Sex differences in blood pressure responsiveness to spontaneous K-complexes during stage II sleep

Abstract

K-complexes are a key marker of nonrapid eye movement sleep, specifically during stages II sleep. Recent evidence suggests the heart rate responses to a K-complexes may differ between men and women. The purpose of this study was to compare beat-to-beat blood pressure responses to K-complexes in men and women. We hypothesized that the pressor response following a spontaneous K-complex would be augmented in men compared with women. Ten men [age: 23 ± 2 yr, body mass index (BMI): 28 ± 4 kg/m²] and ten women (age: 23 ± 5 yr, BMI: 25 ± 4 kg/m²) were equipped with overnight finger plethysmography and standard 10-lead polysomnography. Hemodynamic responses to a spontaneous K-complex during stable stage II sleep were quantified for 10 consecutive cardiac cycles, and measurements included systolic arterial pressure (SAP), diastolic arterial pressure (DAP), and heart rate. K-complex elicited greater pressor responses in men when blood pressures were expressed as SAP (cardiac cycle × sex: P = 0.007) and DAP (cardiac cycle × sex: P = 0.004). Heart rate trended to be different between men and women (cardiac cycle × sex: P = 0.078). These findings suggest a divergent pressor response between men and women following a spontaneous K-complex during normal stage II sleep. These findings could contribute to sex-specific differences in cardiovascular risk that exist between men and women.

NEW & NOTEWORTHY K-complexes during stage II sleep have been shown to elicit acute increases in blood pressure and heart rate, but the role of sex (i.e., male vs. female) in this response is unclear. In the present study, we demonstrate that the pressor response following spontaneous K-complexes were augmented in men compared to age-matched women. The augmented blood pressure reactivity to spontaneous K-complexes during stage II sleep in men advance the field of cardiovascular sex differences, with implications for nocturnal blood pressure control.

INTRODUCTION

The K-complex is a common electroencephalographic (EEG) occurrence during nonrapid eye movement (NREM) sleep (1–3). These characteristic waveforms are present in both stage II (i.e., light sleep) and stage III (i.e., slow wave, delta, deep) sleep (4, 5). As common as K-complexes may be during normal sleep, the generation of these waveforms remains highly debated. Theories include that K-complexes arise from large coordinated firings of cortical neurons (6) and deeper origins within thalamic neurons to generate the slow frequency (i.e., ∼1 Hz) within the brain (7). Contrasting theories posit that the purpose of a K-complex (i.e., spontaneous in nature) is to initiate arousal from sleep (8), promote sleep following an external stimulus like acoustic stimuli (9), and even promote memory consolidation during sleep (10).

The presence of K-complexes during sleep within the central nervous system also appears to be associated with changes to the peripheral nervous system as quantified by postganglionic sympathetic nerve activity. In the early 1990s, Hornyak et al. (11), Somers et al. (12), and Takeuchi et al. (13) all reported significant associations between K-complexes and bursts of muscle sympathetic nerve activity (MSNA), where K-complexes directly preceded a significant proportion of MSNA bursts. Tank et al. (14) later quantified a pressor response following bursts of MSNA that were associated with K-complexes, which produced acute increases in both heart rate (HR) and arterial blood pressure. These prior studies (11–14) examining the relationships between K-complexes, blood pressure, and/or MSNA were not powered to probe for sex differences. This is relevant for several reasons. First, consistent and healthy reductions of nocturnal blood pressure during sleep are associated with reduced cardiovascular morbidity and mortality (15, 16). As such, factors such as K-complexes that could potentially impact nocturnal blood pressure dipping are important to understand. Second, there are studies that report a blunted nocturnal blood pressure dip in males compared with females, particularly when traits such as depression and anger are considered (17–20). Finally, the majority of research studies examining sleep architecture and nocturnal cardiovascular control tend to focus on rapid eye movement (REM) and slow wave sleep, yet it is important to note that sleep stages I and II tend to make up 50% or greater of habitual nocturnal sleep. Accordingly, examining the impact of sex on cardiovascular responses to the characteristic K-complex of stage II sleep has clinical relevance.

Current lines of research on stage II K-complexes and cardiovascular control typically examine either spontaneous and/or experimentally sound-evoked K-complexes. Importantly, spontaneous and sound-evoked K-complexes appear to elicit similar neurovascular and pressor effects (13, 21). de Zambotti (22) recently reported that cardiac autonomic responses to sound-evoked K-complexes are different between adolescent males and females. Specifically, peak HR responses to a sound-evoked K-complex in adolescent males were augmented compared with age-matched females. These sex differences were determined from electrocardiography in adolescents, and it remains unknown if the pressor response to a spontaneous K-complex differs between adult men and women. Accordingly, the purpose of the present study was to determine the impact of sex (i.e., male vs. female) on the pressor response associated with spontaneous K-complexes in stage II sleep. We hypothesized that the blood pressure response to a spontaneous K-complex would be greater in men compared with women.

METHODS

Participants

Twenty young, healthy adults (10 men and 10 women) were recruited by study investigators from the Michigan Technological University campus and surrounding community. Inclusion criteria for each participant were ages 21–40 yr and body mass index (BMI) 18.5–35.0 kg/m². Participants were excluded if they had a history of cardiovascular disease, autonomic dysfunction, asthma, diabetes, or had a prescription to any cardiovascular or antihypertensive medications. All participants were nonsmokers. Participants abstained from alcohol, exercise, and caffeine consumption for a minimum of 12 h before laboratory testing. Women were not pregnant or breastfeeding and were not prescribed any form of birth control medication, intrauterine device, or hormone replacement therapy for at least 6 mo. Eligible women also reported a regular menstrual cycle (i.e., 25–32 days). All participants were screened for moderate to severe obstructive sleep apnea by a board-certified sleep physician (C. A. Smoot) using an at-home ApneaLink device (ResMed, San Diego, CA). The Michigan Technological University Institutional Review Board approved all testing procedures. Laboratory investigators explained all procedures and obtained written informed consent from all participants, and the study conformed to the guidelines contained within the Declaration of Helsinki.

Experimental Design

Each participant arrived to the Michigan Technological University Sleep Research Laboratory at 9:00 PM. Participants were instrumented for a standard polysomnography sleep study with 10–20 electrode placement for acquisition of sleep EEG, respiratory effort, breathing rate, and limb movements. Dual cuff finger plethysmography (Finapres NOVA; Amsterdam, The Netherlands) was instrumented at 10:30 PM and calibrated with a brachial blood pressure via an automated sphygmomanometer (Omron HEM-907XL; Kyoto, Japan). Lights out occurred at 11:00 PM, and 8 h of sleep acquisition was allotted.

Isolated K-complexes were identified in stable portions of stage II sleep. Three isolated K-complexes were selected for blood pressure analysis when 1) no apneic events were present, 2) free of any limb movements, 3) absence of any scorable arousals, and 4) at least 30 s from previous K-complex. Systolic arterial pressure (SAP), diastolic arterial pressure (DAP), and HR were quantified during the cardiac cycle of the K-complex (i.e., baseline) and the 10 succeeding cardiac cycles.

Measurements

Actigraphy.

Objective overnight sleep variables included total sleep time (TST), sleep efficiency (ratio of TST to total time in bed), time spent awake in bed after falling asleep, and awakenings. The default medium sleep-wake threshold of 40 activity counts was selected (23, 24). In addition, the sleep interval detection algorithm was based on 10 immobile minutes to determine sleep onset and sleep termination. Rest intervals were determined when the light intensity fell below 1 lux, and the activity level was zero for five or more epochs. The actiwatch records activity counts each second (sampling rate: 32 Hz) with the mean of each second summed for each 30-s epoch. Data were averaged over the average recording period (∼10 days) for each variable. All actigraphy recordings and software analyses were confirmed by a certified sleep physician.

Sleep acquisition.

A standard polysomnography sleep study (Natus Medical; Middleton, WI) was performed on each participant. Sleep EEG was recorded and scored via 10–20 electrode placement with two frontal, central, and occipital leads referenced electrodes placed at the mastoid process on the opposite side of the head. Electrooculography and electromyography (EMG) were recorded via two electrodes near the eyes and three electrodes placed on the chin, respectively. Thorax and abdomen piezoelectric effort belts were used to monitor respiratory effort, and respiratory flow was also measured using a nasal cannula. Blood oxygen saturation was recorded via pulse oximetry to monitor potential desaturations associated with apneic events. Leg movements were measured via EMG. Sleep staging, apneic/respiratory events, limb movements, and arousals were defined and scored according to the American Association of Sleep Medicine. All sleep studies were reviewed by a board-certified sleep physician (C. A. Smoot).

Blood pressure and HR.

Arterial blood pressures were obtained in the supine position for calibration purposes. Beat-to-beat arterial blood pressure was continuously recorded throughout the entire 8-h sleep study using a Finapres NOVA system. Brachial blood pressures were used to calibrate the system before lights out. A finger cuff was placed on the middle and ring finger. Blood pressure was collected for a 30-min period before changing finger cuffs to minimize potential discomfort. Arterial blood pressures are expressed as SAP and DAP. HR was recorded continuously via a two-lead electrocardiogram.

Data Analysis

K-complex and blood pressure analysis.

Data were imported and analyzed in the WinCPRS software program (Absolute Aliens, Turku, Finland). R-waves were marked in time series, and blood pressure from respective cardiac cycle were calculated. K-complex waveforms were selected by one trained investigator (I. M. Greenlund) during stage two sleep who was blind to participant’s sex and blood pressure to avoid bias. The selection was confirmed by our sleep physician, who was also blind to each participant’s sex and blood pressure. Each K-complex deflection was greater >100 µV and lasted at least 0.5 s. SAP, DAP, and HR were quantified at cardiac cycle 0 (CC0) through cardiac cycle 10 (CC10) after the identified K-complex. Figure 1 depicts representative recordings for one male and one female participant. A description of K-complex distribution between men and women across the night is detailed in results.

Figure 1.

Representative recordings of 1 male and 1 female pressor response to spontaneous, isolated K-complexes during stage II sleep. The blood pressure increase was augmented in men when compared with women. An electrocardiogram (ECG), beat-to-beat blood pressure waveform, and sleep electroencephalogram (EEG) are represented for each participant.

Statistical Analysis

All data were analyzed statistically using commercial software (SPSS 25.0 IBM; Armonk, NY). Independent samples t tests were used to compare subject demographic data. Normality tests were conducted on the SAP, DAP, and HR response to a K-complex at each cardiac cycle via assessment of skewness. An intraclass correlation coefficient and Cronbach’s alpha value were calculated for SAP, DAP, and HR response patterns (i.e., CC0 – CC10) to each of the three K-complexes identified per subject (i.e., within subject reliability). Repeated-measures ANOVA with sex (men vs. women) as the between factor and cardiac cycle (i.e., CC0-CC10) as the within factor was used to compare hemodynamic responses. Independent samples post hoc t tests were performed when significant cardiac cycle × sex interactions were detected to examine sex differences across cardiac cycles (i.e., Men CC1 vs. Women CC1, etc.). All data are presented as means ± SD.

RESULTS

Table 1 compares general participant characteristics between men and women. Age and BMI were similar between groups. SAP was significantly higher in men (P = 0.002), while DAP and HR were similar between sexes. Total sleep time and sleep efficiency were not different between men and women (Table 2).

Table 1.

Participant characteristics: men vs. women

Variable	Men (n = 10)	Women (n = 10)	P Value
Age, yr	23 ± 2	23 ± 5	0.809
Body mass index, kg/m2	28 ± 4	25 ± 4	0.191
Systolic arterial pressure, mmHg	120 ± 8	109 ± 5*	0.002
Diastolic arterial pressure, mmHg	64 ± 6	61 ± 8	0.428
Heart rate, beats/min	68 ± 11	73 ± 11	0.354
Total sleep time, min	406 ± 50	388 ± 46	0.423
Sleep efficiency, %	85 ± 11	81 ± 10	0.415
Apnea hypopnea index, events/min	1.8 ± 2.3	0.4 ± 0.5	0.090

Results are means ± SD.

Table 2.

Objective actigraphic monitored sleep: men vs. women

Variable	Men (n = 10)	Women (n = 8)	P Value
Total sleep time, min	398 ± 36	421 ± 43	0.245
Sleep efficiency, %	83 ± 4	80 ± 7	0.288
Wake after sleep onset, min	34 ± 10	46 ± 21	0.131
Number of awakenings	32 ± 7	38 ± 10	0.169

Results are means ± SD.

The Cronbach’s alphas for SAP and DAP response to K-complexes demonstrated good and excellent within-subject reliability, respectively (α = 0.896 and 0.917). The HR response to a spontaneous K-complex had moderate reliability (α = 0.747). The within-subject reliabilities for SAP, DAP, and HR responses to K-complexes did not differ between men and women.

A sensitivity analysis was performed to compute the required effect size to determine whether the present data can be attributed to a true effect for the outcome variables of SAP, DAP, and HR. An alpha (α) and power (1 – β) of 0.05 and 0.8 were selected, respectively, for a 2 × 11 repeated-measures design at n = 20. Correlation between representative measures was 0.8, due to high correlation between blood pressure and HR on succeeding cardiac cycles. The Greenhouse-Geisser nonsphericity correction ε was 0.251 for SAP, 0.313 for DAP, and 0.458 for HR. Computed effect size for SAP was f = 0.201, DAP was f = 0.186, and HR was f = 0.162.

Each 8-h sleep opportunity was broken into thirds (i.e., 11:00 PM to 1:40 AM vs. 1:40 to 4:20 AM vs. 4:20 to 7:00 AM) to determine if our random K-complex selections were similar in men and women. K-complex distributions were similar between men and women during the first (13% vs. 20%), second (47% vs. 56%), and final (40% vs. 24%) thirds of the night, respectively.

Figure 2 depicts the K-complex pressor response during stable stage II sleep. Significant cardiac cycle × sex interactions were observed in both SAP [F(2.515,45.269) = 4.999, P = 0.007, ${\eta }_{\text{p}}^2$ = 0.217] and DAP [F(3.129,56.318) = 4.893, P = 0.004, ${\eta }_{\text{p}}^2$ = 0.214] responses, with the greater increase following a spontaneous K-complex in men compared with women. Post hoc comparisons revealed significant differences in SAP between men and women at CC9 [Δ8.5 ± 6.4 vs. Δ2.7 ± 4.8 mmHg, t(18) = 2.293, P = 0.034] and CC10 [Δ9.3 ± 5.7 vs. Δ1.6 ± 4.0 mmHg, t(18) = 3.470, P = 0.003]. DAP was significantly different between men and women at CC8 [Δ4.6 ± 4.7 vs. Δ0.8 ± 2.9, t(18) = 2.222, P = 0.039], CC9 [Δ5.1 ± 4.6 vs. Δ−0.1 ± 2.5, t(18) = 3.090, P = 0.006], and CC10 [Δ4.1 ± 4.1 vs. Δ−0.7 ± 2.9, t(18) = 3.060, P = 0.007]. HR responses were not different between men and women following a spontaneous K-complex but was approached significance toward higher HR in men compared with women [F(4.580,82.445) = 2.112, P = 0.078, ${\eta }_{\text{p}}^2$ = 0.105].

Figure 2.

Changes in blood pressure and heart rate (HR) following spontaneous, isolated K-complexes during stage II sleep (C0) and 10 subsequent cardiac cycles (C1-C10). K-complexes increased systolic (SAP) and diastolic arterial pressure (DAP) in both men and women. The pressor response to a spontaneous K-complex was significantly higher in men when compared with women. Repeated-measures ANOVA with sex (men vs. women) as the between factor and cardiac cycle as the within factor were used to compared blood pressure and heart rate. Data depicted as means ± SD. *P < 0.05 between groups (men vs. women).

DISCUSSION

The present study examined the influence of K-complexes on blood pressure during normal, stage II sleep in men and women. In line with a previous study (14), spontaneous K-complexes were associated with acute increases in both SAP and DAP. To our knowledge, the present study is the first to demonstrate sex differences associated with the pressor response to a spontaneous K-complex during stage II sleep. Specifically, cardiac cycle × sex interactions were observed for both SAP and DAP, where K-complexes elicited augmented pressor response in men compared with women. These findings advance our current knowledge of autonomic responses to K-complexes during sleep, and the observed sex differences may have clinical implications with respect to the role of sex in hypertension and other cardiovascular diseases.

A recent study by de Zambotti and colleagues (22) was the first to interrogate the potential role of sex in cardiovascular responses to a K-complex. Briefly, it was reported that peak HR responses to a sound evoked K-complex in adolescent males were augmented compared with age-matched females. While this provides insights into sex-specific cardiac autonomic responses to K-complexes, the relationships between sex, K-complexes, and blood pressure remain unknown. Given the augmented HR response to a K-complex in men, we hypothesized a similar sex difference with respect to blood pressure. Interestingly, while our blood pressure data strongly support this initial hypothesis by demonstrating augmented SAP and DAP responses to K-complexes in men, we observed no significant difference in HR response. However, there was a trend (P = 0.078) toward augmented HR responsiveness to K-complexes in men compared with women. Importantly, and as depicted in Fig. 2, our pattern of a trend toward augmented HR reactivity near the beginning (C3 and C4) and near the end (C7 and C8) are consistent with the patterns observed by de Zambotti et al. (22). Finally, it is worth noting that de Zambotti et al. (22) utilized sound-evoked K-complexes, while the present study utilized spontaneous K-complexes, which could also explain some slight variances.

As noted previously, a number of studies have reported that K-complexes (i.e., spontaneous, associated with sensory stimuli, or disturbed breathing) are associated with, and often precede, bursts of MSNA (11–14, 25–27). While the present study did not measure MSNA, it is reasonable to suggest that there could be sex-differences in MSNA responses to a K-complex that might be contributing to the differences in blood pressure responsiveness. More specifically, it is possible that men could have greater number of MSNA bursts, or greater MSNA amplitude, during K-complexes. Alternatively, it is reasonable to hypothesize that there could be similar associations between K-complexes and MSNA in men and women but that the sympathetic vascular transduction may be more dramatic in men. In support of that concept are several studies reporting sex differences in sympathetic vascular transduction at rest and during various sympathoexcitatory interventions (28–32). Accordingly, future work examining the impact of sex on K-complexes and MSNA may be warranted.

The role of the K-complex during sleep remains controversial. There are three contrasting theories surrounding K-complexes, with three seemingly different and potentially opposing functions during sleep. Specifically, K-complexes have been suggested to have a sleep protective/enhancing influence (i.e., facilitation of slow wave sleep) (33), to have an arousal influence (i.e., transition to momentary wakefulness) (8), or to be a response to sensory stimuli (i.e., marker of disturbed sleep) (34). Sleep protective spontaneous K-complexes are often observed during the transition from stage II sleep to slow wave sleep, as delta power is increased following a spontaneous K-complex (35). In other cases, K-complexes can precede an arousal from stage II sleep (8). Finally, other K-complexes arise from sensory stimuli, most often in the literature as experimental administration of sound to trigger a degree of sleep disturbance (21, 36).

In addition to the acute associations between blood pressure and specific stages of sleep such as K-complexes, nocturnal blood pressure has a characteristic dipping pattern during sleep that has been shown to have cardiovascular benefits. Specifically, greater reductions of nocturnal blood pressure during sleep, and particularly dipping patterns that exceed 10% of daytime blood pressure, are associated with reduced cardiovascular morbidity and mortality (15, 16). At present, there is limited evidence to suggest any conclusive sex differences between nocturnal blood pressure profiles between men and women. However, there are some studies that report a blunted blood pressure dip in males compared with females, particularly when traits such as depression and anger are considered (17–20). As such, it is theoretically possible that the augmented pressor response to the numerous K-complexes over the course of the night could have implications for nocturnal blood pressure dipping patterns.

We acknowledge the following study limitations. First, our sample of women were not tested within a consistent menstrual phase. This was due to the fact that the data were obtained during a screening/familiarization night within the sleep laboratory before formal enrollment into an ongoing randomized control trial (NCT03567434). While menstrual phase has been shown to influence MSNA depending on the sex steroid surges between the early follicular and midluteal phases (37), there is no current evidence that menstrual phase has an impact on hemodynamic responses to K-complexes. In the present study, we had a fairly equal distribution across the ovarian cycle, with ∼40% of our female sample in luteal phase, ∼30% in ovulation phase, and ∼30% in the follicular phase. Second, we acknowledge that we analyzed randomly selected K-complexes across stage II sleep. Our goal was to include isolated and high amplitude K-complexes (i.e., free of other K-complexes at least 30 s prior), which are most indicative of MSNA in the periphery (11). Stage II was selected because of the clear amplitude changes that occur during a K-complex. Stage III K-complexes exist but are more challenging to clearly distinguish due to larger delta waves. As such, it is possible there may be different blood pressure response patterns during stage III K-complexes. Finally, we acknowledge there is a chance our study may be underpowered to detect sex differences given an a priori power analysis was not conducted. However, our sensitivity analysis minimizes this limitation for our primary outcome variables of SAP and DAP but may explain by we did not confirm the cardiac autonomic response from de Zambotti et al (22). Future work should aim to confirm sex differences related to the pressor response to a spontaneous K-Complex in men and women.

In summary, K-complexes are a common feature of normal sleep, a hallmark of stage II sleep, and are associated with acute increases of blood pressure that appears to be driven, in part, through activation of the sympathetic nervous system (12, 13). The present study demonstrates that the pressor response following an isolated, spontaneous K-complexes differs between men and women. These findings provide new insights relevant to well-known cardiovascular sex differences that exist between men and women, with particular insight into nocturnal blood pressure control. Future work examining the role of sex in sympathetic neural responsiveness to K-complexes appear warranted.

GRANTS

This work was supported by National Institute on Alcohol Abuse and Alcoholism Grant AA-024892 and by the Portage Health Foundation.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

I.M.G. and J.R.C. conceived and designed research; I.M.G. and J.R.C. performed experiments; I.M.G., C.A.S., and J.R.C. analyzed data; I.M.G., C.A.S., and J.R.C. interpreted results of experiments; I.M.G. prepared figures; I.M.G. and J.R.C. drafted manuscript; I.M.G., C.A.S., and J.R.C. edited and revised manuscript; I.M.G., C.A.S., and J.R.C. approved final version of manuscript.