Sleep variability, sleep irregularity and nighttime blood pressure dipping

Abstract

Background:

Circadian rhythm regulates many important biological functions in humans. The goal of this study is to explore the impact of day-to-day deviations in sleep-wake cycle on nighttime blood pressure (BP) dipping and further examine whether the ethnic difference in day-to-day deviations in sleep pattern can explain the ethnic difference in nighttime BP dipping.

Methods:

24-h ambulatory blood pressure monitoring and 7-day accelerometer data were obtained from 365 adult participants (age range 18.7-50.1 years; 52.6% African Americans and 47.3% European Americans; 64.1% females). Systolic BP dipping level was used to represent nighttime BP dipping. The standard deviation (SD) of sleep duration was calculated as the index of sleep variability and the SD of sleep midpoint was calculated as the index of sleep irregularity.

Results:

A one-hour increase in the SD of sleep midpoint was associated with a 1.16% decrease in nighttime BP dipping (P<0.001). A one-hour increase in the SD of sleep duration was associated with a 1.39% decrease in nighttime BP dipping (P=0.017). The ethnic difference in the SD of sleep midpoint can explain 29.2% of the ethnicity difference in BP dipping (P=0.008).

Conclusions:

Sleep variability and sleep irregularity are associated with blunted BP dipping in the general population. In addition, data from the present investigation also demonstrate that the ethnic difference in sleep irregularity could partly explain the ethnic difference in BP dipping, an important finding that may help reduce the health disparity between African Americans and European Americans.

Graphical Abstract

INTRODUCTION

Most of the natural processes in humans follow a circadian rhythm (i.e. 24-hour cycle), which is critical for optimal physiological function ^1,2. Biological clocks are genetically determined and regulated by the master clock that is located in the suprachiasmatic nuclei in the hypothalamus of our brain ³. It is well known that blood pressure (BP) demonstrates circadian oscillations ^4,5. The circadian modulation of sympathetic activity during the nighttime is paralleled by an elevation in vagal tone and substantial reductions in biological activities, such as heart rate and cardiac output ^6,7. This intricate interplay of the autonomic system results in a reduction of BP during the resting period ^8,9. Traditionally, the physiological decrease in BP at night, commonly referred to as the "dip," typically falls within the range of 10% to 20% when compared to daytime BP levels, which is considered a protective factor for cardiovascular health. In fact, the non-dipping phenomenon has been linked with increased risk of target organ damage and cardiovascular event ¹⁰.

Sleep represents a multifaceted health behavior ¹¹. In the modern society, environments and lifestyles marked by increased light exposure, nocturnal activities, and pervasive use of electronic media and mobile devices not only lead to insufficient sleep, but also profoundly disrupt the innate rhythms of sleep behaviors. These disruptions in sleep patterns, particularly concerning the duration (i.e. sleep variability) and timing of sleep (i.e. sleep irregularity), have the potential to perturb the natural circadian rhythms that regulate various biological processes ¹². Previous studies ^13,14 using cross-sectional and prospective analyses have found that greater intra-individual day-to-day deviations in sleep patterns were associated with higher prevalence and incidence of multiple disease conditions, including the risk of hypertension. Previous research in laboratory settings ¹⁵ have also found that a rapid 12-hour inversion of the sleep-wake cycle increased 24-hour BP, primarily during the sleep period, leading to reduced BP dipping. However, whether day-to-day deviations in sleep-wake behavior are associated with the BP dipping phenomenon in the population has not been studied. In this study, we aim to fill in this knowledge gap by evaluating the relationship of sleep variability and sleep irregularity with BP dipping in a real-life setting. Based on the previous literature documenting that African Americans (AAs) have larger day-to-day deviations in sleep pattern ^16-18 as well as lower nighttime BP dipping than European Americans (EUs) ^19-23, we further aim to examine whether the ethnic difference in sleep pattern can at least partially explain the ethnic difference in BP dipping in our bi-ethnic sample of adults.

METHODS

Data, Materials, and Code Disclosure Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Participants

The present study included 365 participants (age range 18.7-50.1 years; 52.6% AAs; 64.1% females) from two long-standing cohort studies at the Georgia Prevention Institute: the Georgia Cardiovascular Twin Study (n=243, 80 twin pairs and 83 singletons), and the Georgia Stress and Heart Study (n=122). Participants were included if they had valid 24-h ambulatory BP measurements and 7 days of accelerometer data that were obtained during their study visit conducted between May 2019 - April 2022. The flowchart of participants selected for inclusion in this analysis is illustrated in Figure 1. The Georgia Stress and Heart Study was established in 1989 to study the development of cardiovascular risk factors and the Georgia Cardiovascular Twin Study was established in 1996 to explore the change in relative influence of genetic and environmental factors on the development of cardiovascular risk factors. Study design, selection criteria and the criteria to classify participants as AAs or EUs have been described previously ^19,24,25. Of the 365 participants, 40 were taking anti-hypertensive medications. Sensitivity tests were conducted by excluding these participants from the analyses. The institutional review board at the Medical College of Georgia, Augusta University gave approval for the studies. Each participant gave written consent prior to any involvement, in accordance with the institutional guidelines.

Figure 1.

Flowchart of the selection of the participants for the current study.

During the study visit, anthropometric measurements were obtained and participants’ height and weight were measured using a stadiometer and Healthometer medical scale. Body mass index (BMI) was calculated as weight/height² (kg/m²).

Accelerometer recording and data preprocessing

Participants were asked to wear the accelerometers (ActiGraph Model GT3X+, ActiGraph of Pensacola, FL) on their non-dominant wrist continuously for 7 days, while continuing with their normal activities. At the end of the 7-day period, participants were instructed to return the accelerometer to the Georgia Prevention Institute. All measures were derived by processing raw accelerometer data (.gt3x). Briefly, the ActLife software was used to convert the raw data to .csv files, which were processed with the open source R package GGIR ²⁶ to calculate accelerometer non-wear time, and estimate the sleep period time (SPT) window and episodes within SPT. Accelerometer recordings with less than 3 days of accelerometer data were excluded from further analysis.

Sleep parameters

In GGIR R package, the SPT-window was defined from the median values of the absolute change in estimated z-angle (i.e. the dorsal-ventral direction in the wrist of the anatomical position) across 5-min rolling windows within a 24-h period. The 10th percentile of the output was used to construct an individual’s threshold to distinguish periods with movement from non-movement. Bouts of inactivity lasting ≥ 30 min were defined as inactivity bouts. Inactivity bouts that had a gap < 60 min were combined to represent the inactivity blocks. The SPT-window was defined as the longest inactivity block, with sleep onset as the start of the block and waking time as the end of the block. Sleep episodes within the SPT-window were defined as periods of the z-axis angle change less than 5° for at least 5 minutes. The number of sleep episodes was defined as the number of sleep episodes within the SPT-window. Sleep duration in a SPT-window was calculated as the sum of all sleep episodes²⁶. As suggested by the UK biobank study ²⁷, individuals with an average number of sleep episodes ≤ 5 or ≥ 30 as well as individuals with an average sleep duration < 3 h or > 12 h were excluded from the analyses. Sleep midpoint was calculated as the midpoint between the start and end of the SPT-window. The standard deviation (SD) of sleep duration was calculated as the index of sleep variability and the SD of sleep midpoint was calculated as the index of sleep irregularity.

Due to the U-shaped relationship between sleep midpoint and the risk of diseases, sleep midpoint was examined in terms of deviation from the sample mean (3:21 AM). “Advanced” participants were defined as having a sleep midpoint of < −0.5 SD from the mean (< 2:30 AM) and “Delayed” participants were defined as having a sleep midpoint of > 0.5 SD from the mean (> 4:12 AM).

Ambulatory BP measurement

Participants underwent ambulatory BP monitoring for 24 hours. The 24 hour ambulatory BP assessment methodology have been used by our group for the past 20 years and described previously in detail ¹⁹. Briefly, an ambulatory BP monitor (model 90217; SpaceLabs, Redmond, WA) was fitted to the non-dominant arm. Measurements were obtained every 20 minutes during the daytime (08:00-22:00) and every 30 minutes during the night (22:00-08:00). The daytime period was defined from 8:00 AM until 10:00 PM. The nighttime period was defined from 12:00 AM to 6:00 AM. The adequacy of recordings was defined according to the European Society of Hypertension Working Group on Blood Pressure Monitoring as ≥ 14 readings over the 14hrs designated as daytime, and ≥ 6 readings over the 6hrs designated as nighttime. Systolic blood pressure (SBP) dipping levels were used to represent nocturnal BP dipping and calculated as (Daytime SBP – Nighttime SBP) ×100/Daytime SBP. In addition to the daytime and nighttime periods that were defined by the fixed time window, BP during the wakeup period and BP during the sleep period according to the diary information collected in the Georgia Cardiovascular Twin cohort were also calculated. Both the daytime SBP and nighttime SBP defined by the fixed time window were highly correlated with the wakeup period SBP (r= 0.99) and sleep period SBP (r=0.97) (Supplementary figure S1). The Bland-Altman plot (Supplementary figure S2) also demonstrated the similarity of these two approaches of defining day and night with a mean difference of 0.282 mm Hg for daytime SBP and a mean difference of 0.002 mm Hg for nighttime SBP. Overall, this indicates the daytime and nighttime period defined by the fixed window period is appropriate for the current sample.

Statistical analysis

All analyses were performed using STATA software (V16). Generalized estimating equations (GEE) were used to test whether ethnic differences exist in sleep variability, sleep irregularity and SBP dipping with age, sex and BMI as covariates. GEE is a multiple regression technique that allows for non-independence of twin or family data yielding unbiased standard errors and p-values. GEE was also used to test the associations of sleep variability and sleep irregularity with SBP dipping with age, sex, ethnicity, and BMI as covariates. We further included sleep duration and sleep midpoint as covariates to investigate the independent effect of day-to-day deviations in sleep pattern on SBP dipping. Post hoc power calculation was conducted to check the power of current sample size. According to the observation of a correlation of −0.15 between sleep variability and SBP dipping and a correlation of −0.23 between sleep irregularity and SBP dipping in this study, the current sample size has 82% power to detect the association of sleep variability and 99% power to detect the association of sleep irregularity with SBP dipping, respectively.

We performed mediation analyses ²⁸ to test whether the ethnic differences in day-to-day deviations in sleep can mediate the ethnic differences in SBP dipping. Mediation was considered to present when the ethnic differences in SBP dipping decreased upon the addition of the parameters of day-to-day deviations in sleep to the model. We used bootstrapping ²⁹ to test for mediation significance. The covariates in all mediation models included age, sex, and BMI.

RESULTS

General characteristics of the participants are presented by ethnicity in Table 1. After adjustment of age, sex and BMI, AAs showed significantly lower levels of nighttime SBP dipping (P=0.003) in comparison with EUs. There was no significant difference between AAs and EUs in the SD of sleep duration, however, AAs displayed a larger SD of sleep midpoint than EUs (P=0.029). In addition, AAs had a significantly shorter sleep duration (P<0.001) and were more likely to be delayed sleepers compared with EUs (P=0.009).

Table 1.

Participant Characteristics

Participant characteristics	EA (n=173)	AA (n=192)	P
Age, years	36.4 (5.8)	35.3 (6.0)	0.001
Female, N (%)	105 (60.7)	129 (67.2)	0.129
BMI, kg/m2	28.1 (6.4)	31.1 (7.9)	<0.001
Nocturnal SBP Dipping*, %	12.4 (7.1)	10.3 (7.1)	0.003
Relative amplitude*	0.73 (0.12)	0.67 (0.13)	<0.001
Sleep parameters
SD of sleep midpoint*, hours	1.1 (1.3)	1.3 (1.2)	0.029
SD of sleep duration*, hours	1.1 (0.70)	1.2 (0.54)	0.166
Sleep midpoint+			0.009
Intermediate, N (%)	87 (50.3)	95 (49.5)
Advanced, N (%)	54 (31.2)	39 (20.3)
Delayed, N (%)	32 (18.5)	58 (30.2)
Sleep duration*, hours	6.6 (1.1)	5.9 (1.1)	<0.001

^*Age, sex and BMI adjusted P values

⁺Chi square analysis

The associations of sleep variability and sleep irregularity and SBP dipping are presented in Table 2. A one-hour increase in the SD of sleep midpoint was associated with a 1.16% decrease in SBP dipping (P<0.001). In addition, a one-hour increase in the SD of sleep duration was associated with a 1.39% decrease in SBP dipping (P=0.017). After adjustment of their corresponding mean values, i.e. sleep midpoint or sleep duration, respectively in model 2, the significant results remained the same. These findings were essentially unchanged when participants on anti-hypertensive medication were excluded from the analysis.

Table 2.

Multivariable regression models examining the associations of sleep variability and sleep irregularity with BP dipping

	SBP Dipping
	Model 1		Model 2
	β	P	β	P
SD of sleep midpoint	−1.16	<0.001	−1.51	<0.001
SD of sleep duration	−1.39	0.017	−1.17	0.048

Model 1 adjust age, sex, race, and BMI as covariates

Model 2 further adjust the mean value of sleep duration for SD of sleep duration and sleep midpoint for SD of sleep midpoint

We further tested whether the ethnic difference in the SD of sleep midpoint can explain the ethnic differences in SBP dipping. As shown in Figure 2, the ethnic differences in SBP dipping decrease with the addition of the SD of sleep midpoint to the model (β changes from −1.96 to −1.40). Bootstrapping tests for significance revealed that the mediation effects were significant (P=0.008). Accordingly, the ethnic difference in the SD of sleep midpoint can explain 29.2% of the ethnicity difference in SBP dipping.

Figure 2.

Mediation models illustrating sleep irregularity mediating the effect of ethnicity on BP dipping.

DISCUSSION

This study examined whether day-to-day deviations in sleep patterns were associated with BP dipping phenomenon in a real-life setting. Findings from the present investigation indicate that larger sleep variability and larger sleep irregularity were associated with decreased levels of nighttime BP dipping. Furthermore, for the first time, the present investigation demonstrates that the ethnic differences in sleep irregularity can at least partially explain the ethnic differences observed in BP dipping.

The existing body of research exploring the potential relationship between day-to-day deviations in sleep patterns, as objectively assessed through wrist-worn actigraphy, and BP is still limited in scope, primarily concentrating on office BP levels. For example, Sabra et al ³⁰. demonstrated a significant association between disrupted sleep-wake timing, quantified through a nonparametric metric known as interdaily stability, and increased levels of both SBP and DBP. Additionally, Alberto et al.³¹ have provided evidence that a 10% increase in the sleep fragmentation index correlated with a 5.2% higher prevalence of hypertension, while frequent napping was associated with an 11.6% increased prevalence of hypertension. The insight into the underlying mechanisms of circadian disruption caused by changes in sleep patterns is offered by biological experiments. Laboratory-based studies on animals have indicated that the expression levels of certain clock genes were modulated by sleep-wake cycles ^32,33. Under circumstances of irregular sleep-wake cycles, this regulation can result in deviations of internal clock gene expression patterns from their external oscillatory performance, leading to the disruption of circadian rhythms. Archer et al. ³⁴ found that misaligned sleep affected the circadian regulation of the transcriptome, resulting in a decrease in the number of circadian transcripts. In clinical research conducted in the laboratory, Morris et al. ¹⁵ demonstrated that a rapid 12-hour inversion of the sleep-wake cycle increased 24-hour BP, primarily during the sleep period, leading to reduced BP dipping. It is important to note that the current study provides the first evidence that both sleep variability and sleep irregularity are associated with reduced nighttime BP dipping in a population study. In addition, the present data further extends the laboratory findings of Morris et al. ¹⁵ and provides evidence in a real-life setting.

Ethnic differences in BP dipping ^19-21 and parameters of day-to-day deviations in sleep ^16-18 have been identified, with AAs more likely to exhibit a blunted BP dipping and larger deviations in sleep parameters. The ethnic differences observed in the present study are in line with these previous findings. Additionally, our current results underscore the potential role of sleep irregularity in contributing to the ethnic difference in BP dipping. Specifically, our mediation analysis indicates that as much as 29.2% of the ethnic difference in BP dipping could be explained by the SD of sleep midpoint. Our findings provide important evidence of the likely involvement of sleep irregularity in explaining the blunted BP dipping observed in AAs, and indicate that modification of sleep-wake patterns may be able to reduce the health disparity between AAs and EUs. Future longitudinal studies and randomized clinical trials are needed to investigate whether healthy sleep patterns, specifically by maintaining sleep regularity, would restore a healthy circadian rhythm in BP and improve cardiovascular disease risk.

Our study has several experimental considerations that need to be discussed prior to interpretation. First, as the Georgia Cardiovascular Twin Study and the Georgia Stress and Heart Study are comprised of young and middle-age adults, the generalizability of these results to other age-related populations remains to be determined. Second, the quantification of BP dipping and sleep parameters was based on a one-time ABP measurement or actigraphy data. Continued follow-up of these cohort studies with multiple days at multiple time points will provide more reliable evidence. Last, further studies with large sample sizes involving multiethnic groups are warranted.

In summary, the present investigation demonstrates that sleep variability and sleep irregularity are associated with a blunted BP dipping in young and middle-aged adults. In addition, findings also suggest that the ethnic difference in sleep irregularity could partially explain the established ethnic difference in BP dipping.

Perspective: Exploring Sleep Patterns, Circadian Rhythms and Cardiovascular Health

This study's investigation into the implications of day-to-day sleep-wake cycle deviations on nocturnal blood pressure (BP) dipping adds a unique dimension to our comprehension of the intricate interplay among circadian rhythms, sleep patterns, and cardiovascular health within population-based research. The revealed associations between sleep variability and irregularity and reduced BP dipping emphasize the essentiality of maintaining consistent sleep routines for optimal cardiovascular well-being. This underscores the importance of addressing sleep patterns as a modifiable factor to optimize nocturnal BP regulation, ultimately impacting overall cardiovascular health, especially with the consideration of the increasing prominence of wearable technology, which can moniter sleep patterns in real time. Moreover, the study's exploration of ethnic differences in sleep patterns and their role in explaining ethnic disparities in BP dipping encourages us to consider the potential effectiveness of personalized interventions focused on improving sleep consistency, with the goal of reducing health disparities among diverse ethnic groups.

Looking ahead, this study sets the stage for future investigations that could explore causality, mechanisms and potential interventions to unravel the dynamic interplay between sleep patterns, circadian rhythm and cardiovascular outcomes. As the field of circadian biology and cardiovascular health continues to evolve, these insights have the potential to drive transformative changes in healthcare practices, public health policies, and individual well-being, paving the way for a healthier future.

NOVELTY AND RELEVANCE.

What Is New?

• Uncover the significant role of sleep variability and irregularity in affecting BP dipping in the general population.

• Demonstrate that the ethnic difference in sleep irregularity could partly explain the ethnic disparities in BP dipping, which may prompt us to think about personalized interventions to reduce the health disparity between African Americans and European Americans.

What Is Relevant?

• This study's findings contribute to our understanding of how day-to-day deviations in sleep-wake cycle influence cardiovascular health.

• The research directly addresses BP dipping, a vital indicator of cardiovascular health, and reflects the body's ability to rest and recover. A healthy BP dipping pattern indicates that the cardiovascular system is functioning optimally during sleep, allowing the heart to relax and reduce stress on blood vessels. Disrupted or blunted BP dipping, is associated with a higher risk of developing cardiovascular diseases, stroke, and other related complications.

• Sleep patterns are intricately intertwined with the circadian rhythm, a pivotal regulator of blood pressure dipping. By linking day-to-day sleep deviations to BP dipping, the study bridges a gap in our knowledge of the intricate interplay among circadian rhythms, sleep patterns, and cardiovascular health within population-based research.

Clinical/Pathophysiological Implications?

• The findings emphasize the importance of consistent sleep routines for cardiovascular well-being.

• Insights into the connection between sleep irregularity/variability and BP dipping could inform personalized interventions to manage hypertension.

• Addressing sleep irregularity could potentially mitigate ethnic disparities in cardiovascular health, leading to more targeted health interventions.

Abbreviations and acronyms:

BP

blood pressure

SD

standard deviation

AA

African American

EU

European Americans

SPT

Sleep period time

SBP

Systolic blood pressure

GEE

Generalized estimating equations