
    ​--chronic stress leads to ???the reverse genetic transcriptomal changes found above with RR, leading to increased inflammation, which ???may then lead to the constellation of macro effects of depression, hypertension, diabetes, heart disease…
    ​--and, the Relaxation Response may lead to genetic changes found above (as well as decreased stress), associated with decreased inflammation and perhaps decreased depression, hypertension, diabetes, heart disease

so, the perhaps the real conclusion is that if we cannot decrease the stress in our lives, then there may be a role for relaxation in improving some of the most common and important chronic diseases, decreasing both morbidity and mortality. And that these changes may well be related gene-expression changes in inflammation…