Can ambulatory blood pressure serve as a biomarker for presence of chronic kidney disease?

It is well established that poorly controlled blood pressure (BP) is a major risk factor for chronic kidney disease (CKD) development as well as increased cardiovascular morbidity and mortality [1]. Reduction in BP to levels below 130/80 mmHg has traditionally been a key strategy for reducing the rate of CKD progression and the cardiovascular disease risk [2,3]. Although the diagnosis and treatment of hypertension has been evaluated exclusively using seated resting BP in physician offices, home and 24-h ambulatory blood pressure monitoring (ABPM) are now being utilized as an accepted method for evaluating hypertension [4,5]. ABPM offers a better representation of total BP load over the 24-h period, is noninvasive, easily reproducible, accurate and portable. Studies using ABPM predict cardiovascular events even after adjustments of conventional office BP readings [6,7]. Historically, ABPM also allows for night-time evaluation of BP dipping status in the context of target organ damage as well as cardiovascular morbidity and mortality [8,9]. People with CKD, that is, stage 3 nephropathy and evidence of vascular inflammation such as presence of microalbuminuria, frequently have resistant hypertension and require an average of one or more additional medications to help achieve BP control [10,11]. The current study by Mathisen et al.[12] is unique in that it demonstrates an association between level of BP assessed by ABPM and changes in direct measurement of glomerular filtration rate (GFR) during the earliest stages of kidney disease. The results showed a positive correlation between daytime systolic and night-time SBP and DBP with GFR levels; the higher the measured BP, the higher the measured GFR in normotensive patients. Measurements of GFR in this study were performed using high-performance liquid chromatography (HPLC) decay of serum iohexol concentration. This methodology, although valid, does have the variability to create double peaks on HPLC assessment and may not be as good as iothalamate decay curves [13]. Nevertheless, it is better than indirect measurements for GFR, such as the estimating equations from the Modification of Diet Renal Disease Study Group (MDRD) and Cockroft–Gault formulae [3,14]. The authors speculated that the kidneys undergo an initial phase of glomerular hyperfiltration as a response to elevated BP, resulting in increased GFR [12]. However, there are many problems with this assertion. First, this contention is not valid, as it is drawn from a cross-sectional rather than longitudinal study, with a single measurement of ABPM and GFR. Consequently, it does not assess autoregulation properly. Moreover, autoregulation is presumably intact in the people studied, as they do not have diabetes or other specific diseases that would impact the autoregulatory mechanisms of the kidney. It should also be appreciated that even in early diabetes, when autoregulatory mechanisms are not fully functional and most patients develop hyperfiltration, only about 30% progress to end-stage kidney disease. Thus, hyperfiltration alone is not the only key factor that predicts ultimate kidney failure. Conversely, poor control of BP at levels above 180/100 mmHg over many years does predict kidney failure [15]. The authors also point to a positive correlation between heart rate increases and GFR. The increases in heart rate may also reflect compensatory mechanisms relating to increases in the sympathetic nervous system that will also elevate BP. As previous studies have shown, an elevated heart rate is an independent predictor for development of hypertension [16]. Moreover, an elevated heart rate is a predictor of cardiovascular morbidity and mortality [17]. Thus, heart rate increases could easily result in a transient increase in GFR. In people over the age of 50 years, we expect a decrease in DBP as seen in the data from Mathisen et al. [12] with a negative correlation between DBP, particularly in daytime BP, and GFR. However, in people with stage 2 CKD, there is an initial increase in DBP. The proposed mechanism in these types of patients is the elevation of plasma nitric oxide as part of the buffering effect in the presence of increased BP. This was demonstrated in African–American patients with stage 2 CKD (GFR 60–80 ml/min per 1.73 m2) in whom an elevation in DBP occurred in the presence of elevated nitric oxide [18]. In the current study with early-stage CKD and a white cohort who is relatively young, one may not expect to see this early increase in diastolic pressure. The authors also examined pulse pressure (PP), an indirect measure of arterial stiffness. PP depends on two major factors: left ventricular ejection characteristics and aortic impedance. Although there is direct correlation between advancing stage of CKD and increases in PP, the current study of early-stage CKD noted a positive association between PP and GFR only during the day. This is in contrast to a previous longitudinal study among patients with type 2 diabetes and more advanced CKD, wherein ABPM-assessed PP was a strong predictor for nephropathy progression [19]. Thus, differences between data from the study by Mathisen et al. and existing studies relate to the younger age, namely less than 60 years, absence of diabetes and perhaps most importantly lack of a long history of poor BP control in the current study. In closing, we should note that because a cross-sectional study does not provide the strength to derive associations between ABPM and regulatory changes by the kidney, a longitudinal study is required. Moreover, although increases in GFR may have been observed, they are most likely transient, as only one assessment of GFR was made. Normally functioning kidneys through their autoregulatory ability adjust to increases in BP up to levels of 180/100 mmHg so as to protect themselves [20,21]. Clinically, there is evidence from a 17-year longitudinal study that the highest rates of end-stage kidney disease occur among those with a SBP above 180 mmHg [15] and much lower rates occur at levels below that. Thus ABPM is not a viable biomarker for changes in estimated GFR. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.