Visit-to-visit blood pressure variability: added ‘VALUE’ as a risk marker in low- and high-risk patients

This editorial refers to ‘Blood pressure variability and risk of cardiovascular events and death in patients with hypertension and different baseline risks’†, by M.H. Mehlum et al., on page 2243. The growing interest in the variability of blood pressure is related to the repeated demonstration of its independent predictive value for organ damage and cardiovascular events. In particular, visit-to-visit blood pressure variability has gained attention from physicians since large blood pressure fluctuations from one visit to another are common observations in clinical practice, and have been related to high risk of cardiovascular complications.1 , 2 Several indices are available for the quantification of visit-to-visit blood pressure variability, such as standard deviation and coefficient of variation.3 , 4 Visit-to-visit blood pressure variability can be influenced by a number of factors, including the length of follow-up, number of measures, changes in blood pressure measurement procedures, antihypertensive drugs, patient drug adherence, and environmental conditions including seasonal changes.3 , 4 It is also likely to be influenced by intrinsic characteristics of the patient, among them increased arterial stiffness and/or altered baroreflex function that impairs the buffering of blood pressure changes. Visit-to-visit blood pressure variability can damage target organs, either directly in response to the mechanical effect of blood pressure variations or indirectly in association with the mechanisms at the origin of abnormal cardiovascular regulatory mechanisms leading to higher visit-to-visit blood pressure variability. Post hoc analyses of large randomized clinical trials have been performed since both office blood pressure values at repeated visits and outcomes were available. A consistent finding was that increased visit-to-visit blood pressure variability was predictive of fatal and non-fatal cardiovascular events independently of mean blood pressure. Not surprisingly, the higher the cardiovascular risk at baseline, the higher the predictive value of visit-to-visit blood pressure variability for cardiovascular events. Although the translation of these findings into clinical practice is obvious for high-risk patients, their clinical applicability remains limited in younger patients and those with a lower mean blood pressure. Indeed, these patients are generally considered as at low-risk, a statement that can revisited if visit-to-visit blood pressure variability is available. The article by Mehlum et al. published in this issue of the European Heart Journal provides an important contribution with regard to the issue of the predictive value of visit-to-visit variability in blood pressure for cardiovascular events.5 The authors analyzed the data of the Valsartan Antihypertensive Long-term Use Evaluation (VALUE) trial,6 a multinational randomized-controlled, double-masked trial of valsartan vs amlodipine, in 15 245 patients with hypertension and at least one additional risk factor for cardiovascular event, with a mean duration of follow-up of 4.0 years. They selected 13 803 patients who had no event during the first 6 months and had at least three visits. Visit-to-visit blood pressure variability was assessed by the standard deviation of at least three mean systolic blood pressures, each calculated from three measurements during the visit. An important finding is that patients in the highest quintile of visit-to-visit blood pressure variability had a two- to threefold higher risk of cardiovascular event than those in the lowest quintile, (ischaemic stroke, myocardial infarction, and congestive heart failure). These results were adjusted for a large number of variables, including mean systolic blood pressure during treatment. In that respect, the authors confirm and extend the findings of previous large randomized clinical trials, such as the Anglo-Scandinavian Cardiac Outcomes Trial Blood Pressure Lowering Arm (ASCOT-BPLA)7 and the Losartan Intervention For Endpoint reduction in hypertension (LIFE)8 studies, i.e. increased visit-to-visit blood pressure variability was predictive of fatal and non-fatal cardiovascular events independently of mean blood pressure. The findings of Mehlum et al. 5 are also stimulating because associations between visit-to-visit blood pressure variability and cardiovascular events were stronger among the younger patients (i.e. aged 50–67 years) and patients with lower systolic blood pressure (i.e. <137.8 mmHg). This interaction with age confirms the observations reported in the ASCOT and Women Health Initiative studies,7 , 9 and strengthens them through a detailed analysis according to quintiles (Tables 3 and 4 and Figure 3 in Mehlum et al. 5). In particular, in the present analysis of the VALUE study,5 the increase in risk was steeper among patients with a lower systolic blood pressure during follow-up. These findings may appear surprising since visit-to-visit blood pressure variability and absolute cardiovascular risk were higher among older patients. An interesting explanation suggested by Mehlum et al. 5 is that younger patients are more sensitive to blood pressure variability, whereas in high-risk patients, other cardiovascular risk factors add to the effect of blood pressure variability of organ damage leading ultimately to cardiovascular events. An alternative explanation invokes the healthy survivor bias, i.e. those. who survived are more resistant than those who died. The study by Mehlum et al. 5 has several strengths that are shared by the main VALUE study:6 a large number of patients and visits during a long-duration follow-up and a standardized measurement of office blood pressure. They are also specific to the present study, including multiple sensitivity statistical analyses such as propensity score analysis, matched-case-control analysis, and binary logistic regression analysis, various parameters of blood pressure variability that consistent results. The limitations of the study have been well discussed by the authors. They include the possibility of confounding from other factors than blood pressure variability, for example mean blood pressure level during follow-up, adherence to drug treatment, and the selection of patients from visits from 6 months onward in patients with no events during the first 6 months. However, one aspect of the study that has not been discussed by the authors is the influence of drug treatment on visit-to-visit blood pressure variability. Indeed, although Mehlum et al. 5 reported that valsartan was overrepresented in the fifth quintile of blood pressure variability, and amlodipine twice more frequent in the first quartile of blood pressure variability, they did not comment on the impact of this finding on outcomes. This finding is consistent with post hoc analyses of large randomized clinical trials and their meta-analyses, which indicate that visit-to-visit blood pressure variability may be differentially affected by antihypertensive drug classes and their combinations, with larger reduction in visit-to-visit blood pressure variability with calcium channel blocker-based treatment.7 , 10 VALUE was basically a negative study; however, a marginal benefit of amlodipine over valsartan for ischaemic stroke and myocardial infarction was overcome by a marginal benefit of valsartan on heart failure and major decrease in diabetes.5 , 6 Even after the paper by Mehlum et al., it is not clear what the contribution of visit-to-visit blood pressure variability is to this imbalance regarding the benefits provided by amlodipine or valsartan. Finally, how to apply these concepts in the clinical setting? Mehlum et al. 5 recommend that patients should be monitored closely to achieve consistency of blood pressure control, particularly in younger patients and those with a lower mean blood pressure. Indeed, these patients may be considered as at low risk of cardiovascular events until the measurement of a novel risk marker, i.e. visit-to-visit blood pressure variability, is performed and a high level of variability is unmasked, thus exposing them to a higher risk of cardiovascular events. In this respect, according to the reclassification procedures suggested by Pencina et al. 11 and Hlatky et al.,12 it would be interesting to know which proportion of low-risk patients of the VALUE study would be reclassified as at high-risk after consideration of their visit-to-visit blood pressure variability. Beside the value of visit-to-visit blood pressure variability as a risk marker, a number of practical issues concerning clinical feasibility remain unanswered. One first major flaw for this concept is the fact that visit-to-visit blood pressure variability needs months or years to be assessed, which is incompatible with the current concept of quick normalization of blood pressure. Meanwhile, clinicians are interfering with the phenomenon, either by inertia (true blood pressure variations are neglected, waiting for the next normal blood pressure record) or by overreaction (treatment may change at each abnormal blood pressure reading). It is unclear whether addressing blood pressure variability implies changing algorithms or patients’ related issues. Methodology for assessing blood pressure variability is also unclear. The larger the number of visits, ideally at least eight, the more precise the predictive value of visit-to-visit blood pressure variability.7 Second, the optimal duration of follow-up is unknown. A good balance should be found: short duration, allowing a rapid estimation of visit-to-visit blood pressure variability at the cost of precision, or long duration, precise but at the cost of reactivity and practicality. Third, how can we simply calculate visit-to-visit variability in the office? In clinical practice, three blood pressure measurements can be measured at each visit, allowing the calculation of average systolic blood pressure, and then the SD of several average systolic blood pressures, as illustrated by the VALUE study.6 Dedicated apps are needed for clinical practice. Malhum et al. 5 acknowledge that visit-to-visit variability in blood pressure can be difficult to measure in clinical practice and suggest the use of a graph showing blood pressure at successive clinical visits. Alternatively, ambulatory blood pressure monitoring and mainly home blood pressure monitoring may improve the feasibility of visit-to-visit blood pressure variability measurement and clinical application if data support their use. Further studies will help to identify a standard method to obtain reproducible and valid estimates of visit-to-visit office blood pressure variability, particularly with regard to the optimal interval between visits, the number of measurements, and the measurement intervals between visits.4 In conclusion, although a number of theoretical and practical issues remain to be addressed before using visit-to-visit blood pressure variability in clinical practice, the study by Mehlum et al. 5 provides a valuable contribution to the ongoing research on whether antihypertensive treatment should not only aim at lowering mean blood pressure, but also at reducing visit-to-visit blood pressure variability, and in which category of patients (low-risk vs high-risk), with a final objective of better preventing cardiovascular events (Take home figure). Beside averaged blood pressure, considered as the Gold standard cardiovascular risk marker, there is a growing body of evidence for visit-to-visit blood pressure variability to be a novel risk marker. Particularly, the determinants of visit-to-visit blood pressure variability include methodologic and intrinsic features. Beside averaged blood pressure, visit-to-visit blood pressure variability can contribute to increase target organ damage and thus the risk of cardiovascular and renal events. This work was supported by the Institut national de la santé et de la recherche médicale, Assistance-Publique Hopitaux de Paris, and Paris Descartes University.