home-based blood pressure variability seems best for cardiac risk

 A recent study found that blood pressure variability itself was associated with cardiovascular disease, independent of the measured blood pressure, and that assessing home-based blood pressure variability was superior to ambulatory blood pressure variability (see htn home BP better than ABPM HTN2023 in dropbox, or DOI: 10.1161/HYPERTENSIONAHA.123.2089 )  

   

Details:  

-- 1314 hypertensive patients  from 71 institutions in Japan having office BP> 140/90 mmHg underwent both home and ambulatory blood pressure measurements in the J-HOP (Japan Morning-Surge Home Blood Pressure) study in this post hoc analysis, between 2005-2012  

-- mean age 65, 47% male, BMI 24.6  

-- current smoking 12%, habitual drinking 22%  

-- 86% on antihypertensive meds: ARBs 57%, calcium channel blockers 55%, diuretics 31%, b-blockers 16%, ACE 8%, a-blockers 7%, MRAs 4%  

-- comorbidities: diabetes 24%, CKD 27%, preexisting CVD (cardiovasc disease) 17%, dyslipidemia 34%  

-- urine albumin/creatinine ratio (UACR) 80 mg/g Cr, BNP 36 pg/mL  

-- Baseline office BP: 141/81 mmHg  

   

-- blood pressure measurements:  

    -- office-base blood pressure: 3 BP readings were taken at 15-s intervals in a sitting position. Recorded BP was the mean values of 6 readings obtained during 2 clinic visits  

    -- home-based blood pressure: recorded on a validated blood pressure cuff  

-- number of BP measurements:  

    -- home-based blood pressure: 33 readings (17 morning, 16 evening)  

    -- ambulatory blood pressure: 45 readings (15 included nighttime)  

   

-- Ambulatory BP:  

    -- 24-hr ambulatory systolic BP (ASBP): 131.1, +/- 12.9  mmHg  

    -- 24-hr ambulatory diastolic BP (ADBP): 77.5, +/- 9.3 mmHg  

    -- daytime ASBP: 135.9, +/- 13.6  mmHg  

    -- daytime ADBP: 80.4, +/- 9.9 mmHg  

    -- nighttime ASBP: 120.5, +/- 15.0  mmHg  

    -- nighttime ADBP: 70.6, +/- 9.4 mmHg  

-- Home BP:  

    -- Mean of average morning and evening BP (MEAve), home systolic BP (HSBP): 136.5 +/- 14.5 mmHg  

    -- MEAve, home diastolic BP (HDBP): 76.5 +/- 9.8 mmHg  

    -- Morning HSBP: 140.8 +/- 16.1 mmHg  

    -- Morning HDBP: 79.7 +/- 10.6 mmHg  

    -- Evening HSBP: 131.7 +/- 15.5 mmHg  

    -- Evening HDBP: 73.1+/- 10.1 mmHg  

   

-- Multivariate assessment of the associations of  ambulatory blood pressure variability (ABPV) and home blood pressure variability (HBPV) with cardiovascular risk, using these three statistical models:  

    -- Model 1:  controlling for age; sex; BMI; smoking; alcohol consumption; diabetes; dyslipidemia; antihypertensive medication use; and pre-existing coronary artery disease, stroke, or heart failure  

    -- Model 2: adds office systolic BP (SBP) and the mean 24-h ASBP or MEAve HSBP level to Model 1 of ABPV or HBPV, respectively  

    -- Model 3: adds the mean MEAve HSBP and each MEAve HBPV parameter or the mean 24-h ASBP and each 24-h ABPV parameter to Model 2 of ABPV or HBPV, respectively 

        -- parameters to assess BP variability: the statistical contribution of standard deviation (SD), coefficient of variation (CV), and average real variability (ARV) of the patients’ 24 hour-ambulatory and home systolic blood pressures  

   

-- Cardiovascular outcomes measures: fatal and nonfatal coronary artery disease (acute MI, angina requiring percutaneous coronary intervention, or sudden death within 24 hours of an abrupt symptom onset; fatal and nonfatal stroke; heart failure defined as hospitalization due to acute decompensated heart failure)  

-- They also assessed Brain Natriuretic Peptide (BNP) and urinary albumin-to-creatinine ratio 

   

-- median follow-up 7.0 years  

   

Results;  

--109 cardiovascular events occurred: 49 coronary artery disease events, 45 strokes, and 15 heart failure events  

-- there were modest but highly statistically significant (p<0.001) associations between the ABPM and HBPM measurements by the statistical contributions of ST, CV, and ARV  

   

-- ABPV 24-hour assessments (SD, CV, or ARV): none were significantly associated with cardiovascular risk in any of the adjustment models noted above [none of them were even close to being significant]  

-- HBPV (MEAve): all were statistically significant in each of the adjustment models  

    -- model one:  

        -- SD: 44% increase in cardiovascular events for each 1 SD, aHR 1.44 (1.22-1.70), p<0.001  

        -- CV: 42% increase in cardiovascular events for each 1 SD, aHR 1.42 (1.19-1.69), p<0.001  

        -- ARV: 37% increase in cardiovascular events for each 1 SD, aHR 1.37 (1.18-1.59), p<0.001  

    -- model two:  

        -- SD: 35% increase in cardiovascular events for each 1 SD, aHR 1.35 (1.13-1.61), p=0.001  

        -- CV: 37% increase in cardiovascular events for each 1 SD, aHR 1.37 (1.15-1.64), p=0.001  

        -- ARV: 29% increase in cardiovascular events for each 1 SD, aHR 1.29 (1.10-1.51), p=0.001  

    -- model three model two plus 24 hour ASBP:  

        -- SD: 36% increase in cardiovascular events for each 1 SD, aHR 1.36 (1.14-1.63), p=0.001  

        -- CV: 38% increase in cardiovascular events for each 1 SD, aHR 1.38 (1.16-1.66), p<0.001  

        -- ARV: 29% increase in cardiovascular events for each 1 SD, aHR 1.29 (1.10-1.51), p=0.002  

   

-- biomarkers assessed:  

    -- BNP: more highly correlated with HBPV than ABPV; in the adjusted models, no association with ABPV  

    -- UACR: more highly correlated with HBPV than ABPV; in the adjusted model, no association with either HBPV or ABPV  

   

-- subgroup analyses:  

    -- the ARV of 24-hour ABPV was statistically associated with those people with a prior history of cardiovascular disease, as well as those with well-controlled 24-hour systolic blood pressure  

    -- there was also was a significant association of the ARV of the 24hour ABPV with CVD risk in those not using ARBs/ACE inhibitors  

    -- none of the assessments for HBPV (MEAve) were associated with CVD risk in those >75yo or with the prior history of cardiovascular disease, though these did in the adjusted models  

   

Commentary:  

-- though many prior studies have found a significant and independent role of blood pressure variability as a cardiovascular risk factor, the literature on blood pressure variability is a tad complicated, with some articles defining the variability as BP changes during a short time interval,  or 24-hour blood pressure variation, or day-to-day blood pressure variation over several days, or visit-to-visit clinic blood pressure variation assessed  over several outpatient clinic visits.  the finding of increased cardiovascular events has been found with each of these differing definitions  

    -- as an example, blood variability was found to be associated with increased cardiovascular events, independent of the blood pressure reading itself:
https://gmodestmedblogs.blogspot.com/2016/09/blood-pressure-variability-increases.html  

-- other studies have suggested that both ambulatory and home-based blood pressure variations were significantly associated with cardiovascular outcomes 

  

-- prior studies have documented the benefit of home-based blood pressure readings themselves as a predictor of future cardiovascular events with some suggesting they were almost as good as ambulatory blood pressure readings (in a 2011 systematic review and recommendations, the National Institute for Health an Clinical Excellence, NICE, in the UK strongly supported ambulatory blood pressure monitoring and home-based monitoring for diagnosing and treating hypertension: see Htn nice recs 2011 in dropbox, or their Clinical Guideline 127 from August 2011). 

-- there are, however, fundamental differences in what the ambulatory vs home blood pressure measurements are measuring:  

    --ambulatory blood pressure collects 24 hours of blood pressure recordings (at least 27 are needed for the evaluation, per the editorialists McEvoy JW, Leahy N, and Parati, G), and averages quiet time, daily activities, daily stressors, work, and sleep (which may be associated with lower or higher blood pressures)  

        – and the ABPM results reflect ony one 24-hour period 

    --home blood pressures are measured preferably in quiet time/resting, and preferably the first 1 or 2 are discarded to try to ensure that the readings are the lowest possible  

        – and the HBPM results reflect blood pressures over a longer time interval 

-- this all being said, and bearing in mind that blood pressure and blood pressure variability are independently related to cardiovascular outcomes, it is important to emphasize that ambulatory blood pressure monitoring (ABPM) is the gold standard for hypertension diagnosis: https://gmodestmedblogs.blogspot.com/2015/10/uspstf-guidelines-on-blood-pressure.html  

-- a recent study found that adolescents with hypertension seem to have more adverse cardiovascular events decades later (https://gmodestmedblogs.blogspot.com/2023/12/hypertension-in-teens-predicts-future.html) and that hypertension duration seems  also  to be an independent cardiovasc risk factor. All of this reinforces that atherosclerotic cardiovascular disease is a cumulative process that begins in teens or even before in those with severe familial hyperlipidemia, that it is important to identify and treat hypertension (and lipids, etc) starting in young people (preferably with nonpharmacologic therapies), and that the focus should be on the level of hypertension as well as both hypertension variability and duration.  

  

-- One issue with performing home-based blood pressure readings is the lack of standardization:  

    -- some patients may use inaccurate blood pressure cuffs  

    -- there is no standardized approach overall for instructing patients on how to check their blood pressures  

    -- patients themselves may well not be assessing blood pressure in a consistent manner over time (eg sometimes in the morning after coffee, perhaps in a hectic afternoon, another time when they are winding down at night, …) 

-- my sense from relying on many home-based blood pressure readings (their being more frequently used during Covid) is the following (though this has not been formally tested):  

    -- the patient should be sitting in a quiet room with their feet on the floor and without distractions for two minutes, preferably with their eyes closed and encouraged to think about pleasant things (but not looking at their phone or TV)  

    -- it is likely best if the patient discards their first reading (which might be associated with a catecholamine surge as the cuff inflates), and records the second reading  

    -- the blood pressure should not be measured before 30 minutes after drinking coffee or smoking  

    – given diurnal variations in blood pressure and the fact that some meds have less than 24-hour control, i do ask the patients to check blood pressures  at different times of the day 

-- I do ask the patient to bring in their blood pressure cuff to the office, have them measure the blood pressure on one arm (to assess if they correctly measure their blood pressures) while I simultaneously check the blood pressure in the other arm (to make sure that their cuff is accurate); and given the unusual but possible differences in blood pressures between the 2 arms, I reverse this procedure as well. If the patient measures their blood pressure accurately and with an accurate BP cuff, I feel comfortable using their home measurements (which are more predictive of adverse cardiovasc outcomes than clinic-based reading: eg see the NICE guidelines noted above) to make decisions on both nonpharmacologic and pharmacologic therapies 

   

-- there are some medications that decrease blood pressure variability. the class of meds given as monotherapy that seems to be the best are calcium channel blockers (especially amlodipine) and diuretics (especially chlortalidone), though b-blockers, ACE-Is, and ARBs did much less well. In patients on 2- or 3-drug combinations, those including calcium channel blockers did the best (see htn blood pressure variability by meds JHypertension2021 in dropbox, or DOI:10.1097/HJH.0000000000002618;)  

    – it is important to remember that there may also be significant outcome differences between antihypertensives not related to either blood pressure achieved or blood pressure variability. For example, lisinopril (probably the best of the ACE-I's in terms of duration of effect) may well be inferior to ARBs for stroke prevention (see htn ARBs better than ACE for stroke prevention in dropbox, or doi:10.1093/eurheartj/sup026), though the 24-hour BP variability seems to be about the same (see htn variability and stroke lancet 2010 in dropbox, or Lancet 2010; 375: 906–15)  

   

-- This study has the advantage of measuring ABPV as well as HBPV in the same patients, and supports the following:  

    -- increased HBPV is associated with increased cardiovascular risk, independent of the average home pressure readings  

    -- ABPV is not associated with cardiovascular disease in the adjusted models  

    -- HBPV remained significantly associated with risk after adjusting for  ABPV  

    

-- measuring the BNP is useful because it is associated with activation of the sympathetic nerves and arterial stiffness as well as systolic ventricular volume overload and stress, and therefore a biomarker associated with hypertension (and left ventricular health), and presumably associated with blood pressure variability; UACR is a useful marker of high cardiovascular risk, and is related to renal damage and endothelial dysfunction: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3178015/ . In this study, only the BNP was associated with HBPV and not the ABPV in their adjusted model 

   

Limitations:  

-- there were several factors in this study that might limit its generalizability:  

    -- this was a post hoc analysis of an ongoing study and not a primary goal of the study (this study was primarily aimed to assess the usefulness of HBPM)  

    -- patients who had ambulatory blood pressure monitoring elected to do so as an optional test, and they therefore may be different from the general population  

    -- the study occurred in Japan and may involve demographic and cultural issues specific to that country, including differences from other areas in diet, exercise, homogeneity of the population, stressors, etc  

    -- the study also had an older cohort of participants, and the result may not be applicable to younger folks. Also there were comorbidities in more than a quarter of the individuals. all might limit generalizability to other groups 

    -- this study also was of hypertensive patients with pretty well-controlled hypertension (office BP 141/81, ambulatory BP 131/76 mmHg), and their results may not apply to either normotensive patients or those with higher blood pressures  

    -- there were relatively few blood pressures measured ( home-based blood pressure: 33; ambulatory blood pressure: 45 readings). the editorialists comment that a minimum of 72 readings/d are needed for ambulatory BP monitoring to assess BP variability 

-- ABPM was measured only once over 24-hour period, with might have missed variations in ambulatory blood pressure that might have occurred over time  

-- there was no information of the patients' blood pressures, either ambulatory or home-based, over the seven-year time course of the study. And it may be fraught to conclude that the initial blood pressure evaluation would reflect the blood pressure or its variability over a long period of time (lots of things might change including patient weight, diet, exercise, meds taken, medical comorbidities, etc)  

-- there was very limited information in the study of many of the known determinants of hypertension (diet, exercise, stress, isolation, amounts of alcohol or tobacco exposure, etc) 

-- and, this was an observational study, limiting our ability to determine causality; of note, despite the pretty large number of studies on BP variability, there have not been any randomized controlled trials that assessed similar blood pressure control from different meds having different effects on BP variability, and evaluating cardiovascular and other outcomes accurately 

   

So, another article finding that blood pressure variability is associated with increased adverse cardiovascular events, independent of the baseline blood pressure itself. In this article, blood pressure variability measured at home had the highest association with future cardiovascular events, suggesting a few things:  

-- there seems to be a reasonable imperative for us to develop a clear and systematic approach to doing these home-based measurements, one that is validated and a reasonable surrogate to ABPM as a predictor of future adverse cardiovascular events 

-- employing home-based measurements allows for more telehealth visits, with the attendant convenience to patients and clinicians  

    -- and, home-based BP measurements are more predictive of future cardiovasc events than clinic-based measurements, per the NICE guidelines noted above 

  

-- there are substantial limitations to the design and evaluations in this study, as noted above.  

    -- but, it seems to me that we as clinicians should consider blood pressure variability at home as an important factor in assessing hypertension control and cardiovascular risk, and also to pay attention to the patient’s visit-to-visit variations in the clinic. Blood pressure variability has been shown in several studies to be an independent marker of cardiovascular disease, and we should  incorporate this into our atherosclerosis risk stratification: https://gmodestmedblogs.blogspot.com/2023/10/update-ascvd-risk-factor-critique.html . 

    -- so, it seems that there is little harm, for example, in preferentially initiating therapy with amlodipine, since it works well with hypertension control, is pretty well-tolerated, and seems the best drug to minimize BP variability, pending well-designed RCTs  

-- home-based measurements may well empower patients in terms of their self-care, both by incorporating them directly into their hypertension therapy (pharmacologic and nonpharmacologic), as well as deepening their own understanding of how their environment can affect their blood pressure (e.g. salt, diet, exercise, stress) and the importance of trying to moderate these risk factors  

  

geoff 

 

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list