The Need to Reduce Variability in the Study of Blood Pressure Variability

Related Article, p. 384 Related Article, p. 384 Blood pressure (BP) variability describes the degree to which BP measures are dispersed during a specified time frame. Studying the prognostic significance of BP variability is challenging owing to the myriad options for the time frame used to define it. At one end of the spectrum, very short-term BP variability describes beat-to-beat or minute-to-minute variation, such as when using BP measured continuously using an intra-arterial catheter. At the other end of the spectrum, long-term BP variability describes variation over a period of months or years, such as when using BP measured during periodic clinic or study visits. Intermediate to these options is short-term BP variability measured using 24-hour ambulatory blood pressure monitoring (ABPM).1Parati G. Ochoa J.E. Lombardi C. Bilo G. Assessment and management of blood-pressure variability.Nat Rev Cardiol. 2013; 10: 143-155Crossref PubMed Scopus (536) Google Scholar Further complicating the study of BP variability is the variety of metrics used to calculate BP variability within a given time frame (Table 1). Among the simplest and most common BP variability metric is the standard deviation (SD), which reflects the dispersion of BP measures around the mean BP. Because higher mean BP is associated with higher SD, the coefficient of variation (CV), calculated by dividing the SD by the mean BP, is commonly used. The CV is easy to compute and reduces the correlation with mean BP but does not remove it completely, unlike the metric known as the variation independent of the mean. The variation independent of the mean can be computationally more complex to derive, as it is calculated by dividing the SD by the mean BP raised to the power of x, where x is obtained by fitting a nonlinear regression model.2Rothwell P.M. Howard S.C. Dolan E. et al.Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension.Lancet. 2010; 375: 895-905Abstract Full Text Full Text PDF PubMed Scopus (1315) Google ScholarTable 1Short-Term Blood Pressure Variability Metrics and Selected Studies of Their Association With Kidney OutcomesMetric of BP VariabilityStandard Deviation (SD)Weighted SD (wSD)Coefficient of Variation (CV)Variation Independent of the Mean (VIM)Average Real Variability (ARV)DescriptionDispersion around the meanIndividual daytime and nighttime SD weighted by hours in each periodDispersion normalized to the meanDistribution of BP values normalized to the sample meanAverage absolute difference between consecutive BP measurementsStrengthEasy to calculateLess affected by dipping patternsEasy to calculate, less correlation with mean BPNot correlated with mean BPConsiders the order of measurements, less affected by frequency of measurementsWeaknessesAffected by nighttime dipping patterns, correlated with mean BPNo standard method of determining nighttime periodsAffected by nighttime dipping patternsMore complex to computeCan still be affected by nighttime dipping patternsAssociation with albuminuriaFarrag et al16Farrag H.M.A. Amin A.S. Abdel-Rheim A.R. Relation of short-term blood pressure variability to early renal effects in hypertensive patients with controlled blood pressure.Blood Press Monit. 2019; 24: 221-224Crossref PubMed Scopus (11) Google Scholar: positive association (r = 0.564, P < 0.001)Mulè et al7Mulè G. Calcaterra I. Costanzo M. et al.Average real variability of 24-h systolic blood pressure is associated with microalbuminuria in patients with primary hypertension.J Hum Hypertens. 2016; 30: 164-170Crossref PubMed Scopus (23) Google Scholar: no correlationMadden et al17Madden J.M. O’Flynn A.M. Dolan E. Fitzgerald A.P. Kearney P.M. Short-term blood pressure variability over 24 h and target organ damage in middle-aged men and women.J Hum Hypertens. 2015; 29: 719-725Crossref PubMed Scopus (22) Google Scholar: no association (OR, 0.89 [0.72-1.09])Mulè et al7Mulè G. Calcaterra I. Costanzo M. et al.Average real variability of 24-h systolic blood pressure is associated with microalbuminuria in patients with primary hypertension.J Hum Hypertens. 2016; 30: 164-170Crossref PubMed Scopus (23) Google Scholar: no correlationMadden et al17Madden J.M. O’Flynn A.M. Dolan E. Fitzgerald A.P. Kearney P.M. Short-term blood pressure variability over 24 h and target organ damage in middle-aged men and women.J Hum Hypertens. 2015; 29: 719-725Crossref PubMed Scopus (22) Google Scholar: no association (OR, 1.10 [0.89-1.35])Farrag et al16Farrag H.M.A. Amin A.S. Abdel-Rheim A.R. Relation of short-term blood pressure variability to early renal effects in hypertensive patients with controlled blood pressure.Blood Press Monit. 2019; 24: 221-224Crossref PubMed Scopus (11) Google Scholar: positive association (r = 0.579, P < 0.001)Madden et al17Madden J.M. O’Flynn A.M. Dolan E. Fitzgerald A.P. Kearney P.M. Short-term blood pressure variability over 24 h and target organ damage in middle-aged men and women.J Hum Hypertens. 2015; 29: 719-725Crossref PubMed Scopus (22) Google Scholar: no association (OR, 0.89 [0.73-1.08])Wei et al18Wei F.F. Li Y. Zhang L. et al.Beat-to-beat, reading-to-reading, and day-to-day blood pressure variability in relation to organ damage in untreated Chinese.Hypertension. 2014; 63: 790-796Crossref PubMed Scopus (101) Google Scholar: no associationMulè et al7Mulè G. Calcaterra I. Costanzo M. et al.Average real variability of 24-h systolic blood pressure is associated with microalbuminuria in patients with primary hypertension.J Hum Hypertens. 2016; 30: 164-170Crossref PubMed Scopus (23) Google Scholar: correlation (r = 0.131, P = 0.02)Madden et al17Madden J.M. O’Flynn A.M. Dolan E. Fitzgerald A.P. Kearney P.M. Short-term blood pressure variability over 24 h and target organ damage in middle-aged men and women.J Hum Hypertens. 2015; 29: 719-725Crossref PubMed Scopus (22) Google Scholar: no association (OR, 1.20 [0.97-1.47])Association with KRTWang et al19Wang Q. Wang Y. Wang J. Zhang L. Zhao M.H. C-STRIDE (Chinese Cohort Study of Chronic Kidney Disease). Short-term systolic blood pressure variability and kidney disease progression in patients with chronic kidney disease: results from C-STRIDE.J Am Heart Assoc. 2020; 9e015359Crossref Scopus (10) Google Scholar: association (HR, 1.47 [1.09-1.99])Borrelli et al20Borrelli S. Garofalo C. Mallamaci F. et al.Short-term blood pressure variability in nondialysis chronic kidney disease patients: correlates and prognostic role on the progression of renal disease.J Hypertens. 2018; 36: 2398-2405Crossref PubMed Scopus (23) Google Scholar: no association (HR, 1.16 [0.70-1.92])Borrelli et al20Borrelli S. Garofalo C. Mallamaci F. et al.Short-term blood pressure variability in nondialysis chronic kidney disease patients: correlates and prognostic role on the progression of renal disease.J Hypertens. 2018; 36: 2398-2405Crossref PubMed Scopus (23) Google Scholar: no association (HR, 0.97 [0.89-1.06])Based on BP measurements collected every 15-30 minutes over 24 hours using ambulatory blood pressure monitoring. Values in brackets are 95% CI. Abbreviations: BP, blood pressure; HR, hazard ratio; OR, odds ratio; KRT, kidney replacement therapy (dialysis or kidney transplant). Open table in a new tab Based on BP measurements collected every 15-30 minutes over 24 hours using ambulatory blood pressure monitoring. Values in brackets are 95% CI. Abbreviations: BP, blood pressure; HR, hazard ratio; OR, odds ratio; KRT, kidney replacement therapy (dialysis or kidney transplant). The timing and order of BP measurements can also be important to consider when calculating BP variability, particularly for short-term BP variability measured using 24-hour ABPM. Blood pressure usually falls during the asleep period as part of the circadian rhythm. This normal nocturnal BP dipping is generally associated with lower risks of adverse clinical events3de la Sierra A. Gorostidi M. Banegas J.R. Segura J. de la Cruz J.J. Ruilope L.M. Nocturnal hypertension or nondipping: which is better associated with the cardiovascular risk profile?.Am J Hypertens. 2014; 27: 680-687Crossref PubMed Scopus (92) Google Scholar but will contribute to higher measured BP variability. To minimize this issue, short-term BP variability can be calculated separately for daytime and nocturnal periods with use of weighting to account for the differences in the durations of these periods.4Bilo G. Giglio A. Styczkiewicz 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: 2058-2066Crossref PubMed Scopus (177) Google Scholar Another metric, called the average real variability (ARV), accounts for the order of BP measurements on an even finer scale, as it averages the absolute differences between each consecutive BP measurement taken over a period of time.5Mena L. Pintos S. Queipo N.V. Aizpurua J.A. Maestre G. Sulbaran T. A reliable index for the prognostic significance of blood pressure variability.J Hypertens. 2005; 23: 505-511Crossref PubMed Scopus (338) Google Scholar In this issue of AJKD, Jhee et al6Jhee J.H. Oh D. Seo J. et al.Short-term blood pressure variability and incident CKD in patients with hypertension: findings from the Cardiovascular and Metabolic Disease Etiology Research Center-High Risk (CMERC-HI) Study.Am J Kidney Dis. 2023; 81: 384-393https://doi.org/10.1053/j.ajkd.2022.08.017Abstract Full Text Full Text PDF Scopus (2) Google Scholar studied 1,173 participants in the Cardiovascular and Metabolic Disease Etiology Research Center–High Risk (CMERC-HI) Study with hypertension and normal baseline kidney function (mean baseline estimated glomerular filtration rate [eGFR]: 89.6 ± 13.8 mL/min/1.73 m2). Using BP readings from one 24-hour ABPM, the authors sought to examine the association of short-term BP variability with incident chronic kidney disease (CKD), defined as a sustained decline in eGFR of ≥30% from baseline, eGFR <60 mL/min/1.73 m2, or overt proteinuria (urinary albumin-creatinine ratio ≥300 mg/g). One notable strength of this study is that the laboratory tests used to define incident CKD were performed at 3-month intervals according to a prespecified protocol, in contrast to studies that used laboratory measurements collected during routine clinical care and that therefore may have been prone to ascertainment bias. Jhee et al used 3 different metrics to define short-term systolic and diastolic BP variability: ARV, SD, and CV. They found that higher systolic and diastolic BP variability defined using ARV was consistently associated with a higher risk of incident CKD whether BP variability was modeled as a continuous or categorical variable, in fully adjusted models that included mean 24-hour BP, or in models that treated death as a competing risk. Even though ARV was strongly correlated with SD in their study (Pearson r = 0.79 for systolic BP variability), they found no significant association of BP variability with incident CKD when defined using SD or CV. Dipping status also had no significant association with incident CKD. A significant association of ARV and kidney outcomes has also been seen in other studies. For example, in a study of a similar cohort of patients without CKD but with hypertension, higher systolic but not diastolic short-term BP variability defined using the ARV was associated with incident albuminuria in models adjusting for mean 24-hour systolic BP, whereas BP variability defined using weighted SD was not.7Mulè G. Calcaterra I. Costanzo M. et al.Average real variability of 24-h systolic blood pressure is associated with microalbuminuria in patients with primary hypertension.J Hum Hypertens. 2016; 30: 164-170Crossref PubMed Scopus (23) Google Scholar A series of studies conducted in a healthy cohort of self-identified African Americans showed that higher systolic and diastolic ARV was associated with decreased endothelial and smooth muscle function, but no other metrics of short-term BP variability were used in those studies for comparison.8Diaz K.M. Veerabhadrappa P. Kashem M.A. et al.Visit-to-visit and 24-h blood pressure variability: association with endothelial and smooth muscle function in African Americans.J Hum Hypertens. 2013; 27: 671-677Crossref PubMed Scopus (58) Google Scholar,9Diaz K.M. Veerabhadrappa P. Kashem M.A. et al.Relationship of visit-to-visit and ambulatory blood pressure variability to vascular function in African Americans.Hypertens Res. 2012; 35: 55-61Crossref PubMed Scopus (102) Google Scholar A systematic review and meta-analysis found that 17 of 19 studies using ARV showed a significant association of higher ARV with risk of target organ (mostly cardiovascular) damage, whereas alternative BP variability metrics such as SD, weighted SD, and CV did not perform as well.10Mena L.J. Felix V.G. Melgarejo J.D. Maestre G.E. 24-Hour blood pressure variability assessed by average real variability: a systematic review and meta-analysis.J Am Heart Assoc. 2017; 6e006895Crossref PubMed Scopus (106) Google Scholar The ARV may have other advantages over other short-term BP variability metrics in that it is easy to calculate, reflects the sequential order of BP measurements, and is less sensitive to normal circadian rhythms, leading some researchers to advocate for ARV to be the standard short-term BP variability metric. However, whether ARV versus SD or CV better represents vascular pathophysiology remains unclear. It is also unclear whether systolic or diastolic BP variability is more consistently associated with outcomes. In the meta-analysis discussed above,10Mena L.J. Felix V.G. Melgarejo J.D. Maestre G.E. 24-Hour blood pressure variability assessed by average real variability: a systematic review and meta-analysis.J Am Heart Assoc. 2017; 6e006895Crossref PubMed Scopus (106) Google Scholar 9 of the included studies showed that both systolic and diastolic ARV were significantly associated with cardiovascular outcomes, 5 studies showed an association with systolic ARV only, and 2 studies showed an association with diastolic ARV only. Similarly mixed outcomes were shown among the studies examining the association of systolic and diastolic ARV with kidney outcomes. What is clear is that the lack of standardization of BP variability metrics will continue to hinder the field from moving forward. A scoping review of methodological factors in 110 short-term BP variability studies using ABPM confirmed that these studies varied considerably: frequency of BP readings ranged from every 15 to every 75 minutes, the number of BP readings ranged from 32 to 64, and the definition of the daytime/awake period was determined using fixed times or patient diaries and ranged from 10 to 17 hours.11Veloudi P. Sharman J.E. Methodological factors affecting quantification of blood pressure variability: a scoping review.J Hypertens. 2018; 36: 711-719Crossref PubMed Scopus (14) Google Scholar These differences make it difficult to compare results between studies to draw any definitive conclusions about the usefulness of BP variability to reliably identify a group at higher risk for adverse clinical events among patients with similar mean BP. Even if a consensus is reached on how best to define short-term BP variability, questions remain on whether it can and should be treated, and what is considered normal versus elevated BP variability. Cross-sectional studies suggest that in patients treated with 1 antihypertensive, short-term BP variability is lower when calcium channel blockers (CCB) or thiazides are used, compared to angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or β-blockers.12Levi-Marpillat N. Macquin-Mavier I. Tropeano A.I. Parati G. Maison P. Antihypertensive drug classes have different effects on short-term blood pressure variability in essential hypertension.Hypertens Res. 2014; 37: 585-590Crossref PubMed Scopus (46) Google Scholar,13de la Sierra A. Mateu A. Gorostidi M. Vinyoles E. Segura J. Ruilope L.M. Antihypertensive therapy and short-term blood pressure variability.J Hypertens. 2021; 39: 349-355Crossref PubMed Scopus (5) Google Scholar In patients with dual antihypertensive therapy, a combination that included a CCB had lower BP variability.12Levi-Marpillat N. Macquin-Mavier I. Tropeano A.I. Parati G. Maison P. Antihypertensive drug classes have different effects on short-term blood pressure variability in essential hypertension.Hypertens Res. 2014; 37: 585-590Crossref PubMed Scopus (46) Google Scholar,13de la Sierra A. Mateu A. Gorostidi M. Vinyoles E. Segura J. Ruilope L.M. Antihypertensive therapy and short-term blood pressure variability.J Hypertens. 2021; 39: 349-355Crossref PubMed Scopus (5) Google Scholar The third-generation non-dihydropyridine CCB amlodipine was the most commonly used in these studies. Amlodipine is a long-acting CCB with delayed absorption and >95% protein binding leading to delayed peak levels and longer duration of action,14Abernethy D.R. Schwartz J.B. Calcium-antagonist drugs.N Engl J Med. 1999; 341: 1447-1457Crossref PubMed Scopus (368) Google Scholar,15Messerli F.H. Calcium antagonists in hypertension: from hemodynamics to outcomes.Am J Hypertens. 2002; 15: 94S-97SCrossref PubMed Google Scholar which could help to explain its association with lower BP variability. Future interventional studies are needed to confirm these findings not only on lowering BP variability but on kidney and cardiovascular outcomes. Such studies will be challenging to design without consensus and standardization of BP variability metrics as a first step. Advancing our understanding of the clinical significance of BP variability and whether treatments targeting BP variability will improve outcomes requires reducing the variability of studies of BP variability. Mario Funes Hernandez, MD, and Tara I. Chang, MD, MS. None. The authors declare that they have no relevant financial interests. Received October 11, 2022, in response to an invitation from the journal. Direct editorial input from an Associate Editor and a Deputy Editor. Accepted in revised form October 27, 2022. Short-term Blood Pressure Variability and Incident CKD in Patients With Hypertension: Findings From the Cardiovascular and Metabolic Disease Etiology Research Center–High Risk (CMERC-HI) StudyAmerican Journal of Kidney DiseasesVol. 81Issue 4PreviewThe association between short-term blood pressure variability (BPV) and kidney outcomes is poorly understood. This study evaluated the association between short-term BPV and kidney disease outcomes in people with hypertension. Full-Text PDF