Abstract
The prognostic value of nighttime blood pressure (BP) load in patients with chronic kidney disease (CKD) remains unknown. The prognostic value of nighttime BP load in a cohort of Chinese patients with nondialysis CKD was investigated. The authors monitored ambulatory BP and followed health outcomes in 588 Chinese CKD patients. Multivariable‐adjusted Cox regression analyses indicated that nighttime BP load was a significant risk factor for all clinical outcomes in CKD patients, even when adjusted for clinic BP. Tertile 3 of systolic BP load (vs tertile 1) was associated with an increased risk of renal events (hazard ratio [HR], 2.21; 95% confidence interval [CI], 1.12–4.38) and cardiovascular events (HR, 5.34; 95% CI, 1.58–18.04); tertile 3 of diastolic BP load (vs tertile 1) was associated with an increased risk of all‐cause mortality (HR, 6.73; 95% CI, 1.79–25.20), cardiovascular mortality (HR, 7.18; 95% CI, 1.47–35.03), renal events (HR, 2.40; 95% CI, 1.17–4.92), and cardiovascular events (HR, 5.87; 95% CI, 1.97–17.52). Higher nighttime BP load, especially nighttime diastolic BP load, was associated with a poorer prognosis in Chinese nondialysis CKD patients.
Keywords: chronic kidney disease, nighttime blood pressure load, prognosis, mortality, cardiovascular events
1. Introduction
Hypertension has been a large challenge for patients with chronic kidney disease (CKD) and contributes to 45% of deaths in men and 46% deaths in women.1, 2 Ambulatory blood pressure (BP) monitoring (ABPM) has been widely used in the management of hypertension and is a better predictor than clinical monitoring of BP for target organ damage and cardiovascular events in patients with CKD.3, 4 ABPM can provide detailed information on BP during a 24‐hour period, including mean daytime and nighttime BP as well as BP load.
As introduced by Zachariah and colleagues5, 6 and White,7 BP load is defined as the proportion of BP readings above set thresholds or the integrated area under the BP curve above the same values. BP load is recommended to be computed separately for the awake and sleeping periods of the day.7 Several experts in the field of ABPM suggest BP load as a more accurate predictor of clinical outcome than BP level. BP load for prediction of target organ damage and cardiovascular risk has been reported in hypertensive patients: nighttime systolic BP (SBP) load has shown a strong association with left ventricular mass index in children with primary hypertension8; both SBP load and diastolic BP (DBP) load are well correlated with left ventricular mass index in essential hypertension9; DBP load is correlated with microalbuminuria and intima‐media thickness independent of mean BP levels10; and daytime SBP load has been shown to predict cardiovascular events in treated octogenarians with hypertension.11 However, some experts propose that BP load predicts cardiovascular risk but does not clinically meaningfully refine the risk prediction based on 24‐hour BP level.12, 13
Information on the role of BP load in CKD patients is limited. Toprak and colleagues14 reported that nighttime SBP load was closely related to the increase in left ventricular mass index in renal transplant recipients.14 We previously reported that nighttime SBP load was associated with target organ damage in patients with nondiabetic chronic kidney disease independent of BP level.15 The question about whether nighttime BP load provides any additional information on predicting the prognosis of CKD patients remains unknown. We hypothesized that nighttime BP load has an additional role in the prognosis of CKD patients based on these cross‐sectional studies. We carried out this prospective cohort study to explore the prognostic value of nighttime BP load in nondialysis Chinese CKD patients.
2. Materials and methods
2.1. Study population
The study protocol was approved by the ethics committee of the Third Affiliated Hospital of Sun Yat‐sen University (Guangdong, China). The study protocol was approved by the institutional review board of our university. Informed consent was obtained from patients before enrollment. Consecutive patients were recruited from the Third Affiliated Hospital of Sun Yat‐sen University (Guangdong, China) from July 2010 to December 2014.
Inclusion criteria were: (1) age 14 years and older and younger than 75 years; (2) CKD; and (3) patients who promised to return to the Third Affiliated Hospital of Sun Yat‐sen University to complete the first follow‐up (ie, 6 months after the baseline survey). Exclusion criteria included patients with acute changes in the estimated glomerular filtration rate (eGFR) >30% in the previous 3 months; those undergoing dialysis; those who had undergone kidney transplantation; those with atrial fibrillation; those with cardiovascular events in the previous 3 months; those who were pregnant; those who had night‐work or shift‐work employment; those who could not tolerate ABPM; and those who had invalid ABPM data.
A total of 697 CKD patients fulfilled the inclusion criteria, 71 patients were excluded based on the detailed reasons shown in Figure 1,and 38 patients were lost to follow‐up after their first visit. Finally, 588 CKD patients were enrolled in this study (Figure 1). Patients were divided into three groups by tertiles of nighttime SBP or DBP load.
Figure 1.
Patient selection. ABPM indicates ambulatory blood pressure monitoring; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate
2.2. Measurements
2.2.1. Ambulatory BP Monitoring
Patients underwent 24‐hour ABPM using a TM‐2430 Monitor (A&D, Tokyo, Japan). Cuff size was chosen based on arm circumference and was applied to the nondominant arm. BP was recorded every 15 minutes in the daytime, and every 30 minutes at night. Monitoring was performed on a working day. Patients were asked to complete their usual activities but to keep motionless at the time of measurement. Patients had no access to ABPM values. Strenuous physical activity was discouraged in all patients during the monitoring period, and their daily activities were comparable. BP series were eliminated from the analyses if any of the following were applicable: >30% of measurements were lacking; they had missing data for >3‐hour spans; they were collected from patients who were experiencing an irregular rest‐activity schedule; or a nighttime sleep span <6 or >12 hours during monitoring.
2.2.2. BP measurement in the clinic
BP was measured for each patient during a visit to the physician. Briefly, measurements were taken from the nondominant arm in a quiet environment using a mercury sphygmomanometer with the patient in a sitting position after 5 minutes of rest. BP was not measured if the patient had consumed tobacco, ingested caffeine, or ate within the previous 30 minutes. SBP and DBP values (Korotkoff phase I and phase V, respectively) at each visit enabled recording of a minimum of three BP measurements at intervals of ≥1 minutes. Reported values of clinic BP were the mean of values recorded during the 2 days in which the ABPM device was installed and removed. For all patients, sphygmomanometric measurements were recorded by the same physician, who was not aware of the results of ABPM recordings.
2.3. Collection of other data
We collected urine samples from 7 am to 7 am the next day to detect the extent of proteinuria and sodium levels over 24 hours. These patients were asked to void their bladders before and after the urine collection. Proteinuria was measured by immunoturbidimetry. In addition, medical history, demographic information, laboratory data (hemoglobin, albumin, calcium, phosphorus, intact parathyroid hormone, serum fasting glucose, cholesterol, uric acid, serum creatinine, and serum urea nitrogen), and current therapy were obtained at the initial study visit. All experimental data were measured using a 7180 Biochemistry Auto‐Analyzer (Hitachi, Tokyo, Japan).
2.4. Definitions
ABPM daytime and ABPM nighttime were defined according to patients' schedules, respectively. BP load was the percentage of BP values reaching or exceeding 135 mm Hg systolic or 85 mm Hg diastolic during daytime, or 120 mm Hg systolic or 70 mm Hg diastolic during nighttime.5 CKD was defined according to Kidney Disease: Improving Global Outcomes (KDIGO) 2012 clinical practice guideline.16 eGFR was calculated using 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) creatinine equation.17 We divided these CKD patients into five stages (1–5) according to the KDIGO 2012 clinical practice guideline.16 Diabetes mellitus was defined as the need for antidiabetic drugs or meeting the diagnostic criteria based on the American Diabetes Association's Standards of Medical Care in Diabetes.18
2.5. Outcomes
Primary end points were time to all‐cause mortality and time to cardiovascular mortality. Secondary end points were time to renal events and time to cardiovascular events. Cardiovascular mortality was defined as death caused by cardiovascular events. Renal events were a composite of doubling of serum levels of creatinine or end‐stage renal disease (ESRD), whichever occurred first. The end point of ESRD was reached on the day of the first dialysis session.19 Cardiovascular events included a fatal or nonfatal cardiovascular event: myocardial infarction, heart failure, revascularization, stroke, and other events (acute arterial occlusion of lower extremities and thrombotic occlusion of the retinal artery), whichever occurred first. The cause of death was identified according to death certificates and autopsy reports based on the International Classification of Disease—10th Revision. Hospital records were collected to establish the diagnosis based on criteria set by the American College of Cardiology and the European Society of Cardiology.20, 21 Patients were followed up until March 31, 2016, or death, and censored on the date of the last visit to the nephrology clinic.
2.6. Statistical analyses
Data were analyzed using SPSS version 20.0 (IBM Corp, Armonk, NY, USA) and Stata version 14.0 (StataCorp, College Station, TX, USA). Descriptive statistics are presented as mean±standard deviation for continuous variables and median/interquartile range for nonparametric variables. Frequencies and percentages were used for categorical variables. Comparisons of continuous variables among groups were evaluated by analysis of variance or the Kruskal‐Wallis test. Differences among categorical variables were analyzed using the χ2 test. P values for multiple comparisons were corrected according to the Bonferroni method. We used Stata to calculate the prevalence of end points. Crude rates as well as rates standardized by the direct method for sex and age are reported. Comparison of the prevalence of events among groups was performed by the log‐rank test. The P value for multiple comparisons was corrected according to the Bonferroni method. We employed Kaplan‐Meier estimates of survival (plotted according to current recommendations) and the log‐rank test to compare survival in different groups.22 We used multivariable Cox regression models, adjusting for important predictors, to evaluate the prognostic value of nighttime BP load. Adjustment factors included age, sex (female=0, male=1), diabetes mellitus (no=0, yes=1), smoking and drinking (no=0, yes=1), body mass index, history of cardiovascular disease (no=0, yes=1), eGFR, hemoglobin, phosphate, cholesterol, proteinuria, and renin‐angiotensin system blockade (no=0, yes=1). The assumption of proportional hazards was assessed by visual judgment of log‐minus‐log survival plots. Statistical significance was set at an α level of <0.05 on two‐sided tests.
3. Results
3.1. Baseline characteristics of the study population
Of the 588 CKD patients enrolled, 397 patients had chronic glomerulonephritis, 63 had diabetic nephropathy, 32 had hypertensive nephropathy, 28 had lupus nephritis, and 68 patients had other causes of renal disease (Figure 1).
The mean age of patients was 42.8±16.7 years, and 336 were men (57.1%). The median course of disease was 6 months, and 92 patients (15.6%) had diabetes mellitus. At enrollment, 109 patients (18.5%) were current smokers, and 55 patients (9.3%) reported alcohol intake. Forty‐two patients (7.1%) among the participants had a history of cardiovascular disease (Table 1). The number of patients with CKD stage 1/2/3/4/5 was 219/102/122/72/73, respectively.
Table 1.
Baseline characteristics of patients by tertiles of SBP and DBP load
| Total (N=588) | Nighttime SBP load | Nighttime DBP load | |||||
|---|---|---|---|---|---|---|---|
| Tertile 1 | Tertile 2 | Tertile 3 | Tertile 1 | Tertile 2 | Tertile 3 | ||
| Limits, % | <16.7 | 16.7‐69.6 | ≥69.6 | <27.3 | 27.3‐75.0 | ≥75.0 | |
| Age, y | 42.8±16.7 | 35.9±13.7 | 42.7±16.9a | 49.7±16.4a , b | 35.2±15.8 | 46.5±17.3a | 46.6±14.3a |
| Men/women | 336/252 | 98/98 | 126/70 | 112/84 | 109/87 | 112/84 | 115/81 |
| Disease course, mo | 6 (1–24) | 3 (1–12) | 6 (1–24) | 6 (1–22)a | 3 (1–12) | 6 (1–24) | 6 (1–24)a |
| Diabetes mellitus, No. (%) | 92 (15.6) | 12 (6.1) | 22 (11.2)a | 58 (29.6)a , b | 14 (7.1) | 35 (17.9)a | 43 (21.9)a , b |
| Current smoker, No. (%) | 109 (18.5) | 21 (10.7) | 46 (23.5)a | 42 (21.4)a | 27 (13.8) | 47 (24)a | 35 (17.9)a |
| Alcohol intake, No. (%) | 55 (9.3) | 12 (6.1) | 19 (9.7) | 24 (12.2) | 14 (7.1) | 22 (11.2) | 19 (9.7) |
| BMI, kg/m2 | 23.2±3.6 | 22.5±3.6 | 23.1±3.6 | 23.9±3.5a | 22.3±3.7 | 23.5±3.6a | 23.7±3.5a |
| History of cardiovascular disease, No. (%) | 42 (7.1) | 5 (2.6) | 9 (4.6)a | 28 (14.3)a , b | 4 (2) | 13 (6.6) | 25 (12.8)a |
| eGFR, mL/min per 1.73 m2 | 67.8 (29.8–106.7) | 101.8 (66.6–120.5) | 64.3 (32.2–105.4)a | 35.9 (16.2–80.2)a , b | 105.1 (66.8–121.9) | 58.2 (29.1–99.0)a | 37.0 (16.3–77.6)a , b |
| Hemoglobin, g/L | 122.2±24.1 | 129.5±19.7 | 124.1±24.3a | 113.0±25.1a , b | 130.6±21.2 | 120.7±25.0a | 115.3±23.7a , b |
| Albumin, g/L | 33.3±9.0 | 33.5±9.7 | 33.4±9.1 | 33.1±8.2 | 32.7±9.9 | 33.2±9.1 | 33.9±8.0 |
| Total calcium level, mmol/L | 2.2±0.2 | 2.2±0.2 | 2.2±0.2 | 2.2±0.2 | 2.2±0.2 | 2.2±0.2 | 2.2±0.2 |
| Phosphate level, mmol/L | 1.3±0.2 | 1.2±0.2 | 1.3±0.2 | 1.3±0.3a | 1.2±0.2 | 1.2±0.2 | 1.3±0.3a |
| iPTH, pg/mL | 51.9 (31.6–87.9) | 38.6 (23.6–61.3) | 49.0 (30.6–92.6)a | 62.6 (41.7–115.0)a , b | 39.0 (23.0–63.5) | 51.1 (30.5–102.8)a | 60.3 (39.6–112.2)a , b |
| Serum fasting glucose, mmol/L | 5.1±1.4 | 4.9±1.1 | 5.0±1.2 | 5.4±1.8 | 4.9±1.1 | 5.2±1.4 | 5.2±1.7 |
| Cholesterol, mmol/L | 6.1±3.0 | 6.3±3.0 | 6.1±3.0 | 6.0±2.9 | 6.4±2.9 | 6.2±3.1 | 5.9±2.8 |
| Uric acid, mmol/L | 444.6±130.7 | 390.5±127.5 | 462.1±121.4a | 480.9±125.8a | 404.8±127.3 | 443.0±122.6a | 485.8±129.8a , b |
| Blood urea nitrogen, mmol/L | 6.8 (4.9–11.1) | 5.2 (3.8–6.7) | 7.0 (5.0–10.8)a | 9.7 (6.5–15.0)a , b | 5.0 (3.7–6.4) | 7.3 (5.4–11.9)a | 9.0 (6.5–14.7)a , b |
| Serum creatinine, μmol/L | 107.2 (71.9–192.0) | 78.7 (60.9–112.0) | 118.2 (74.2–188.0)a | 159.7 (91.3–321.6)a , b | 77.3 (59.8–110.9) | 118.3 (78.7–191.8)a | 165.1 (93.6–323.1)a , b |
| Urinary sodium excretion, mmol/24 h | 130.7±80.6 | 129.1±93.5 | 129.2±69.8 | 133.7±77.7 | 126.4±98.0 | 139.5±75.4 | 125.8±63.7 |
| Proteinuria, g/24 h | 1.7 (0.5–4.5) | 0.9 (0.2–3.8) | 1.7 (0.6–4.5)a | 2.2 (1.0–5.5)a | 1.2 (0.2–4.2) | 1.7 (0.5–4.5) | 2.0 (0.9–5.2)a |
| Clinic SBP, mm Hg | 140.2±22.9 | 125.7±16.6 | 138.6±17.2a | 156.4±23.1a , b | 17.3±1.2 | 140.4±20.5a | 153.4±22.5a , b |
| Clinic DBP, mm Hg | 85.4±13.2 | 79.3±10.7 | 85.9±11.6a | 91.0±14.4a , b | 78.1±10.0 | 84.4±11.2a | 93.6±13.3a , b |
| 24‐h SBP, mm Hg | 129.0±17.2 | 112.4±8.1 | 128.5±9.0a | 146.3±13.0a , b | 114.2±10.5 | 129.7±12.8a | 143.3±13.9a , b |
| 24‐h DBP, mm Hg | 77.8±10.0 | 69.3±5.8 | 78.4±6.9a | 85.6±9.3a , b | 68.5±5.2 | 77.3±5.7a | 87.6±7.7a , b |
| Saytime SBP, mm Hg | 130.6±17.2 | 114.5±8.9 | 130.4±9.9a | 146.9±13.8a , b | 116.4±11.0 | 131.1±13.4a | 144.2±14.2a , b |
| Daytime DBP, mm Hg | 79.0±10.1 | 70.8±6.1 | 79.8±7.5a | 86.2±9.6a , b | 70.1±5.6 | 78.5±6.6a | 88.2±8.2a , b |
| Nighttime SBP, mm Hg | 121.0±19.6 | 101.2±7.3 | 119.1±7.2a | 142.6±13.5a , b | 103.2±10.5 | 121.6±13.0a | 138.1±16.1a , b |
| Nighttime DBP, mm Hg | 71.8±11.6 | 61.4±6.6 | 71.5±6.5a | 82.4±9.7a , b | 59.9±5.3 | 71.2±3.9a | 84.2±7.8a , b |
| Receiving no antihypertensive drugs, No. (%) | 147 (25) | 66 (33.7) | 57 (29.1)a | 24 (12.2)a , b | 71 (36.2) | 48 (24.5)a | 28 (14.3)a , b |
| RAS blockade, No. (%) | 341 (58.0) | 122 (62.2) | 110 (56.1) | 109 (55.6) | 118 (60.2) | 115 (58.7) | 108 (55.1) |
| Calcium channel blocker, No. (%) | 196 (33.3) | 14 (7.1) | 59 (30.1)a | 123 (62.8)a , b | 16 (8.2) | 66 (33.7)a | 114 (58.2)a , b |
| α‐Blocker, No. (%) | 46 (7.8) | 2 (1) | 8 (4.1) | 36 (18.4)a , b | 4 (2) | 12 (6.1) | 30 (15.3)a |
| β‐Blocker, No. (%) | 94 (16.0) | 9 (4.6) | 28 (14.3)a | 57 (29.1)a , b | 8 (4.1) | 34 (17.3)a | 52 (26.5)a |
| Statins, No. (%) | 103 (17.5) | 35 (17.9) | 24 (12.2) | 44 (22.4)b | 35 (17.9) | 26 (13.3) | 42 (21.4) |
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; iPTH, intact parathyroid hormone; RAS, renin‐angiotensin system; SBP, systolic blood pressure.
Indicated comparison with tertile 1, P<.05.
Indicated comparison with tertile 2, P<.05. P value for multiple comparisons was corrected according to the Bonferroni method (3 comparisons).
Compared with patients in tertile 1, patients in tertile 3 were older; had a higher prevalence of diabetes mellitus; were current smokers; had a history of cardiovascular diseases; had higher BMIs; had a longer disease course; had higher levels of uric acid, phosphate, intact parathyroid hormone, blood urea nitrogen, and creatinine; had lower levels of hemoglobin and eGFR; more often had proteinuria; had higher clinic BP, 24‐hour BP, daytime BP, and nighttime BP; and had a higher frequency of use of calcium channel blockers, α‐blockers, and β‐blockers (P<.05) (Table 1).
3.2. Prognostic value of nighttime BP load
The median duration of follow‐up was 2.9 years (interquartile range, 2.0–4.1 years). During this period, 44 patients died and 39 of them died of cardiovascular events. A total of 140 renal and 74 cardiovascular events were recorded.
In partially multivariable‐adjusted models (adjusted for important predictors), nighttime SBP or DBP load was associated with all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events. Further adjustment for clinic SBP or DBP did not change the associations (Table 2).
Table 2.
Exploration of the prognostic value of nighttime BP load with multivariable‐adjusted cox analyses
| All‐cause mortality | Cardiovascular mortality | Renal events | Cardiovascular events | |
|---|---|---|---|---|
| Nighttime SBP load, per 1% | ||||
| Not adjusted | 1.028 (1.017–1.039), P<.001 | 1.031 (1.019–1.043), P<.001 | 1.026 (1.021–1.032), P<.001 | 1.029 (1.021–1.038), P<.001 |
| Partially adjusted | 1.017 (1.004–1.031), P=.008 | 1.019 (1.004–1.034), P=.01 | 1.013 (1.006–1.018), P<.001 | 1.015 (1.006–1.025), P=.001 |
| Fully adjusted (+clinic SBP) | 1.016 (1.002–1.03), P=.02 | 1.016 (1.001–1.032), P=.04 | 1.011 (1.004–1.018), P=.002 | 1.013 (1.003–1.023), P=.009 |
| Nighttime DBP load, per 1% | ||||
| Not adjusted | 1.025 (1.014–1.036), P<.001 | 1.025 (1.013–1.037), P<.001 | 1.026 (1.020–1.032), P<.001 | 1.024 (1.016–1.033), P<.001 |
| Partially adjusted | 1.024 (1.010–1.038), P=.001 | 1.022 (1.007–1.037), P=.003 | 1.014 (1.006–1.021), P<.001 | 1.020 (1.01–1.031), P<.001 |
| Fully adjusted (+clinic DBP) | 1.026 (1.011–1.041), P<.001 | 1.024 (1.008–1.04), P=.003 | 1.013 (1.005–1.021), P=.001 | 1.019 (1.008–1.03), P=.001 |
Data are presented as hazard ratios (95% confidence intervals), followed by P values, which express the risk per 1% increase in the nighttime blood pressure (BP) load variables.
Partially adjusted hazard ratios were adjusted for age, sex (female=0, male=1), diabetes mellitus (no=0, yes=1), smoking and drinking (no=0, yes=1), body mass index, history of cardiovascular disease (no=0, yes=1), estimated glomerular filtration rate, hemoglobin, phosphate, cholesterol, proteinuria, and renin‐angiotensin system blockade (no=0, yes=1). In fully adjusted models, nighttime BP load was additionally adjusted for clinic BP. Pearson's product‐moment correlation coefficient (r=.559, P<.001) for clinic systolic BP (SBP) and nighttime SBP load (r=.486, P<.001) for clinic diastolic BP (DBP) and nighttime DBP load.
3.3. Incidence of events
For nighttime SBP load, crude and standardized prevalence of all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events were highest in patients in tertile 3 (P<.05), and crude and standardized prevalence of cardiovascular mortality and renal and cardiovascular events were higher in patients in tertile 2 than in patients in tertile 1 (P<.05).
For nighttime DBP load, crude and standardized prevalence of all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events were higher in patients in tertile 3 than in patients in tertile 1 (P<.05); crude and standardized prevalence of all‐cause mortality and renal and cardiovascular events were higher in patients in tertile 3 than in patients in tertile 2 (P<.05); and crude and standardized prevalence of renal and cardiovascular events were higher in patients in tertile 2 than in patients in tertile 1 (P<.05) (Table 3).
Table 3.
Incidence of events by tertiles of nighttime BP load
| Nighttime SBP load | Nighttime DBP load | |||||
|---|---|---|---|---|---|---|
| Tertile 1 | Tertile 2 | Tertile 3 | Tertile 1 | Tertile 2 | Tertile 3 | |
| All‐cause mortality | ||||||
| Crude rate | 0.49 (−0.06 to 1.05) | 1.90 (0.79–3.01) | 5.22 (3.40–7.05)a , b | 0.67 (0.02–1.33) | 1.92 (0.80–3.05) | 4.86 (3.13–6.58)a , b |
| Standardized rate | 0.49 (−0.06 to 1.05) | 1.90 (0.79–3.01) | 5.23 (3.41–7.05)a , b | 0.68 (0.02–1.34) | 1.92 (0.80–3.05) | 4.87 (3.14–6.60)a , b |
| Cardiovascular mortality | ||||||
| Crude rate | 0.16 (−0.16 to 0.48) | 1.88 (0.78–2.98)a | 4.70 (2.97–6.43)a , b | 0.50 (−0.06 to 1.07) | 1.91 (0.79–3.02) | 4.18 (2.58–5.79)a |
| Standardized rate | 0.16 (−0.16 to 0.48) | 1.90 (0.78–3.02)a | 4.71 (2.98–6.44)a , b | 0.51 (−0.06 to 1.08) | 1.91 (0.79–3.02) | 4.20 (2.59–5.81)a |
| Renal events | ||||||
| Crude rate | 1.83 (0.76–2.9) | 7.12 (4.94–9.30)a | 19.86 (16.21–23.51)a , b | 1.71 (0.66–2.77) | 8.52 (6.14–10.9)a | 17.65 (14.24–21.05)a , b |
| Standardized rate | 1.81 (0.74–2.88) | 7.19 (4.94–9.44)a | 19.93 (15.88–23.98)a , b | 1.73 (0.67–2.79) | 8.52 (6.14–10.9)a | 17.69 (14.28–21.10)a , b |
| Cardiovascular events | ||||||
| Crude rate | 0.49 (−0.06 to 1.04) | 3.51 (2.00–5.02)a | 9.53 (7.04–12.02)a , b | 0.84 (0.11–1.58) | 4.46 (2.75–6.17)a | 7.83 (5.61–10.05)a , b |
| Standardized rate | 0.49 (−0.06 to 1.04) | 3.53 (2.01–5.05)a | 9.71 (7.20–12.22)a , b | 0.84 (0.11–1.58) | 4.46 (2.75–6.17)a | 7.86 (5.64–10.08)a , b |
Values are presented as rates (95% confidence intervals), expressed as number of events per 100 patient‐years. Rates are crude or standardized for sex and age (<40 and ≥40 years) by the direct method. Comparison of event rates among groups was performed by log‐rank test. P values for multiple comparisons were corrected according to the Bonferroni method (3 comparisons).
Comparison with tertile 1, P<.05.
Comparison with tertile 2, P<.05.
BP indicates blood pressure; DBP, diastolic blood pressure; SBP, systolic blood pressure.
3.4. Risks associated with nighttime BP load
The survival curves associated with tertiles of nighttime BP load are shown in Figures 2 and 3. With respect to all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events, there was a significant difference among the three survival curves of different level of nighttime SBP or DBP load (both P<.001).
Figure 2.
Kaplan‐Meier survival curves as a function of patients with different levels of nighttime systolic blood pressure (SBP) load. (a) Cumulative survival curves for all‐cause mortality in patients with different levels of nighttime SBP load. (b) Cumulative survival curves for cardiovascular mortality in patients with different levels of nighttime SBP load. (c) Cumulative survival curves for renal events in patients with different levels of nighttime SBP load. (d) Cumulative survival curves for cardiovascular events in patients with different levels of nighttime SBP load. P<.001 indicated a comparison between tertiles of nighttime SBP load
Figure 3.
Kaplan‐Meier survival curves as a function of patients with different levels of nighttime diastolic blood pressure (DBP) load. (a) Cumulative survival curves for all‐cause mortality in patients with different levels of nighttime DBP load. (b) Cumulative survival curves for cardiovascular mortality in patients with different levels of nighttime DBP load. (c) Cumulative survival curves for renal events in patients with different levels of nighttime DBP load. (d) Cumulative survival curves for cardiovascular events in patients with different levels of nighttime DBP load. P<.001 indicated a comparison between tertiles of nighttime DBP load
Multivariable‐adjusted Cox regression analyses were performed to identify the hazard ratios (HRs) with tertiles of BP load. In these models, tertile 3 of SBP load (vs tertile 1) was associated with an increased risk of renal events (HR, 2.21; 95% confidence interval [CI], 1.12–4.38) and cardiovascular events (HR, 5.34; 95%, 1.58–18.04); tertile 3 of DBP load (vs tertile 1) was associated with an increased risk of all‐cause mortality (HR, 6.73; 95% CI, 1.79–25.20), cardiovascular mortality (HR, 7.18; 95% CI, 1.47–35.03), and renal (HR, 2.40; 95% CI, 1.17–4.92) and cardiovascular events (HR, 5.87; 95% CI, 1.97–17.52) (Table 4).
Table 4.
Multivariable‐adjusted cox regression analyses of all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events in CKD patients with different levels of nighttime BP load
| All‐cause mortality | Cardiovascular mortality | Renal events | Cardiovascular events | |
|---|---|---|---|---|
| Nighttime SBP load | ||||
| Tertile 1 | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| Tertile 2 | 0.88 (0.20–3.85), P=.86 | 2.94 (0.33–25.95), P=.33 | 1.11 (0.54–2.30), P=.77 | 2.98 (0.83–10.62), P=.09 |
| Tertile 3 | 2.58 (0.69–9.64), P=.16 | 6.81 (0.85–54.67), P=.07 | 2.21 (1.12–4.38), P=.02 | 5.34 (1.58–18.04), P=.007 |
| Nighttime DBP load | ||||
| Tertile 1 | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| Tertile 2 | 3.16 (0.82–12.14), P=.09 | 4.44 (0.91–21.64), P=.06 | 1.30 (0.63–2.70), P=.47 | 3.59 (1.22–10.58), P=.02 |
| Tertile 3 | 6.73 (1.79–25.20), P=.005 | 7.18 (1.47–35.03), P=.01 | 2.40 (1.17–4.92), P=.02 | 5.87 (1.97–17.52), P=.002 |
Data are expressed as hazard ratios (95% confidence intervals), followed by P value.
Hazard ratios were adjusted for age, sex (female=0, male=1), diabetes mellitus (no=0, yes=1), smoking and drinking (no=0, yes=1), body mass index, history of cardiovascular disease (no=0, yes=1), estimated glomerular filtration rate, hemoglobin, phosphate, cholesterol, proteinuria, and renin‐angiotensin system blockade (no=0, yes=1). BP indicates blood pressure; CKD, chronic kidney disease; DBP, diastolic blood pressure; SBP, systolic blood pressure.
4. Discussion
In this prospective cohort study, we were the first to investigate the prognostic effect of nighttime BP load in Chinese nondialysis CKD patients. Nighttime BP load was a significant factor for predicting all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events even after adjusting for clinic BP, and patients with higher nighttime BP load had more clinical events than patients with lower nighttime BP load. With respect to nighttime SBP load, the highest tertile was associated with the highest risk of renal and cardiovascular events. With respect to nighttime DBP load, the highest tertile was associated with the highest risk of all‐cause mortality, cardiovascular mortality, and renal and cardiovascular events. Taken together, these results suggest that high nighttime BP load is a significant risk factor for the prognosis of Chinese nondialysis patients with CKD, and nighttime BP load might offer additional information on predicting prognosis in CKD patients in clinical practice. Special attention should be paid to nighttime BP load in the management of hypertension in patients with CKD.
Previous studies on BP load as a risk stratification factor have included only hypertensive patients, comprised only a small sample size, or were cross‐sectional.8, 9, 10, 11 However, data on the prognostic role of BP load in CKD patients are lacking. CKD patients have a different BP pattern from those without CKD and it has been reported that among uncontrolled hypertensive patients with CKD, 90.7% have nocturnal hypertension,23 Moreover, the etiology, environment, and genetic background of Chinese CKD patients are different from those in Western countries.24 It is important to further investigate the prognostic role of nighttime BP load in Chinese nondialysis CKD patients.
BP is variable during the whole day even in normotensive states. In this context, it is important to emphasize the dynamic and around‐the‐clock nature of such enhanced BP transmission. In terms of physical and mental activity, as well as body position, nighttime BP represents the minimal BP needed for adequate organ perfusion.25 A higher nighttime BP load means more opportunity for target organ damage.26 Some patients might present with higher BP load but normal mean BP levels, and BP load could provide extra information on the assessment and management of hypertension.27 Given that BP lability is further exaggerated in hypertensive states, such data also suggest that isolated measurements of mean BP level are likely to be more limited in providing an index of target organ damage, which might help to partly explain the phenomenon that the aggressive control of BP to retard the progression of CKD has yielded disappointing outcomes28 and imply that nighttime BP load may become a new target for the management of hypertension. Further prospective randomized clinical trials are needed to examine whether lowering nighttime BP load has a beneficial effect in improving the prognosis and attenuating the progression of cardiovascular and renal disease in CKD patients.
4.1. Study strengths and limitations
To the best of our knowledge, we are the first to investigate the prognostic effect of nighttime BP load in nondialysis CKD patients. We found that nighttime BP load was a significant predictor for clinical outcomes, and high nighttime BP load was associated with poor prognosis in Chinese nondialysis patients with CKD, which might provide an opportunity to improve the management of hypertension.
Some limitations should be considered when we interpret these results. First, patients in our cohort were from a single center. Second, all patients underwent only one ABPM, therefore we could not rule out subsequent changes in ABPM. Third, in order to complete the assessment, all CKD patients were inpatients. Some outpatients with CKD who presented with nonsevere proteinuria or nonsevere renal damage might have been excluded, which may lead to bias. Fourth, only one ethnic group (Chinese) was enrolled in this cohort, which limits the generalization of the study outcomes. Finally, patients did not accept standard therapy at follow‐up, and the effect of drugs cannot be ruled out.
5. Conclusions
We have provided the first evidence that nighttime BP load is a significant predictor for the prognosis of Chinese patients with nondialysis CKD. Further prospective randomized clinical trials are needed to clarify whether lowering nighttime BP load has a beneficial effect in CKD patients.
Conflict of interest
No conflicts of interest, financial or otherwise, are declared by the authors.
Acknowledgments
We would like to thank all of the patients, and their families, for participating in this study.
Li Y, Deng Q, Li H, et al. Prognostic value of nighttime blood pressure load in Chinese patients with nondialysis chronic kidney disease. J Clin Hypertens. 2017;19:890–898. 10.1111/jch.13017
Funding information
This work was supported by a training project for excellent younger scholars of the Third Affiliated Hospital of Sun Yat‐sen University (2010)
Yan Li and Qiongxia Deng contributed equally to this work.
References
- 1. Agarwal R, Andersen MJ. Blood pressure recordings within and outside the clinic and cardiovascular events in chronic kidney disease. Am J Nephrol. 2006;26:503‐510. [DOI] [PubMed] [Google Scholar]
- 2. Global Burden of Metabolic Risk Factors for Chronic Diseases Collaboration . Cardiovascular disease, chronic kidney disease, and diabetes mortality burden of cardiometabolic risk factors from 1980 to 2010: a comparative risk assessment. Lancet Diabetes Endocrinol. 2014;2:634‐647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Agarwal R, Andersen MJ. Prognostic importance of ambulatory blood pressure recordings in patients with chronic kidney disease. Kidney Int. 2006;69:1175‐1180. [DOI] [PubMed] [Google Scholar]
- 4. Minutolo R, Agarwal R, Borrelli S, et al. Prognostic role of ambulatory blood pressure measurement in patients with nondialysis chronic kidney disease. Arch Intern Med. 2011;171:1090‐1098. [DOI] [PubMed] [Google Scholar]
- 5. Zachariah PK, Sheps SG, Bailey KR, Wiltgen CM, Moore AG. Age‐related characteristics of ambulatory blood pressure load and mean blood pressure in normotensive subjects. JAMA. 1991;265:1414‐1417. [PubMed] [Google Scholar]
- 6. Zachariah PK, Sheps SG, Ilstrup DM, et al. Blood pressure load—a better determinant of hypertension. Mayo Clin Proc. 1988;63:1085‐1091. [DOI] [PubMed] [Google Scholar]
- 7. White WB. Blood pressure load and target organ effects in patients with essential hypertension. J Hypertens Suppl. 1991;9:S39‐S41. [PubMed] [Google Scholar]
- 8. Sharma AP, Mohammed J, Thomas B, Lansdell N, Norozi K, Filler G. Nighttime blood pressure, systolic blood pressure variability, and left ventricular mass index in children with hypertension. Pediatr Nephrol. 2013;28:1275‐1282. [DOI] [PubMed] [Google Scholar]
- 9. Bauwens F, Duprez D, De Buyzere M, Clement DL. Blood pressure load determines left ventricular mass in essential hypertension. Int J Cardiol. 1992;34:335‐338. [DOI] [PubMed] [Google Scholar]
- 10. Tatasciore A, Renda G, Zimarino M, et al. Awake systolic blood pressure variability correlates with target‐organ damage in hypertensive subjects. Hypertension. 2007;50:325‐332. [DOI] [PubMed] [Google Scholar]
- 11. Andrade SS, Serro‐Azul JB, Nussbacher A, et al. Daytime systolic blood pressure load and previous stroke predict cardiovascular events in treated octogenarians with hypertension. J Am Geriatr Soc. 2010;58:2232‐2234. [DOI] [PubMed] [Google Scholar]
- 12. Li Y, Thijs L, Boggia J, et al. Blood pressure load does not add to ambulatory blood pressure level for cardiovascular risk stratification. Hypertension. 2014;63:925‐933. [DOI] [PubMed] [Google Scholar]
- 13. Liu M, Li Y, Wei FF, Zhang L, Han JL, Wang JG. Is blood pressure load associated, independently of blood pressure level, with target organ damage? J Hypertens. 2013;31:1812‐1818. [DOI] [PubMed] [Google Scholar]
- 14. Toprak A, Koc M, Tezcan H, Ozener IC, Oktay A, Akoglu E. Night‐time blood pressure load is associated with higher left ventricular mass index in renal transplant recipients. J Hum Hypertens. 2003;17:239‐244. [DOI] [PubMed] [Google Scholar]
- 15. Wang C, Zhang J, Deng W, et al. Nighttime systolic blood‐pressure load is correlated with target‐organ damage independent of ambulatory blood‐pressure level in patients with non‐diabetic chronic kidney disease. PLoS One. 2015;10:e131546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Andrassy KM. Comments on ‘KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease’. Kidney Int. 2013;84:622‐623. [DOI] [PubMed] [Google Scholar]
- 17. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604‐612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. American Diabetes Association . Classification and diagnosis of diabetes. Diabetes Care. 2015;38(suppl):S8‐S16. [DOI] [PubMed] [Google Scholar]
- 19. Gabbai FB, Rahman M, Hu B, et al. Relationship between ambulatory BP and clinical outcomes in patients with hypertensive CKD. Clin J Am Soc Nephrol. 2012;7:1770‐1776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redefined—a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol. 2000;36:959‐969. [DOI] [PubMed] [Google Scholar]
- 21. Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies: a statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung, and Blood Institute. Circulation. 2003;108:2543‐2549. [DOI] [PubMed] [Google Scholar]
- 22. Pocock SJ, Clayton TC, Altman DG. Survival plots of time‐to‐event outcomes in clinical trials: good practice and pitfalls. Lancet. 2002;359:1686‐1689. [DOI] [PubMed] [Google Scholar]
- 23. Mojon A, Ayala DE, Pineiro L, et al. Comparison of ambulatory blood pressure parameters of hypertensive patients with and without chronic kidney disease. Chronobiol Int. 2013;30:145‐158. [DOI] [PubMed] [Google Scholar]
- 24. Liu ZH. Nephrology in China. Nat Rev Nephrol. 2013;9:523‐528. [DOI] [PubMed] [Google Scholar]
- 25. Gonzalez RE, Hernandez A, Dibner C, Koehler BB, Pechere‐Bertschi A. Arterial blood pressure circadian rhythm: significance and clinical implications. Rev Med Suisse. 2012;8:1709‐1712, 1714–1715. [PubMed] [Google Scholar]
- 26. Bidani AK, Griffin KA, Williamson G, Wang X, Loutzenhiser R. Protective importance of the myogenic response in the renal circulation. Hypertension. 2009;54:393‐398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Zachariah PK, Sumner WR. The clinical utility of blood pressure load in hypertension. Am J Hypertens. 1993;6:194S‐197S. [PubMed] [Google Scholar]
- 28. Bidani AK, Griffin KA, Epstein M. Hypertension and chronic kidney disease progression: why the suboptimal outcomes? Am J Med. 2012;125:1057‐1062. [DOI] [PMC free article] [PubMed] [Google Scholar]