Prognostic Value of Blood Pressure Rhythmicity for Estimated Glomerular Filtration Rate in Male Hypertensive Patients Aged 55 and Older

Abstract

Introduction

Blood pressure (BP) exhibits a circadian rhythm characterized by higher levels during wakefulness and lower levels during sleep; however, the functional and structural impact of the rhythms of BP remains uncertain.

Methods

Two hundred hypertensive males aged 55 and older without overt cardiovascular or cerebrovascular diseases were enrolled in this longitudinal study. Of these, 188 were included in the analyses (12 lacked valid BP records for part of the 24-h period). Rhythmic profiling of BP was performed using ARSER, and rhythmicity was considered significant at p < 0.05. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology (CKD-EPI) formula. The primary outcome was the change in eGFR.

Results

The average age was 64.9 ± 7.2 years. For systolic BP (SBP), 38 of the subjects exhibited a 12-h rhythm and 43 subjects a 24-h rhythm; for diastolic BP (DBP), 38 exhibited a 12-h rhythm, and 36 exhibited a 24-h rhythm. During the 3-year follow-up period, 16 of the subjects died, and 36 were lost to follow-up. The mean eGFR at baseline and follow-up were, respectively, 86.6 ± 14.0 and 81.0 ± 17.1 mL min⁻¹ 1.73 m⁻² (p = 0.001). The urinary albumin:creatinine ratio did not vary significantly among the groups (p = 0.059). Subjects with 12-h rhythmic SBP exhibited a smaller reduction in eGFR than those with arrhythmic SBP (p = 0.014). However, the changes in eGFR were similar among the groups displaying 12-h or 24-h rhythmic DBP or arrhythmic DBP. We defined a decline in eGFR as a reduction of >1/2 SD between baseline and follow-up. Adjusting for confounding factors (including age, smoking, alcohol consumption, diabetes mellitus, BMI, albumin levels, administration time of antihypertensive drugs, and duration of hypertension), the risk of a decline in eGFR was 70% lower in subjects with 12-h rhythmic SBP than in those with arrhythmic SBP (heart rate = 0.307 [0.108–0.874], p = 0.027).

Conclusion

SBP with a 12-h period is a protective predictor of the decline in eGFR in hypertensive males. It is, therefore, necessary to focus on the rhythmic profiling of BP.

Introduction

After diabetes, hypertension is the second leading cause of end-stage renal disease [1]. Recent evidence suggests that there may be common molecular mechanisms, including oxidative stress, the renin-angiotensin system, endothelial dysfunction, and sex differences, underlying its impact on kidney disease [2]. Lowering blood pressure (BP) is critical to slow the progression of kidney disease [3]. Nocturnal BP is the strongest predictor of renal events, owing to the dipper BP profile pattern, defined as a reduction in BP of at least 10% at night [4, 5]. The Hygia Chronotherapy Trial found that significantly fewer cardiovascular events occurred when hypertension treatment was administered at night than in the morning [6]. Further, that trial found that the highest drug concentration coincided closely with the peak of most of the key circadian determinants of the 24-h BP profile, with clinically relevant effects on the endogenous BP rhythm [6].

In the general population, BP exhibits a characteristic rhythmic pattern, with two daytime peaks (morning and afternoon/early evening), a minor midday nadir, and a decline during nighttime sleep [7]. The rhythm of BP is controlled by the central clock located in the suprachiasmatic nucleus of the hypothalamus and peripheral clocks distributed throughout the body. The suprachiasmatic nucleus directly receives photic cues for entrainment of the light-dark cycle and for synchronizing the peripheral clocks. These peripheral clocks can be resynchronized by eating behavior [8]. The molecular clocks in the kidney, vasculature, heart, brain, and nervous system are all peripheral-clock regulators of BP [7].

However, nycthemeral variation in BP is altered in hypertensive individuals [9]. Several clinical and epidemiological studies have demonstrated that hypertension and disruption of BP rhythm contribute to adverse health outcomes, including cardiovascular disease, dementia, and chronic kidney disease [10]. A 10% increase in the diurnal difference in systolic BP (SBP) is associated with a 1.21-fold higher risk of chronic kidney disease [11]. Notably, BP rhythm is typically defined based on the dipping ratio or using the coefficient of variation of BP calculated based on its mean and standard deviation. Nonetheless, circadian BP patterns in humans remain significantly underexplored.

To elucidate these aspects, we focused on men aged 55 and older with hypertension and investigated their circadian phenotypes and the effects of BP on renal outcomes. These findings elucidate the interplay between hypertension, circadian rhythm, and renal outcomes, expanding the knowledge of renal complications of hypertension in men aged 55 and older.

Methods

Study Population

The participants were from the cohort of Brain and Vascular Health in the Elderly (BRAVE-1). The study consecutively recruited 200 older males (aged ≥55 years). Inclusion criteria required participants to have a diagnosis of primary hypertension ≥2 months and with an education level above primary school. Exclusion criteria were history of neurologic diseases, cardiovascular diseases or peripheral artery disease, psychological disorders, inability to read or write, life expectancy <1 year and receiving renal replacement therapy. All individuals provided written informed consent. The study protocol was approved by the Institutional Ethics Committee of the Second Affiliated Hospital of Nanjing Medical University.

Ambulatory ABP Monitoring

A 24-h ambulatory ABP monitoring was performed using an automatic oscillometric device (SpaceLabs Medical, Redmond, WA, USA), and data were analyzed with manufacturer’s software (SpaceLabs Report Manager System). The recorder was set to take readings at 20-min intervals during the day (6:00 a.m. to 10:00 p.m.), and every 30 min during the night (10:00 p.m. to 6:00 a.m.). All patients were encouraged to carry out their normal daily awake and sleep periods. A recording was considered acceptable if there were at least 20 readings from 8:00 a.m. to 8:00 p.m. and at least 16 readings from 8:00 p.m. to 8:00 a.m. [12].

Data Analysis by ARSER

We further analyzed BP from 8:00 a.m. at 4-h intervals. The circadian rhythmicity was analyzed by ARSER analysis [13] using the R statistics package: “MetaCycle” [14]. ARSER analysis was an algorithm for identifying periodic expression profiles in analyzing circadian microarray data created by Yang and Su [13]. Further information is available on https://bioinformatics.cau.edu.cn/ARSER. BP was considered rhythmic at the significance threshold p < 0.05. The period was fixed at 10–14 h or 22–26 h, which is referred to as 12-h or 24-h period. For cosinor regression analysis, data were fitted to a linear model using the least squared method, minimizing the residual sum of the squares: xi = M + Acos (θi + φ), where M = Mean, φ = Acrophase, θi = Time of sample collection.

Data Collection

General information, including demographic characteristics and medical information, was determined by a research staff member using the combination of patient interviews and medical records. Routine clinical laboratory tests were performed using fasting blood samples. Renal function was calculated as an estimated glomerular filtration rate (eGFR) using the Chronic Kidney Disease Epidemiology (CKD-EPI) equation.

Longitudinal Follow-Up

At the end of the 3-year follow-up period, fasting blood samples were collected. The primary outcome was the change in eGFR. We defined eGFR decline as the falling difference over 1/2 SD between follow-up and baseline.

Statistical Analyses

For descriptive analysis, mean ± SD or median (IQR) and frequency (%) are presented as appropriate. Comparison between the two groups was calculated with the unpaired 2-tailed Student’s t test or chi-squared test. The ARSER analysis was performed using R software. Univariate and multivariate Cox regressions were used to assess the relationship between BP rhythm and eGFR. Potential confounders included age, smoking, alcohol consumption, diabetes mellitus, BMI, albumin levels, administration time of antihypertensive drug time, and hypertension duration. Data were analyzed using SPSS 25.0. A p < 0.05 was considered statistically significant.

Results

Subject Demographics and Baseline Characteristics

Among the 200 subjects (all male) enrolled in the BRAVE-1 study, 12 had insufficient BP records, with <70% valid measurements; consequently, 188 men were included (mean age, 64.9 ± 7.2 years; BMI, 25.4 ± 2.5 kg/m²; Table 1). The median time of first onset of hypertension was 384.5 (183.0–623.5) months. Most of the subjects (85.6%) took their BP medication in the morning. The subjects had an average resting heart rate (HR) of 69.1 ± 11.1 bpm, SBP/diastolic BP (DBP) of 126/76.9 mm Hg, eGFR of 86.6 ± 14.0 mL⁻¹ min⁻¹ 1.73 m⁻², and urinary albumin:creatinine ratio (uACR) of 7.5 (3.8–19.4) mg/g.

Table 1.

Characteristics of the study population at baseline (N = 188)

	Mean ± SD/n (%)
Age, years	64.9±7.2
BMI, kg/m2	25.4±2.5
Current smoker	53 (28.2)
Current alcohol	75 (39.9)
Hypertension time, months	384.5 (183.0–623.5)
Administration of antihypertensive drug time
None	9 (4.8)
Morning	161 (85.6)
Night	4 (2.1)
Both morning and night	14 (7.4)
Diabetes	39 (20.7)
HR, bpm	69.1±11.1
SBP, mm Hg	126.0±12.2
DBP, mm Hg	76.9±8.1
eGFR, mL/min/1.73 m2	86.6±14.0
UACR, mg/g	7.5 (3.8–19.4)
Albumin, g/L	47.0±2.7
Total cholesterol, mmol/L	4.7±0.8
Triglycerides, mmol/L	1.3 (1.0–2.0)
HDL cholesterol, mmol/L	1.3±0.3
LDL cholesterol, mmol/L	2.9±0.8
CRP, mmol/L	0.9 (0.6–1.7)

BMI, body mass index; eGFR, estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; UACR, urinary albumin-to-creatinine ratio.

BP Rhythm

We performed an ARSER analysis to assess the circadian rhythm of BP at 4-h intervals. BP oscillates with two distinct periods, of approximately 12 h and 24 h. For SBP, 38 subjects exhibited a 12-h period and 43 exhibited a 24-h period, while the remaining 107 exhibited no discernible rhythm (Fig. 1a, b). For DBP, 38 (20.2%) exhibited a 12-h period and 36 (19.1%) exhibited a 24-h period, while 114 (60.6%) exhibited arrhythmics. Average SBP was 125.2 ± 12.9 mm Hg in the 12-h group, 125.2 ± 11.0 mm Hg in the 24-h group, and 126.6 ± 12.5 mm Hg in the arrhythmia group (p = 0.749). Average DBP was 77.5 ± 9.3 mm Hg in the 12-h group, 77.5 ± 8.1 mm Hg in the 24-h group, and 76.5 ± 7.8 mm Hg in the arrhythmia group (p = 0.719) (online suppl. Tables 1, 2; for all online suppl. material, see https://doi.org/10.1159/000544992).

Fig. 1.

Circadian patterns of BP. SBP (a) and DBP (b) of the subjects at different times and ARSER analysis of the 12- and 24-h rhythms (p < 0.05).

Follow-Up and Outcomes

Figure 2 presents a flow diagram of the study. During the 3-year follow-up period, 36 subjects (19.1%) were lost to follow-up and 16 (8.5%) died. We also compared the baseline characteristics between participants with follow-up and loss of follow-up. There was no significant difference in age, SBP, DBP, HR, eGFR, hypertension time, urinary albumin-to-creatinine ratio and CRP between the two groups (online suppl. Table 3). It suggested the follow-up data were representative. Thus, we compared the BP and renal function results at baseline and follow-up (Table 2). Relative to baseline, eGFR was lower at follow-up (Baseline: 86.6 ± 14.0; follow-up: 81.0 ± 17.1; p = 0.001) and SBP higher at follow-up (baseline: 129.4 ± 13.5; follow-up: 126.0 ± 12.2; p = 0.020). DBP and uACR did not differ significantly between baseline and follow-up (p > 0.05).

Fig. 2.

Flow-chart of the study.

Table 2.

Renal function and BP of participants at baseline and follow-up

	Baseline (N = 188)	Follow-up (N = 136)	p value
SBP	126.0±12.2	129.4±13.5	0.020
DBP	76.9±8.1	77.1±8.4	0.865
eGFR	86.6±14.0	81.0±17.1	0.001
ACR	7.5 (3.8–19.4)	8.3 (7.0–12.7)	0.059

Association between BP Rhythm and eGFR

For SBP, relative to the arrhythmia group, the 12-h period group exhibited a lower annual reduction in eGFR (arrhythmia group: −1.98 [−3.99 to −0.14] mL⁻¹ min⁻¹ 1.73 m⁻²; 12-h period group: −0.63 [−2.29 to 1.34] mL⁻¹ min⁻¹ 1.73 m⁻²; p = 0.014; Fig. 3), whereas this reduction did not differ significantly between the 24-h period and arrhythmia groups (24-h period group: −2.1 [−3.84 to −0.19] mL⁻¹ min⁻¹ 1.73 m⁻²; arrhythmia group: −1.98 [−3.99 to −0.14] mL⁻¹ min⁻¹ 1.73 m⁻², p = 0.858; Fig. 3a). For DBP, the 12-h period, 24-h period, and arrhythmia groups exhibited similar annual differences in eGFR (12-h period group: −1.4 [−3.43 to 0.62] mL⁻¹ min⁻¹ 1.73 m⁻²; 24-h period group: −2.1 [−3.68 to 0] mL⁻¹ min⁻¹ 1.73 m⁻²; arrhythmia group: −1.5 [−3.67 to −0.17] mL⁻¹ min⁻¹ 1.73 m⁻², p > 0.05) (Fig. 3b).

Fig. 3.

Association between BP rhythm and change in eGFR per year. a For SBP, the annual differences in eGFR in the 12-h period, 24-h period, and arrhythmia groups were −0.63 (−2.29 to 1.34), −2.1 (−3.84 to −0.19), and −1.98 (−3.99 to −0.14), respectively. b For DBP, these annual differences were −1.4 (−3.43 to 0.62), −2.1 (−3.68 to 0), and −1.5 (−3.67 to −0.17), respectively.

To further investigate the impact of SBP rhythmicity on eGFR, we defined a decline in eGFR as a reduction of >1/2 SD (△eGFR ≥7) between follow-up and baseline. For SBP, the 12-h period group exhibited a 70% lower risk of decline in eGFR than the arrhythmia group (HR = 0.344 [0.126–0.938]; p = 0.037; Table 3). After adjusting for confounding variables (including age, smoking, alcohol consumption, diabetes mellitus, BMI, albumin levels, duration of antihypertensive drug administration, hypertension duration), 12-h period SBP remained an independent protective predictor of the decline in eGFR (HR = 0.307 [0.108–0.874]; p = 0.027).

Table 3.

Association between BP rhythm and eGFR decline

	Model 1		Model 2
	β (95% CI)	p value	HR (95% CI)	p value
SBP without rhythm	1.0 (ref)		1.0 (ref)
SBP with 12-h rhythm	0.344 (0.126–0.938)	0.037	0.307 (0.108–0.874)	0.027
SBP with 24-h rhythm	1.375 (0.589–3.211)	0.462	1.547 (0.616–3.833)	0.353

Model 1: unadjusted.

Model 2: adjusted for age, smoking, alcohol consumption, diabetes mellitus,  BMI, albumin levels, administration time of antihypertensive drug, and hypertension duration.

eGFR decline defined as a follow-up score on eGFR that was 1/2 SD or more below the baseline.

Discussion

In this prospective cohort study, we identified distinct patterns of fluctuations in BP, with both 12-h and 24-h periods, in men aged 55 and older with hypertension. Notably, approximately 60% of the subjects exhibited irregular BP rhythms. For SBP, the 12-h period group exhibited lower declines in eGFR than the arrhythmia group, whereas, for DBP, eGFR did not vary significantly between the 12-h period, 24-h period, or arrhythmia groups. For SBP, the 12-h period group had a 70% lower risk of decline in eGFR than the arrhythmia group, whereas the 24-h period in SBP did not show protective effects against a decline in eGFR.

Biological rhythms enable organisms to anticipate and adapt to environmental changes in the light/dark cycle. In addition to the well-characterized 24-h circadian rhythms, organisms exhibit oscillations with a 12-h period, consistent with the periodicity of the ocean tides [15]. Such 12-h rhythms occur regularly and autonomously within cells, and their oscillations can be synchronized by specific external stimuli. The 12-h biological clocks operate independently of the 24-h biological clocks [16].

The mechanisms underlying the regulation of BP rhythm by peripheral clocks remain unclear. Evidence from rodent and human studies suggests that the central circadian systems and peripheral clocks within various physiological systems, including the vasculature, liver, adrenal glands, kidneys, gut microbiota, immune system, and autonomic nervous system, contribute to regulating the circadian rhythm of BP [17]. At the molecular level, cellular-clock control of BP rhythm is mediated by four core clock proteins: brain and muscle ARNT-like 1 (Bmal1), circadian locomotor output cycles protein kaput, cryptochrome (Cry), and period (Per) [18]. Studies targeting these core clock genes in knockout rodent models have advanced our understanding of how circadian mechanisms contribute to BP regulation. In perivascular adipose tissue, Bmal1 regulates resting-phase BP via transcriptional regulation of angiotensinogen [19]. Cry1/Cry2 knockout or Cry-null mice lack circadian oscillations in BP and develop salt-sensitive hypertension [20]. Furthermore, Per1 coordinates sodium handling in the distal nephrons to regulate the 24-h BP rhythm [21, 22].

Shea et al. [23] examined the circadian rhythm of BP in 28 normotensive adults (16 males) for 24 h, finding that BP peaked in the evening at approximately 9:00 p.m. and reached a nadir in the early morning. In a sample of six normotensive, healthy, non-smoking, non-medicated males undergoing a 24-h constant-routine protocol, endogenous circadian rhythm was observed in HR but not in BP [23]. In our study, however, 40% of the hypertensive subjects exhibited 12-h or 24-h oscillations in BP. These discrepancies among studies may be due to differences in sample sizes and in the characteristics of the subjects. The other studies of BP rhythms mentioned here examined young, healthy individuals. A cross-sectional study of hypertensive patients revealed that BP peaked 4 h or 12 h, and dipped 20 h, after awakening [24, 25]. In those studies, the oscillations showed no rhythmicity, owing to the different methods used to assess the rhythm in BP. As demonstrated by our findings, other populations, such as older hypertensive males, may exhibit different circadian rhythms of BP.

The relationship between kidney function and BP is complex. Renal dysfunction can lead to hypertension, which further accelerates renal damage [26]. For older Chinese men, eGFR was observed to decline linearly, by 0.91 (0.86–0.95) mL⁻¹ min⁻¹ 1.73 m⁻², with every year of age [27]. Here, however, subjects with arrhythmic SBP exhibited a faster decline in eGFR than those with 12-h periodic SBP. In a study of 906 hypertensive and CKD patients with 7.8 years of follow-up, non-dipping BP status was associated with an 82% greater risk of kidney disease progression among patients with ambulatory BP at goal [28]. In a study utilizing the JTK_CYCLE algorithm to quantify 24-h BP rhythmic components, subjects without rhythmic BP components had 1.8- and 1.48-times higher risks of end-stage renal disease or composite renal outcome, respectively, than those with rhythmic BP components (although these associations were not statistically significant after being were fully adjusted) [29]. Similarly, we found that the decline in eGFR was not altered in the 24-h rhythmic SBP group, thus highlighting the stronger prognostic value of 12-h rhythmic SBP.

The effects of SBP with 12-h periodicity were notably different from those of SBP with and 24-h periodicity. For SBP, BP at 12:00 a.m. was reduced in the 12-h period group, indicating that this group exhibited a greater nocturnal reduction in BP than the 24-h period SBP group. While further research is necessary to validate the pathophysiology, we hypothesize that the protective effects of this periodicity may be linked to “dipping,” which exhibits complex pathophysiology and involves factors such as sleep-activity or the sleep-wake cycle, autonomic nervous system function, water and sodium regulation, and gut microbiota [30].

Interestingly, the cyclical periods of the BP rhythms are interchangeable. Katinas et al. [31] continued monitoring the BP of GSK, finding that an abrupt increase in the dosage of antihypertensive drugs led to suppression of the 24-h BP rhythm, which was then replaced by a 12-h rhythm. This observation has prompted us to identify different classes of antihypertensive drugs and determine which class is beneficial in correcting the loss of BP rhythm.

This study has several limitations. First, the population was limited to older hypertensive males. Larger clinical trials across multiple ethnic and age groups, and including females, are required to elucidate variation in circadian rhythm within and between individuals. Second, we obtained only a single-point measurement of ambulatory BP at enrollment and only one measurement of BP during follow-up. It is unclear whether BP fluctuated over time in these subjects; however, we confirmed that none of the subjects altered their medication schedule. Finally, the follow-up period spanned the COVID-19 pandemic, the mortality rate of the subject exceeded expectations, and we were unable to effectively determine the cause of death.

We evaluated the circadian rhythmicity of BP and its associations with adverse renal outcomes in hypertensive males aged 55 and older. Beyond the traditional risk factors for a decline in eGFR, we found that a 12-h rhythmic SBP serves as an independent protective predictor of a decline in eGFR. Monitoring 24-h BP rhythms using ambulatory BP monitoring and recognizing the importance of restoring a normal BP rhythm should be considered crucial aspects of BP management. Restoring the 12-h circadian rhythm of BP represents a potential therapeutic approach for addressing hypertension.

Acknowledgment

We would like to thank Editage (www.editage.cn) for English.

Statement of Ethics

This study protocol was reviewed and approved by the Institutional Ethics Committee of The Second Affiliated Hospital of Nanjing Medical University, Approval No. ([2018]KY081). All individuals provided written informed consent.

Conflict of Interest Statement

Author Chunsun Dai was a member of the journal’s Editorial Board at the time of submission. The others do not have any conflicts of interest to declare.

Funding Sources

This works was supported by National Natural Science Foundation of China Grants (81870502) and Jiangsu Province’s Key Provincial Talents Program (Qnrc2016669) to Lei Jiang, and Natural Science Foundation of Jiangsu Province (BK20241161) and School Fund of Nanjing Medical University (NMUB20230003) to Lulu Wang.

Author Contributions

Lei Jiang and Chunsun Dai were involved in the research idea, study design, and supervision or mentorship. Lulu Wang, Han Tian, Xinxin Xu, Xinyan Gu, Liu Li, Hui Zheng, and Jie Xu were responsible for data acquisition. Lulu Wang and Han Tian were involved in data analysis and interpretation. All authors read and approved the final manuscript.

Funding Statement

This works was supported by National Natural Science Foundation of China Grants (81870502) and Jiangsu Province’s Key Provincial Talents Program (Qnrc2016669) to Lei Jiang, and Natural Science Foundation of Jiangsu Province (BK20241161) and School Fund of Nanjing Medical University (NMUB20230003) to Lulu Wang.

Data Availability Statement

The data that support the findings of this study are not publicly available due to the privacy of research participants but are available from the corresponding author (Lei Jiang) upon reasonable request.