Ambulatory BP Phenotypes and Their Association with Target Organ Damage and Clinical Outcomes in CKD

Hypertension among people with CKD is common, remains often inadequately controlled, and represents an established risk factor for adverse cardiovascular and kidney outcomes (¹). Routine office BP measurements, the current cornerstone of hypertension management, provide imprecise reflection of the actual BP load recorded with the “gold standard” method of ambulatory BP monitoring. A 2009 meta-analysis of six studies (incorporating data from 980 participants with CKD) showed that the overall prevalence of masked hypertension was 8.3% and that the prevalence of white coat hypertension was 18.3% (²). Notably, 40.3% of participants with CKD who were classified as normotensives (or adequately controlled hypertensives) on the basis of office BP recordings, in fact, had uncontrolled hypertension outside of the office (²). The prevalence of masked hypertension was reported to be even higher in studies assessing conjointly daytime and nighttime BP measurements in the definition of masked hypertension (³). This is to be expected if we take into consideration that circadian BP rhythms are markedly impaired and that the burden of nocturnal hypertension is high among patients with CKD (¹). Because ambulatory BP monitoring facilitates the identification of specific BP phenotypes (such as masked, white coat, isolated nocturnal hypertension, etc.), the wider adoption of this technique may improve the management of hypertension, particularly in those with CKD.

In this editorial, we explore the complex association of ambulatory BP phenotypes with target organ damage and clinical outcomes in CKD on the basis of two analyses of the Jackson Heart Study and the Chronic Renal Insufficiency Cohort (CRIC) study published in this issue of CJASN.

In the first study, a comparative analysis of blacks with drug-treated hypertension participating in the Jackson Heart Study, Mwasongwe et al. (⁴) showed that, compared with participants without CKD (n=447), those with CKD (n=144) had a higher multivariable-adjusted prevalence of uncontrolled office BP (prevalence ratio, 1.44; 95% confidence interval [95% CI], 1.02 to 2.02) and sustained uncontrolled hypertension (prevalence ratio, 1.66; 95% CI, 1.16 to 2.36) (⁴). When the analysis was stratified by the level of eGFR and by the presence of albuminuria, it was albuminuria—not reduced eGFR—that was associated with higher multivariable-adjusted prevalence of uncontrolled office, daytime, and nighttime BP; a nondipping BP pattern; and sustained uncontrolled hypertension (⁴). These associations persisted in sensitivity analyses using 24-hour instead of daytime BP and the 2017 American Heart Association/American College of Cardiology BP classification to define hypertension phenotypes (⁴).

These associations are not surprising and come in accordance with the results of a 2005 analysis aiming to identify the independent determinants of systolic hypertension in 232 veterans with CKD (⁵). Among 17 risk factors tested for interaction in multivariable regression models, it was the spot urinary protein-creatinine ratio that exhibited the strongest correlation with systolic BP, and this correlation remained unmodified regardless of the technique used to diagnose hypertension (office, home, or ambulatory BP monitoring) (⁵). In contrast, eGFR was not an independent determinant of systolic BP by any technique (⁵).

In a 2009 analysis of 336 veterans with or without CKD, when eGFR and protein-creatinine ratio were inserted conjointly in regression models, proteinuria—not eGFR—remained an independent determinant of the severity of hypertension (⁶). When circadian variation in BP was assessed with a cosinor model, compared with the staging of CKD, proteinuria was more influential in the disruption of diurnal BP rhythms (⁶).

Taken together, the established belief that the severity of hypertension and disruption in circadian BP variation parallels more closely with the eGFR decline needs to be reappraised. It seems that, at any stage of CKD, patients with versus those without proteinuria are more likely to have uncontrolled ambulatory hypertension and/or a nondipping BP pattern.

In this report from the Jackson Heart Study, Mwasongwe et al. (⁴) also evaluated the association of ambulatory BP phenotypes with target organ damage. Among participants without CKD, there was no association between hypertension categories and echocardiographically documented left ventricular hypertrophy. However, among those with CKD, uncontrolled daytime BP (prevalence ratio, 2.19; 95% CI, 1.04 to 4.65), uncontrolled nighttime BP (prevalence ratio, 3.71; 95% CI, 1.15 to 11.93), and masked uncontrolled hypertension (prevalence ratio, 2.23; 95% CI, 1.00 to 4.98) were cross-sectionally associated with higher multivariable-adjusted prevalence of left ventricular hypertrophy (⁴).

The longitudinal inter-relation of left ventricular mass growth with office and ambulatory BP was explored in a recent cohort study of 274 veterans with CKD who were prospectively followed for ≤4 years (⁷). There was a rapid increase in the rate of left ventricular mass growth over the first 2 years of follow-up, but this rate plateaued thereafter. Systolic BP increased linearly during follow-up, with a rate of 1.7 mm Hg/yr for office recordings and with a rate of 1.8 mm Hg/yr for 24-hour ambulatory recordings. Routine office BP (triplicate recordings at the timing of echocardiography without a prespecified 5-minute seated rest) could not predict the longitudinal growth of left ventricular mass index. In contrast, the predictive value of “research-grade” office BP (triplicate recordings obtained after a 5-minute seated rest over three consecutive visits) was equal to that of 24-hour ambulatory BP (⁷). A linear increment in left ventricular mass index was evident among participants with controlled, masked uncontrolled, and sustained uncontrolled hypertension when classification was on the basis of 24-hour and daytime BP (P=0.05 for the trend) (⁷) but not on nighttime BP. Additive value in prediction of left ventricular mass index both cross-sectionally and longitudinally was provided by the masked hypertension effect of nighttime BP (i.e., the difference between nighttime and office systolic BP) (⁷).

If the above-described cross-sectional and longitudinal associations are causal, then interventions to normalize these BP phenotypes (i.e., antihypertensive therapy guided by ambulatory BP monitoring) may be more effective in causing regression of left ventricular hypertrophy.

In the second study discussed in this editorial, a cross-sectional and longitudinal analysis of ambulatory BP data from 1502 participants in the CRIC study, Ghazi et al. (⁸) showed the lack of association of BP and dipping patterns with prevalent or incident frailty. Paradoxically, there was also no association of dipping status with physical activity assessed with the short physical performance battery (SPPB) score (⁸). It has to be noted, however, that such associations may exist and may be bidirectional. Either in the cross-sectional or in the longitudinal part of this analysis, ambulatory BP pattern was the exposure; frailty or SPPB score was the outcome. The vice versa may also occur. Because the sleep-activity cycle is the master regulator of circadian BP rhythms (¹), it is likely that a disrupted sleep-activity cycle in a frail patient with minimal physical activity during daytime may be the cause of a disrupted circadian variation in BP. Similarly, the high physical activity during nighttime is the profound explanation for the blunted dipping BP pattern in night-shift workers. Another example is that of astronauts, in whom circadian variation in BP is also blunted, mainly because the zero gravity environment in space results in relative physical inactivity that disturbs their normal sleep-activity cycle (¹).

Despite the fact that this analysis of the CRIC study failed to show that ambulatory BP patterns correlate with SPPB score, an earlier analysis of 103 veterans with CKD showed that the level of physical activity as assessed directly with the use of actigraphy over 2 weeks was an independent determinant of circadian BP profile described by either the dichotomous dipper-based or the phase-based rhythm classification (⁹). Once again, the observational nature of these data cannot demonstrate direct cause and effect associations. Randomized trials are clearly warranted to overcome the limitation of reverse causality and elucidate whether interventions aiming to improve physical functioning translate into a benefit on restoration of diurnal BP rhythms.

In this analysis of the CRIC study, dipping BP patterns were not associated with prevalent cognitive impairment assessed with the Modified Mini-Mental Status Examination (⁸). In longitudinal analysis, compared with normal dipping, extreme dipping was associated with a marginally greater incidence of cognitive impairment (hazard ratio, 1.83; 95% CI, 0.99 to 3.34) (⁸). Nondipping and reverse dipping had no association with incident cognitive dysfunction (⁸). Once again, this lack of association should not be interpreted as absence of a direct cause and effect relationship between disrupted circadian BP patterns and cognitive impairment. Emerging clinical trial evidence suggests that intensive BP lowering may be beneficial in causing regression of cerebral microvascular damage. As an example, in the Intensive Versus Standard Ambulatory Blood Pressure Lowering to Prevent Functional Decline in the Elderly trial (¹⁰), 199 elderly (≥75 years of age) patients with hypertension and evidence of white matter disease on screening magnetic brain imaging were randomized to a lower (<130 mm Hg) versus a standard (<145 mm Hg) 24-hour systolic BP target. These prespecified BP goals were achieved over a median treatment period of 3–4 months, and the average separation thereafter was a between-group difference of 16.3 mm Hg in 24-hour systolic BP. Despite the fact that changes in mobility outcomes or cognitive function during the 3-year-long follow-up did not differ between groups, intensive ambulatory BP lowering provoked a significant regression of the white matter hyperintensity lesions (¹⁰). These preliminary results offer promise that larger trials with longer treatment periods in the future may prove that intensive therapy guided by ambulatory BP monitoring is an effective strategy to improve the brain health and mobility outcomes.

Disclosures

Dr. Agarwal is a member of data safety monitoring committees for Astra Zeneca and Ironwood Pharmaceuticals; is a member of steering committees of randomized trials for Akebia, Bayer, Genzyme US Companies, Glaxo Smith Kline, Janssen, Relypsa, and Sanofi; is a member of adjudication committees for Bayer, Boehringer Ingelheim, and Janssen; and is a member of scientific advisory boards or a consultant for Celgene, Daiichi Sankyo, Inc., Eli Lilly, Relypsa, Reata, Takeda Pharmaceuticals, USA, and ZS Pharma. Dr. Georgianos has nothing to disclose.

Funding

Dr. Agarwal is supported by National Institutes of Health grant 5 R01 HL126903-02 and US Department of Veterans Affairs grant VA Merit Review 5I01CX000829-04.