Journal Logo


Assessment of drug effects on blood pressure variability

which method and which index?

Stergiou, George S.; Kollias, Anastasios; Ntineri, Angeliki

Author Information
doi: 10.1097/HJH.0000000000000201
  • Free

Blood pressure (BP) is known to be a continuous variable with dynamic characteristics of variability in response to daily physical and mental stimuli. The idea that the variation of BP puts additional burden on the heart and vasculature beyond that of average BP has been regarded for decades as a reasonable concept by researchers, practitioners and even patients. However, in contrast to the straightforward approach required for the evaluation of average BP, the quantification of the BP variability (BPV) turned out to be a tricky task. Interestingly, several different lines of evidence suggest that the BPV has independent prognostic value beyond that of average BP. However, to date the plethora of methodological approaches applied in different studies and the lack of evidence on critical relevant research issues did not allow BPV to be used in practice as a tool for improving the management of hypertensive patients.

For the quantification of BPV, several measurement methods and sampling of BP readings have been used, and several mathematical approaches have been applied (Tables 1–2) [1–10].

Indices of blood pressure variability
Indices of blood pressure variability per measurement method


The different methods available for office and out-of-office BP measurement have been used to provide information on different aspects of BPV. Evaluation of BPV using each of these methods has been shown to independently predict cardiovascular risk [1–3,11–17]. However, each BPV component seems to reflect different mechanisms, is likely to provide different information on cardiovascular regulation and might have different clinical implications [12,18].

Very-short-term BPV can be assessed by continuous beat-to-beat intraarterial BP monitoring or noninvasive finger-cuff photoplethysmography. The use of the former is limited because of the invasive nature and the latter because of questionable measurement accuracy.

Short-term BPV is based on intermittent BP sampling at 15–30-min intervals in a routine 24-h period, obtained using noninvasive oscillometric ambulatory monitors. This appears to be an ideal method for routine BPV evaluation, by providing information on the dispersion of BP values in different conditions of posture and activity. However, the independent prognostic value of ambulatory BPV has been questioned, and its application might not be well accepted by patients for repeated use in the long-term management of hypertension. In the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT), ambulatory BPV had less effect on vascular events than that assessed by office measurements [2]. Moreover, analysis of the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcome (IDACO) showed that the 24-h ambulatory BPV did not contribute much to risk stratification over and beyond the average ambulatory BP [15]. It should be mentioned, however, that most of the published studies have been limited by infrequent ambulatory BP sampling, whereas it has been shown that measurements at 15-min intervals are required to provide an accurate assessment of BPV [19].

Mid-term day-by-day BPV based on self-home BP monitoring has also been shown to provide prognostic information independent of average home BP [1,12,13,17]. Home monitoring might be more appropriate for repeated BPV assessment in clinical practice, because it is widely available and well accepted by hypertensive patients for long-term use [12,13]. However, its application requires validated devices, measurement standardization and prevention of patients’ reporting bias using automated memory [12,13].

Long-term BPV can be assessed by repeated office BP measurement in succeeding visits. Recent outcome studies demonstrated the prognostic relevance of visit-to-visit BPV [2,3,16], which might be superior to short-term ambulatory BPV [2]. In these studies, however, carefully standardized office measurements were obtained, which might not be feasible to replicate in routine clinical practice.


Multiple mathematical approaches have been applied to determine BPV evaluated by different BP measurement methods. The SD is the classic index for BPV quantification, yet its major limitations are that it is proportional to the mean BP value and might have inferior prognostic ability than newer BPV indices. Several SD-based formulas have been introduced to eliminate the impact of mean BP (e.g. coefficient of variation), or of diurnal BP variation [4]. More recent studies tested novel BPV indices, which appear to have superior prognostic ability to SD-based indices. Some indices have been developed to handle BP readings obtained using specific measurement methods (Table 2).

While SD merely reflects the BP excursions around the mean, the time rate of BP variation measures the speed of ambulatory BP fluctuations between successive readings and integrates the direction of these changes [5,20]. There is evidence that the time rate of BP variation is independently associated with target organ damage [5,20].

The average real variability (ARV) also overcame deficiencies of SD and added unique additional prognostic value to short-term ambulatory BPV, being more sensitive to the individual BP measurement order and less so to the low sampling frequency of ambulatory monitoring [6,15,21].

The variability independent of mean (VIM), introduced in the analysis of prognostic ability of visit-to-visit BPV in the ASCOT and the Medical Research Council (MRC) Trial, transformed SD into a new statistical tool uncorrelated with mean BP and with independent prognostic value [2]. Asayama et al.[1] explored the prognostic ability of ARV, VIM and the maximum–minimum BP difference of home measurements, and concluded that none of them incrementally predicts cardiovascular outcome over and beyond mean SBP. Finally, the residual BPV (RSD) reflects the erratic component of short-term BPV, and has been shown to be positively associated with left ventricular mass index [7] and cardiovascular risk [14].

Interindividual versus intraindividual BPV has been an issue of debate in the analysis of outcome trials. In the ASCOT, interindividual BPV was lower with amlodipine than with atenolol, mainly because of lower intraindividual BPV, and in the MRC study, both interindividual and intraindividual BPV increased with atenolol compared with placebo or diuretic [2]. Although there was a relatively good correlation between intravariability and intervariability, they cannot be regarded as interchangeable indices but probably reflect different physiological or therapeutic phenomena [22,23]. Interindividual BPV during treatment may primarily depend on the individual's BP response to treatment rather than the variance of BP response over time [22,23]. This view is supported by the European Lacidipine Study on Atherosclerosis (ELSA) that showed higher interindividual than intraindividual visit-to-visit office and ambulatory BPV, suggesting that only the latter is able to reflect precisely the treatment-induced changes [22].

Specific indices have been developed to quantify the effect of antihypertensive drugs on ambulatory BP. The trough-to-peak ratio (TPR), introduced to evaluate the duration of antihypertensive drug action, reflects the ambulatory BP decline in two narrow time intervals, is poorly reproducible and has wide interpatient variability [24]. In contrast, the smoothness index is based on the entire 24-h recording period and takes into account both the degree of BP reduction and its 24-h distribution [8,24]. Thus, it reflects the homogeneity of 24-h drug effect, and is more reproducible and more closely related to treatment-induced regression of organ damage than the TPR [8,25].

In this issue of the journal, Parati et al.[9] introduced the treatment-on-variability index (TOVI), calculated by dividing the treatment-induced hourly BP reduction to the degree of absolute BPV under the same treatment. Thus, the TOVI allows the assessment of the impact of antihypertensive drug treatment on both mean BP and BPV over 24 h. TOVI was compared with smoothness index in a retrospective analysis of 10 drug trials receiving placebo, monotherapy or two-drug combination, with ambulatory BP available before and during treatment [9]. Although both smoothness index and TOVI reflect the effectiveness of drug-induced hourly BP decline, the former assesses the 24-h variation of this decline, whereas the latter quantifies the short-term BPV ridded of the impact of the nocturnal BP change [9]. The two indices showed similar behavior among treatments by discriminating their differential impact on BP and BPV. Although both indices were improved with all treatments compared with placebo, combination therapy resulted in higher values than monotherapies [9]. Higher smoothness index or TOVI was attributed to stronger BP-lowering effect and longer duration of drug action [9]. Thus, TOVI appears to be at least as useful as smoothness index, yet the clinical relevance of treatment-induced effects on these indices in terms of clinical outcomes remains to be proved.


In recent years, there has been increasing interest in the effect of antihypertensive drugs on BPV and its independent impact on cardiovascular event prevention. Several outcome trials suggested that calcium channel blockers (CCBs) are superior to other drug classes in reducing BPV, which might independently contribute to more effective cardiovascular protection.

In the ASCOT, clinic and ambulatory BPV were higher in the CCB compared with the β-blocker-treated group, independently of their effects on mean BP [2]. Interestingly, the cardiovascular event rates were lower in the CCB group, which could be partially attributed to the drug effects on BPV. In the MRC trial, visit-to-visit BPV was increased in the β-blocker compared with the diuretic and the placebo group, with these BPV trends in the β-blocker group being associated with stroke risk [2].

The above outcome data are in line with a meta-analysis of 389 trials that showed BPV to be reduced by CCBs and diuretics, and increased by β-blockers, angiotensin converting enzyme inhibitors and angiotensin receptor blockers [26]. In 21 trials with outcome data, the above effects contributed to differences in stroke risk, independently of effects on mean BP [26]. Interestingly, the aforementioned opposing effects of antihypertensive drug classes on BPV were dose–dependent and persisted when used in combinations [27]. The authors suggested that high-dose CCB monotherapy or combination with other drugs might be particularly effective for stroke prevention [27]. A recent systematic review of home BPV trials also suggested favorable effects of CCBs and not of β-blockers or angiotensin receptor blockers on BPV [13].

The abovementioned trials had different design and different methodology for BPV evaluation, yet the message is consistent in favor of CCBs. Thus, an important new chapter has opened, and there is an urgent need to establish the optimal methodology and indices for quantifying the drug treatment effects on BPV.


A mathematical approach to select the optimal BPV index that accurately reflects the dispersion of BP values around the average unaffected by the latter is reasonable and was the rationale for introducing novel indices, such as the ARV and VIM. On the other hand, indices specifically developed to assess drug-induced BPV changes (e.g. smoothness index and TOVI) are clearly attractive. However, the ultimate test for a BPV index is to show whether its change with treatment can independently influence the risk of cardiovascular events.

The outcome trials, that showed CCBs to have favorable effects on BPV compared with other drugs with subsequent enhanced cardiovascular protective abilities, have used various methods for BPV quantification. In the ASCOT and the MRC trial, Rothwell et al.[2] tested the SD, coefficient of variation, VIM, ARV and RSD of clinic BP measurements in terms of intraindividual visit-to-visit BPV, and also SD, coefficient of variation and ARV of ambulatory BP, and concluded that mainly the effects on systolic visit-to-visit BPV and partly on systolic ambulatory BPV could account for reduced event rates in the CCB group. A meta-analysis of 21 outcome trials applied the variance ratio as an expression of interindividual visit-to-visit variability to differentiate the BPV changes with different drugs [26].

Cross-sectional and short-term studies also demonstrated the ability of several BPV indices to differentiate the effect of different drugs. Some studies obtained home BP readings and used the SD or coefficient of variation of morning BP [28], or SD of daily mean or daily morning or evening BP [13,29]. Other studies performed ambulatory BP monitoring and used the SD of systolic daytime, night-time and 24-h BP [30]. Ambulatory BP-specific indices, such as TPR, smoothness index and TOVI, have been explored in a meta-analysis of 11 trials [31], in the analysis of 10 trials in the current issue of the journal [9] and in other trials [32] with successful results. The prognostic relevance of the latter indices requires verification in outcome trials.

The optimal method and index for assessing BPV should combine technical and clinical features. Evidence should be provided that the BPV index: is easily measurable to be applicable in clinical practice; is reproducible; has defined normalcy and intervention thresholds; independently contributes to cardiovascular risk; is modifiable by treatment; and patients’ prognosis is improved when additional treatment targets are set for BPV beyond those for average BP. At present, evidence for most of these questions is missing, and therefore BPV remains a challenging research issue deserving thorough investigation. Future prospective trials should perform head-to-head comparisons of different BPV indices and test BPV as an additional target of treatment, aiming to more efficient prevention of organ damage and cardiovascular disease.


Conflicts of interest

There are no conflicts of interest.


1. Asayama K, Kikuya M, Schutte R, Thijs L, Hosaka M, Satoh M, et al. Home blood pressure variability as cardiovascular risk factor in the population of Ohasama. Hypertension 2013; 61:61–69.
2. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B, et al. ASCOT-BPLA and MRC Trial Investigators Effects of beta blockers and calcium-channel blockers on within-individual variability in blood pressure and risk of stroke. Lancet Neurol 2010; 9:469–480.
3. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375:895–905.
4. Bilo G, Giglio A, Styczkiewicz K, Caldara G, Maronati A, Kawecka-Jaszcz K, et al. A new method for assessing 24-h blood pressure variability after excluding the contribution of nocturnal blood pressure fall. J Hypertens 2007; 25:2058–2066.
5. Zakopoulos NA, Tsivgoulis G, Barlas G, Papamichael C, Spengos K, Manios E, et al. Time rate of blood pressure variation is associated with increased common carotid artery intima-media thickness. Hypertension 2005; 45:505–512.
6. Mena L, Pintos S, Queipo NV, Aizpúrua JA, Maestre G, Sulbarán T. A reliable index for the prognostic significance of blood pressure variability. J Hypertens 2005; 23:505–511.
7. Sega R, Corrao G, Bombelli M, Beltrame L, Facchetti R, Grassi G, et al. Blood pressure variability and organ damage in a general population: results from the PAMELA study. Hypertension 2002; 39:710–714.
8. Parati G, Omboni S, Rizzoni D, Agabiti-Rosei E, Mancia G. The smoothness index: a new, reproducible and clinically relevant measure of the homogeneity of the blood pressure reduction with treatment for hypertension. J Hypertens 1998; 16:1685–1691.
9. Parati G, Dolan E, Ley L, Schumacher H. Impact of antihypertensive combination and monotreatments on blood pressure variability: assessment by old and new indices. Data from a large ambulatory blood pressure monitoring database. J Hypertens 2014; 32:1326–1333.
10. Stauss HM. Identification of blood pressure control mechanisms by power spectral analysis. Clin Exp Pharmacol Physiol 2007; 34:362–368.
11. Parati G, Ochoa JE, Lombardi C, Bilo G. Assessment and management of blood-pressure variability. Nat Rev Cardiol 2013; 10:143–155.
12. Stergiou GS, Nasothimiou EG. Home monitoring is the optimal method for assessing blood pressure variability. Hypertens Res 2011; 34:1246–1248.
13. Stergiou GS, Ntineri A, Kollias A, Ohkubo T, Imai Y, Parati G. Blood pressure variability assessed by home measurements: a systematic review. Hypertens Res 2014; [Epub ahead of print].
14. Mancia G, Bombelli M, Facchetti R, Madotto F, Corrao G, Trevano FQ, et al. Long-term prognostic value of blood pressure variability in the general population: results of the Pressioni Arteriose Monitorate e Loro Associazioni Study. Hypertension 2007; 49:1265–1270.
15. Hansen TW, Thijs L, Li Y, Boggia J, Kikuya M, Björklund-Bodegård K, et al. International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcomes Investigators Prognostic value of reading-to-reading blood pressure variability over 24 h in 8938 subjects from 11 populations. Hypertension 2010; 55:1049–1057.
16. Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The relationship between visit-to-visit variability in systolic blood pressure and all-cause mortality in the general population: findings from III NHANES, 1988 to 1994. Hypertension 2011; 57:160–166.
17. Johansson JK, Niiranen TJ, Puukka PJ, Jula AM. Prognostic value of the variability in home-measured blood pressure and heart rate: the Finn-Home Study. Hypertension 2012; 59:212–218.
18. Stergiou GS, Parati G. How to best assess blood pressure? The ongoing debate on the clinical value of blood pressure average and variability. Hypertension 2011; 57:1041–1042.
19. di Rienzo M, Grassi G, Pedotti A, Mancia G. Continuous vs intermittent blood pressure measurements in estimating 24-h average blood pressure. Hypertension 1983; 5:264–269.
20. Manios E, Tsagalis G, Tsivgoulis G, Barlas G, Koroboki E, Michas F, et al. Time rate of blood pressure variation is associated with impaired renal function in hypertensive patients. J Hypertens 2009; 27:2244–2248.
21. Pierdomenico SD, Di Nicola M, Esposito AL, Di Mascio R, Ballone E, Lapenna D, Cuccurullo F. Prognostic value of different indices of blood pressure variability in hypertensive patients. Am J Hypertens 2009; 22:842–847.
22. Mancia G, Facchetti R, Parati G, Zanchetti A. Visit-to-visit blood pressure variability in the European Lacidipine Study on Atherosclerosis: methodological aspects and effects of antihypertensive treatment. J Hypertens 2012; 30:1241–1251.
23. Zanchetti A. Wars, war games, and dead bodies on the battlefield: variations on the theme of blood pressure variability. Stroke 2011; 42:2722–2724.
24. Omboni S, Parati G, Mancia G. The trough:peak ratio and the smoothness index in the evaluation of control of 24 h blood pressure by treatment in hypertension. Blood Press Monit 1998; 3:201–204.
25. Rizzoni D, Muiesan ML, Salvetti M, Castellano M, Bettoni G, Monteduro C, et al. The smoothness index, but not the trough-to-peak ratio predicts changes in carotid artery wall thickness during antihypertensive treatment. J Hypertens 2001; 19:703–711.
26. Webb AJ, Fischer U, Mehta Z, Rothwell PM. Effects of antihypertensive-drug class on interindividual variation in blood pressure and risk of stroke: a systematic review and meta-analysis. Lancet 2010; 375:906–915.
27. Webb AJ, Rothwell PM. Effect of dose and combination of antihypertensives on interindividual blood pressure variability: a systematic review. Stroke 2011; 42:2860–2865.
28. Ishikura K, Obara T, Kato T, Kikuya M, Shibamiya T, Shinki T, et al. J-HOME-Morning Study Group Associations between day-by-day variability in blood pressure measured at home and antihypertensive drugs: the J-HOME-Morning study. Clin Exp Hypertens 2012; 34:297–304.
29. Matsui Y, O’Rourke MF, Hoshide S, Ishikawa J, Shimada K, Kario K. Combined effect of angiotensin II receptor blocker and either a calcium channel blocker or diuretic on day-by-day variability of home blood pressure: the Japan Combined Treatment With Olmesartan and a Calcium-Channel Blocker Versus Olmesartan and Diuretics Randomized Efficacy Study. Hypertension 2012; 59:1132–1138.
30. Zhang Y, Agnoletti D, Safar ME, Blacher J. Effect of antihypertensive agents on blood pressure variability: the Natrilix SR versus candesartan and amlodipine in the reduction of systolic blood pressure in hypertensive patients (X-CELLENT) study. Hypertension 2011; 58:155–160.
31. Parati G, Schumacher H, Bilo G, Mancia G. Evaluating 24-h antihypertensive efficacy by the smoothness index: a meta-analysis of an ambulatory blood pressure monitoring database. J Hypertens 2010; 28:2177–2183.
32. Ulusoy S, Ozkan G, Konca C, Kaynar K. A comparison of the effects of fixed dose vs. single-agent combinations on 24-h blood pressure variability. Hypertens Res 2012; 35:1111–1117.
© 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins