VARIABILITY: THE HALLMARK OF BLOOD PRESSURE MEASUREMENT – TYPES, ASSESSMENT AND PROGNOSTIC SIGNIFICANCE

: Systemic blood pressure, recorded by various parameters, always shows variation. This is known as ‘variability’, an entity not assessed routinely. Research has shown greater prognostic importance of it than routine parameters. Various intrinsic and extrinsic factors modulate it. Depending on the interval of successive readings, blood pressure variability is of five types, with different non-invasive methods utilized for recording. It is calculated by various statistical parameters, the most common being standard deviation, but average real variability is the most accurate and easily applied. Clinical evidence is increasing rapidly, indicating variability as a prognostic marker for stroke, ischemic heart disease, renal failure, cognitive dysfunction, heart failure, and mortality. Therapeutic measures for the control of variability have also been forwarded. The paucity of clinical application of blood pressure variability is the stimulus to narrate this review, especially for physicians managing hypertensive patients.


INTRODUCTION
Blood pressure (BP), not a constant number in any individual, varies from beat to beat.Variation occurs due to several factors emanating from person, environment, atmosphere, time of day, and season.BP variability (BPV) has been noted in both normotensive and hypertensive subjects, especially in the latter.Frattola et al., for the first time, reported the prognostic importance of BPV. 1 All parameters of BP are associated with sub-clinical organ damage culminating in end-organ dysfunction, either acutely or chronically.Most prognostic studies utilized absolute and mean BP values as indicative of target organ damage (TOD), with BPV not given due importance, probably due to the non-availability of a standardized method for its assessment.Lately, many observational studies and analysis of clinical trial data (post-hoc) have shown prognostic importance. 2 BPV can be 'physiological', enabling an individual to cope with the stresses of daily life, or 'pathological', being a harbinger of diseases.No matter what time interval is taken for BP measurements, readings differ, at times, substantially.Different devices can record this variation.Initially, BPV was assessed intraarterially, not applicable in routine practice.Noninvasively, beat-to-beat variability is measured by finger plethysmographan oscillometric method.Ambulatory blood pressure monitor (ABPM) and time-triggered home blood pressure monitor are used for 24-hour BP monitoring, whereas digital, mercurial, and aneroid BP monitors are utilized for day-to-day, i.e., home and visit-to-visit BP monitoring.
Depending on the duration of successive measurements, BPV is of five types: very short-term (beat-to-beat), short-term (within a day), mid-term (within a week), long-term (visit-to-visit), and very long-term (visit-to-visit > five years). 3A difference in the prognostic impact of these BPVs has been noted. 4 a systematic review of 33 studies (over one million subjects), Stevens SL et al. examined the association of three types of BPVs (short-, mid-and long-term) with various endpoints.They concluded: "Long-term variability in SBP was associated with risk of all-cause mortality (HR 1.15, 95% CI 1.09 to 1.22), CVD mortality (HR1.18,95%CI 1.09 to 1.28), CVD events (HR 1.18, 95% CI 1.07 to 1.30), CHD (HR 1.10, 95% CI 1.04 to 1.16), and stroke (HR 1.15, 95% CI 1.04 to 1.27).Mid-term and short-term variability in daytime SBP were also associated with all-cause mortality (HR 1.15, 95% CI 1.06 to 1.26 and HR 1.10, 95% CI 1.04 to 1.16, respectively)". 5The hazard ratio for BP variability with respect to cardiovascular disease (CVD) mortality has been found to be 1.18.
In a similar meta-analysis (41 cohorts), Diaz KM et al found modest association of visit-to-visit BPV with various endpoints, "for each 5 mmHg higher SD of SBP, the pooled hazard ratios for stroke across seven cohorts was 1.17 (95% CI:1.07-1.28),for CHD across four cohorts was 1.27 (95% CI:1.07-1.51),for CVD across five cohorts was 1.12 (95% CI:0.98-1.28),for CVD mortality across five cohorts was 1.22 (95% CI:1.09-1.35),and for all-cause mortality across four cohorts was 1.20 (95% CI:1.05-1.36)". 6e prognostic impact of BPV has been noted in normotensives also.Liu M et al. studied 7065 patients with optimal systolic blood pressure (SBP) for MACE (major adverse cardiac events), which was higher by 21% in subjects with the highest SBP variability (quartile 4) as compared to those with the lowest (quartile 1) (HR 1.21; 95% CI, 1.09-1.35). 7r the assessment of BPV without any loss of prognostic information, at least 48 readings are needed. 8rious statistical parameters assess the variability of BP, the foremost being 'standard deviation (SD)', a measure of the dispersion of BP readings around a central mean value.The number of readings obtained during day and nighttime are sometimes different, and 'dipping' in nighttime BP is also noted.Many researchers have used another measure of ' weighted SD ' to remove bias created by these two factors.To obviate the total dependency of BPV on the mean value, the 'coefficient of variation (CV)', calculated by dividing SD with the mean value, was advocated by some researchers, but this also did not totally remove the effect of mean BP. 'Variation independent of mean (VIM)' and 'average real variability (ARV)' obviate the effect of mean maximally.In VIM, SD is divided by mean BP raised to the power of x, which nullifies the correlation between these two.In ARV, the sum of differences between successive BP measurements is divided by the number of total readings minus one and has been shown to provide the best estimate of BPV. 9 The therapeutic impact of BPV is yet to be established.However, certain studies have shown that Calcium channel blockers decrease, whereas beta-blockers and diuretics are inferior in this regards. 10areness of BPV amongst physicians could be better, as shown by Setia S et al. in a survey of 60 Singaporean physicians.Approximately 82% of physicians had no training for BPV. 11though the literature is loaded with evidence for BPV, basic knowledge (definition, types, acquisition methods, advantages, and drawbacks along with prognostic informationtotal or differential), has not yet been detailed for a busy clinician.To remove this paucity, it is entirely rational to narrate a text where the subject has been explained in an understandable manner.Due to the very low application of this important phenotype of BP, the main objective of this review is to apprise the treating physician of the applicability of BPV in routine clinical practice, making the management of BP in the long term more stringent.

CRITICAL OVERVIEW OF LITERATURE
The appreciable spurt of clinical evidence has been noted for BPV.A PubMed search for this term fetched only three articles in 1948.Since then, the number of studies has constantly risen, reflected by 2754 studies last year and 1537 this year till this manuscript was written.

TYPES OF BLOOD PRESSURE VARIABILITY (PHENOTYPES)
Variability in BP is time-constrained and has been divided into five classes, as shown in Figures 1 (Ultra short-term and mid-term) and 2 (mid-term and longterm) with intervals for measurements mentioned in Table 1.Devices used for measurement of different types of variability and the advantages/disadvantages of each are shown in Table 1.

Same as above
Well suited for the long term as more than 20 years of follow-up reported.
Patient migration and physician switching may not allow such long records to be maintained.

PARAMETERS TO MEASURE BLOOD PRESSURE VARIABILITY
Initially, the standard deviation (SD) of BP records was taken as the criterion of variability.It is a measure of the dispersion of BP measurements around a central mean value, Figure 3A.Many drawbacks have been noted for SD as a measure of variability.Since SD and mean BP are correlated, both cannot be used in multivariate analysis simultaneously.SD gives one average value of the whole dataset, disregarding the sequence of change in variability.
In a 24-hour study of BP records, more measurements are obtained during daytime compared to nighttime.This preponderance of daytime readings has a biased effect on the 24-hour mean value (and any derived measurement).Nocturnal dipping of BP also creates bias.To obviate these confounders, another method has been introduced in which the measurements of day and nighttime are weighted according to their numbers, and the SD derived is known as weighted SD (wSD), Figure 3B. 12other method to remove the dependency on the mean is the coefficient of variation, in which SD is divided by mean BP and expressed as a percentage, Figure 3C.Being a proportion, data recorded in different units can be compared by this parameter.
During routine chores, BP sometimes varies so abruptly that transient changes go unnoticed, which causes more target organ damage than steady state higher BP values.To catch these transient changes, another parameter came into vogue, 'time rate of BP variation (TRBP)' by Zakopoulos. 13In this, the mean of absolute ratios in the difference of successive BP readings and time between them are noted.
Two measures of assessment of BPV known as 'variation independent of the mean (VIM)' and 'average real variability (ARV)' totally disregard the mean BP, hence providing a better estimate of variability.VIM is derived by dividing SD with the mean, which has been raised to the power of x (derived from a fitting curve obtained by plotting mean SD against mean BP), rendering the correlation between the two almost zero, Figure 3D.

Figure 3: Statistical parameters used to assess BP variability
The ARV provides a better estimate of BP dispersion and therapeutic measures.It has shown greater predictive power for subclinical organ damage, mortality, and non-fatal events.ARV is the preferred index for very short-term and short-term variability assessment.Absolute differences between successive BP readings are summed up and divided by the total number of BP readings minus one, Figure 3E.
Although derived from the same data, the prognostic information obtained by these parameters differs.This has been explicitly shown by Mena LJ et al. in their remarkable meta-analysis of 19 studies comparing the predictive powers of these indexes. 14

FACTORS AFFECTING BLOOD PRESSURE VARIABILITY
Many intrinsic and extrinsic factors affect BPV, varying with type.However, most specifically affects short-term variability, as shown in Figure 4 (blue ovals).Behavioral and environmental factors affect both types of variability (white margins), and drugs or inappropriate estimation of blood pressure (yellow margins) affect long-term variability more often.Sodium causes an elevation in BP more at nighttime, especially in salt-sensitive patients, as excretion of this mineral also shows diurnal variation. 15Physiology of other minerals like potassium, magnesium, and calcium is exactly the opposite as they lower BP.
Neurohormonal factors in the form of sympathetic activation and baro-receptor reflexes play an important role in very short-term and short-term BPV.G Mancia et al. have elaborated on it in their study of 82 patients. 16rmones also have a role in the modulation of BPV.Catecholamines, Insulin, and Angiotensin II have been studied clinically.Vasodilators like bradykinin, nitric oxide, and endothelins cause an effect on BPV.These agents mostly affect very short-term and short-term BPV.
Seasonal alteration by sympathetic modulation mainly affects both short-term and long-term variability as higher blood pressures are recorded in winter, especially in the elderly. 17ch 1-degree centigrade fall in temperature results in a 1% increase in mortality due to alteration in BP.A seasonal variation in SBP of 2.9/3.4 mmHg and 1.4/0.9mmHg in the Northern/Southern hemispheres has been reported, respectively, from data across three continents and 24 studies (cross-sectional). 18working group of the European Society of Hypertension issued a consensus statement on seasonal variation in BP, mentioning the evidence and recommendations for clinical practice. 19havioral factors, especially mental stress, physical activity, and postural changes, by producing changes in sympathetic activity, volume status, and cardiac output, affect both short-and long-term BPV.
Adherence to medical treatment and restriction of social habits (smoking and alcohol consumption) ensure effective control of BPV.

BLOOD PRESSURE VARIABILITY AS A PROGNOSTIC MARKER
The The association of increased incidence of stroke due to arterial stiffness and carotid intima malformation with short-and mid-term BPV has been studied.Zhou   29 Chang JY et al. studied 90 patients who had undergone successful endovascular thrombectomy but had poor collateral circulation and found that BPV (assessed by SD, CV, and VIM of SBP and mean PR) during the first 24 hours after recanalization had a greater impact on functional outcome (assessed by early neurological recovery on day one and mRS at three months) in patients with poor collateral http://www.pakheartjournal.comcirculation. 30This necessitates adequate control of BPV for effective functional recovery.
Neurological and cardiac insults in their earlier stages demand stricter control of BPV as these patients show fluctuation in BP levels more often in the first seven days.Control at this time makes recovery smooth and better for long-term functional recovery.
 Cognitive function and dementia : Chiu TJ et al.,  in a systemic review of 20 cohort studies (7 924 168 persons), found SBP variation assessed by CV and SD was associated with a higher risk of dementia of all causes, especially during shorter follow-ups in elderly subjects (> 65 years). 31verkamp RA et al., in a study of 279 elderly patients, assessed the risk of mortality with cognitive decline and BP variability by 7-day HBPM record. 32No correlation was found for day-to-day variability irrespective of the mean (VIM) for mortality and cognitive decline, but a significant association was noted for morning systolic BP VIM and mortality (adjusted HR: 1.09, 95%-CI 1.01-1.18,p = 0.033).
Dementia and cognitive decline are associated with small cerebrovascular disease (sCVD).This entity is identified on MRI by white matter hyperintensities, microbleeds, lacunes, and perivascular spaces' enlargement.In a study of 82 middle-aged hypertensive patients, an association of BPV with sCVD was evaluated by de Heus RA et al. by HBPM. 33Significant association of systolic BPV and evening SBP was noted with sCVD.Association for DBP, mean BP, or morning BP was not significant.
In 3511 Chinese patients comprised of two groups [45-59 years (mid to old life) and 60 years or more (old to old life)], Xu et al. found a significant negative association between SBP variability and various parameters of cognitive function like orientation, language, recall, and total (assessed by Mini-mental state exam-MMSE). 34For every one-unit increase in SBP variability, the MMSE score decreased by 9.5 points.No such correlation was noted for DBP and pulse pressure variability.A significant association was stated only in the older group, emphasizing better control of SBP in this group to reduce cognitive dysfunction.
 Heart failure: Admission BP and variability in cases of heart failure have shown prognostic value with regard to mortality.A higher admission blood pressure is associated with a better prognosis, but a higher BPV in heart failure showed an adverse prognosis. 35Wei FF et al., in 1006 patients with acute decompensated heart failure (preserved ejection fraction), assessed mean admission BP, SD, and CV of BP during multiple hospitalizations for a median follow-up of 1.54 years. 36For every 1% increment in SD and CV of SBP, the risk of all-cause mortality increased by 10% and 11%, respectively (SD: HR, 1.In a large Swedish study of 9855 diabetic patients followed up for a median of 4 years, the correlation of BPV (by SD, CV, and VIM) with cardiovascular and all-cause mortality was studied with no change in hypertension treatment. 39No significant association was noted for the entire group, but in the subset of patients not taking any antihypertensive medicine (n=2949), an association was noted for all-cause mortality with only a small increase in discrimination when this variable was added to other measures.This study doesn't support the use of BPV in this group of patients as a marker of mortality.
 Chronic kidney disease: The role of BPV in assessing CKD patients is taken with care as volume and pressure status vary significantly due to interventions (dialysis, etc.).However, in certain areas, BPV plays a significant prognostic role.
Hsieh MY et al., in their study of 1011 patients on regular dialysis, found BPV was associated with an increased risk of access thrombosis [HR= 1.27, 95% CI, 1.18-1.44,/SD increase in BPV]. 40The risk in patients was 2.45 times in the highest BPV quartile.
In the HEMO trial comprising 1844 patients on regular hemodialysis, Chang TI et al. studied the association of BPV (by CV and ARV) with allcause and cardiovascular mortality. 41They found an 18% higher risk of death from any cause with every 10% rise in BPV by CV, and ARV also showed a similar association."Black race, a history of heart failure and diabetes mellitus, catheter use, and having more frequent intradialytic hypotension are associated with higher visit-to-visit blood pressure variability." Chia YC et al., in their long (15 years) retrospective study of 874 patients, studied the relation of BPV with a decline in eGFR. 42A significant negative correlation was noted (SD: r=0.16, p<0.001;CV: r=0.14, p<0.001).They found that an SD of 13.5 mmHg and CV of 9.74% were associated with the onset of chronic kidney disease.
Wang G et al., in their study of 245 non-dialyzed patients of renal failure (grade 1-4), found a correlation of high BPV (assessed by SD, CV, and VIM) with the progression of renal disease (seen only in unadjusted model and not in fully adjusted model) and cardiovascular disease but not with total mortality. 43They found hyperkalemia, LV end-diastolic diameter, hypertension, and BMI as markers of high BPV.
Kumanan T et al., in their one-year study of longterm blood pressure variability of 406 patients, found that female sex (p=0.023),DM (p=0.013),CKD (p=0.007), and the tendency for developing OSA (p=0.004) were associated with significant variability. 9A median value of 11.69 differentiated low and high variability groups.In multivariate analysis, only DM and CKD were significantly associated with high variability.
 Obstructive sleep apnea: Obstructive sleep apnea (OSA) is associated with hypertension.
Steinhorst has studied the association of BPV and OSA, AP et al. in 107 patients. 44A significantly higher BPV has been noted in patients with apneahypopnea index ≥ 10.
Damage to targets of high blood pressure is accrued by every component of it (SBP, DBP, and mean BP), but the narration mentioned above shows clinical evidence of a higher damaging role of blood pressure variability, emphasizing its recognition and effective control.

TARGETING BP VARIABILITY THERAPEUTICALLY
Although the position of BPV as a prognostic marker has been clearly established, there are clinical lacunae in its implementation as a therapeutic target.First, there is no clear demarcation for normality at present; secondly, pharmacological trials especially directed against BPV, need to be improved.However, evidence is accumulating from earlier large-scale trials (ALLHAT trial) that calcium channel blockers reduce BPV significantly more than ACE inhibitors (lisinopril).In a retrospective analysis of MRC trial and ASCOT-BPLA trials, beta-blockers proved inferior to diuretics (Chlorthalidone) and calcium channel blockers (Amlodipine). 45Combinations of antihypertensive agents, including calcium channel blockers, are better at reducing BPV than combinations lacking this agent (COPE trial). 46The superiority of calcium channel blockers for modulation of BPV extends to all types of it and combination with ACEI/ARB, outperforming other combinations, especially those involving beta blockers.However, more dedicated research is still needed in this regard.

DIFFERENTIAL PROGNOSIS OF BLOOD PRESSURE VARIABILITY PHENOTYPES
The prognostic impact of the different phenotypes of BPV differs, as demonstrated by comparative clinical studies.
Tahir ZA studied 152 hypertensive patients retrospectively, comparing short-term (ABPM record) and long-term (home BP record) BPVs for a period of 10 years.The endpoints of the study were "acute coronary syndrome (ACS), chronic ischemic heart disease (IHD), heart failure (HF), or stroke".Systolic BPV of day (OR=1.94;95% CI=1.09-3.45;p=0.025) and nighttime (OR=1.23;95% CI=1.00-1.51;p=0.048) showed significant association with IHD, whereas SD of visit-to-visit BPV was significantly associated with ACS (OR=1.10;95% CI=1.01-1.21;p=0.04). 47n another comparative study (n = 508) of long-term (visit-to-visit BPV) vs. short-term (24-hour BPV) for mortality in treated hypertensive patients (age > 65 years), Chowdhury EK et al. also showed differential results. 48SD of daytime SBP and SDw of 24-hour SBP were more significantly associated with mortality, whereas SD of visit-to-visit SBP was not.The low association of VV BPV with endpoint was attributed to a small number of patients by the researchers, as other studies have shown an association in this regard.

CONCLUSION
Blood pressure variability, a parameter of great prognostic importance, is yet to gain widespread clinical acceptance in daily practice.Long-term variability can be gauged easily by serial BP measurements obtained from patients' medical records.The application needs the knowledge of the treating physician in this regard, which is the primary purpose of this manuscript.Short-term variability assessment needs some sophisticated measures (home BP measurement or ABPM), which are now becoming more readily available with advancements in medical care.The medical community must refrain from showing inertia in using these gadgets and apply all the available parameters to ensure better health care.Reasonable control of BPV in acute phases of vascular events ensures better survival and limited damage in the long term.
Similarly, encouraging results can be obtained if this parameter is closely observed in cognitive and renal dysfunction areas.No specific therapeutic agent is available at present for BPV normalization, but there are clues from large-scale studies showing the efficacy of certain agents (calcium channel blockers).This area needs special attention for better clinical outcomes in the future.BPV phenotypes differ regarding prognostic capability in different clinical conditions, and variability of various blood pressure parameters also show prognostic differences for clinical endpoints.A thorough knowledge of these subtleties of blood pressure management will ensure high-quality care for hypertensive patients

Figure 1 :Figure 2 :
Figure 1: A-Ultra short-term BP variability.The red line is for systolic BP, and the blue line is for diastolic BP,

Figure 4 :
Figure 4: Factors affecting BP variability Eating causes a slight rise in BP, which usually falls 1 hour later due to vasodilation in viscera and is seen especially after carbohydrate intake in elderly and hypertensive subjects.
prognostic significance of BPV has been shown in a large number of studies.This has varied from individual target organs to cumulative evidence of damage, as demonstrated by all phenotypes.Intraphenotypic comparison has also been done for various endpoints, showing mixed results.BPV has demonstrated clear superiority over other BP parameters like SBP, DBP, and mean BP.  Target organ damage: Clinical evidence for target organ damage (TOD) has been noted in earlier studies.Gianfranco H et al. studied 108 hypertensive patients by intra-arterial 24-hour BP measurement. 20Parameters for LVH (ECG), cardiac volume (CXR), ocular fundi (Keith-Wegner classification), heart failure, cerebral and peripheral vascular insufficiency (clinical measures), and renal dysfunction (BUN, creatinine, and proteinuria) were scored for TOD.Short-term and long-term variabilities were assessed within half-hour SD and among halfhour SD of mean BP.High BPV, both short-and long-term (p < 0.05 and < 0.01), respectively, were found to have significant prognostic values for TOD.

Table 1 : Variability types, devices used for recording, advantages, and disadvantages S. No. Type of variability Time interval and duration Device used Advantages Disadvantages
White coat effect.Provision of validated devices may not be possible.Personnel expertise during different time frames may include bias.