The Relationship between Sleep and Autonomic Health

Abstract

Objective

To examine the relationship between sleep and resting Autonomic Nervous System (ANS) functioning in college students.

Participants

Participants were 141 undergraduate students (52 males) recruited from a large southeastern university during September-October 2017.

Methods

Participants completed self-report inventories (demographic and sleep characteristics). Resting state skin conductance (SC) and heart rate variability (HRV) were measured in a laboratory setting for ANS functioning.

Results

SC was positively associated with sleep quality (p=0.027), sleep latency (p=0.040), and use of sleep medication (p<0.001). Analyses yielded a negative association between the standard deviation of the normal-normal interval of heart beats (SDNN) and the self-reported amount of time to fall asleep each night (p=0.041). Sleep efficiency was negatively correlated with low frequency HRV (p=0.002).

Conclusions

Sleep components are associated with resting ANS activity, and targeted interventions focused on improved ANS functioning may benefit sleep quality in college students.

The maintenance and promotion of health is vital to college student success. As such, one of the overarching goals of the Healthy Campus 2020 initiative outlined by the U.S. Department of Health and Human Services is to promote quality of life, healthy development, and positive health behaviors on college campuses. However, in a recent survey, the American College Health Association reported that students attributed stress (30%) and sleep difficulties (21%) as factors that negatively affected their academic performance (e.g., lower grade or dropped courses). For example, in an examination of 1,125 college students, sleep habits were explored in conjunction with academic performance and physical health factors. Results revealed that students who had poor sleep quality reported significantly more problems associated with physical and psychological health than those who reported better sleep quality. Moreover, academic and emotional stress have been cited as factors that contribute to poor sleep quality. ^1,2 College students in particular are faced with innumerable stressors (e.g., academic performance, pressure to succeed, maintaining social relationships, dating, work-life balance, and post-graduation plans) and oftentimes sleep is sacrificed. ^3,4

Sleep quality and duration are critical to cognitive, emotional, and physical well-being. ^5–7 College students in particular are often subjected to irregular sleep schedules and often experience insufficient sleep and poor sleep quality. ^8,9 Generally speaking, lack of sleep can hinder academic success ¹ and lead to poor health outcomes such as obesity, ¹⁰ depression, ¹¹ and other chronic diseases. ¹² However, overall satisfaction with the sleep experience has been shown to be positively associated with well-being and positive affect. ¹³ For example, researchers have shown that sufficient and consistent sleep is linked to improved self-regulatory behaviors and decreased strain in college students. ¹⁴ Similarly, Lebensohn and colleagues reported that the two most significant wellness behaviors associated with higher well-being in medical students were exercise and restful sleep. ¹⁵ It is evident that sufficient sleep is important for overall well-being. However, research about the influence of sleep on resting autonomic nervous system (ANS) functioning in college students is lacking. In terms of the ANS, greater vagal activity (i.e. parasympathetic dominance) is viewed as conducive to health, whereas a dominant or overactive sympathetic branch proves to be detrimental to health. ^16–18 The purpose of this study was to examine the extent to which college students are at risk for autonomic imbalances as a function of sleep.

Sleep deficits cause over-activity in the stress response systems, and this deficit may induce alterations in autonomic control. ¹⁹ For example, cardiovascular researchers have demonstrated that poor quality of sleep and sleep deprivation can produce a reduction in vagal tone, and thus disrupt the ANS ability to inhibit sympathetic dominance. ^20–22 In a study of hypertensive patients, results revealed that poor sleep quality was associated with greater sympathetic nervous system (SNS) activation. ²³ Because sleep deficiencies are associated with increased neuroendocrine activity, sleep deprivation in particular is considered a chronic stressor that often leads to negative consequences in the brain and body systems. ^24,25

Modal responses of the ANS include cardiovascular, electrodermal, and respiratory changes. In the current study, we examined cardiovascular response via resting heart rate variability (HRV) and electrodermal response utilizing skin conductance (SC) functioning in healthy college students within our laboratory setting. SC provides a unique method for measuring autonomic functioning because alterations in sweating are mediated by cholinergic nerve activity. ²⁶ SC responses are associated with increasing stress and anxiety. ^27–29 In addition, elevated SC responses are associated with enhanced memory consolidation, emotional processing, and attention. ³⁰ In a recent study, Liu and colleagues found college students who were sleep deprived displayed increased SC levels following a difficult perceptual task compared to their well-rested counterparts. ³¹ Thus, explorations into SC can provide important information about ANS arousal, specifically in relation to sleep. HRV is defined as the variance between consecutive heartbeats or the RR interval, ³² and the parasympathetic/vagal activity is associated with the high-frequency component of HRV. ³³ Moreover, resting HRV is a biomarker for physical and psychological health. ^34,35

Despite the well-documented link between sleep and overall well-being, understanding the effect of sleep on autonomic balance in college students is warranted. Therefore, the purpose of the current study was to examine the relationship between aspects of sleep and the aforementioned indices of ANS functioning at rest. Our central hypothesis was that aspects of sleep would show a positive association with autonomic function (operationalized by SC and HRV indices). Accordingly, less sympathetic activity (i.e. lower SC level, & minimal low frequency HRV power) would be reflective of better sleep quality.

Methods

Participants

Participants were 141 undergraduate students (52 males) currently enrolled in a large Southern predominately White university. In terms of ethnicity, 83.6% identified as White/European American, 7.9% identified as African-American, 6.4% identified as Asian-Americans, 0.7% identified as Latino(a)/Hispanic, 0.7% identified as Native American, and 0.7% identified as Middle Eastern. The age of participants ranged from 18 to 43 with a mean age of 20.72 (SD = 3.26). Participants were recruited using the Psychology department’s undergraduate research website. Inclusion criteria were as follows: (a) must be at least 18 years of age, and (b) no medical diagnosis of sleep disorders (e.g., narcolepsy, sleep apnea, or sleep terrors). This study was conducted during the 2017 Fall semester, and was approved by the University’s Institutional Review Board (IRB) prior to data collection.

Apparatus Used

A desktop Dell computer with a 22-inch color monitor was used to administer the survey and demographic sheet for each student as they individually came to the research lab. In an adjacent examination room, an additional Dell desktop computer with two 22-inch color monitors was used to run the physiological software program used in this study. The physiological measures (SC, HR, & HRV) were sampled via the multichannel Procomp InfinitiTM and Biograph software system (Thought Technology Ltd., Montreal, Canada). All sensor leads were connected to this 8-channel encoder, and each participant’s physiological data were exported into a data file. SC was measured using cuffs with sensors fastened around the middle and index fingers of the non-dominant hand. SC is measured in micro-Siemens units, and these data were collected at 32 Hz. A higher level of skin conductance response indicates a higher level of arousal.

HR (beats per minute) and HRV were measured via 3-lead electrode sensors placed on both forearms and the left wrist of the participant as outlined by the Thought Technology Manual, and these data were collected at 256 Hz. The raw inter-beat interval (IBI) data were analyzed in the time and frequency domains. More specifically, HRV was assessed via the following frequency domains: very low frequency (VLF) power between the limits of 0.003 Hz and 0.04 Hz, low frequency (LF) power in the range of 0.04 Hz and 0.15 Hz, high frequency (HF) power in the range of 0.15 Hz and 0.40 Hz, and LF/HF ratio. The HF component reflects parasympathetic cardiac control and LF reflects sympathetic/parasympathetic balance. ³⁶ The time-domain measure included the standard deviation of normal-to-normal intervals (SDNN) of heart beats. In a recent study, borderline abnormal and abnormal SDNN readings were associated with increased risk of cardiovascular disease in middle aged adults.³⁷

Measures

Demographic Questionnaire

A self-report questionnaire was used to obtain information about the demographics of our sample population. Information such as: age, race, sex, and classification in school were collected.

Sleep Questionnaire

Quality of sleep was measured using the Pittsburgh Sleep Quality Index (PSQI). ³⁸ This self-administered questionnaire assesses quality of sleep during the previous month and contains 19 self-rated questions yielding seven components: subjective sleep quality (i.e., rating of sleep quality from poor to good), sleep latency (i.e., time it takes to fall asleep), sleep duration (i.e. how long asleep), sleep efficiency (i.e., percentage of time sleeping in a bed), sleep disturbance (i.e. how sleep is affected by disturbances), use of sleep medications (i.e. amount of usage), and daytime dysfunction (i.e. how sleep affects activities of daily living). Each component is scored from 0 to 3, yielding a global PSQI score between 0 and 21, with higher scores indicating lower quality of sleep. The PSQI is useful in identifying self-reported good sleepers and poor sleepers. A global PSQI score > 5 indicates that a person is a poor sleeper having severe difficulties in at least two areas or moderate difficulties in more than three areas.

Procedure

Upon arrival to the lab, all participants were asked to read and electronically sign the informed consent document. Once consent was given, participants were asked to complete a survey package (demographics and sleep quality). Once completed, participants were asked to sit quietly for 5 minutes prior to obtainment of physiological measures. After this acclimation period, participants were prepared for physiological recordings using the Thought Technology sensors interfaced with their software program. A measure of SC, HR, and HRV were measured simultaneously for 5 minutes in order to assess ANS functioning at rest. Physiological data were only collected at this time and no other measures were obtained. Participants were thanked for their cooperation and participation in this study.

Data analytic techniques

Data analyses for both the physiological and self-report data were performed using SPSS Version 25 for Windows. Artifacts (i.e. muscle tension, eye-blinks, etc.) from the physiological data were removed via the auto-edit functions of the Biograph (Physiological Suite) software program (Thought Technology Ltd., Montreal Canada). The frequency domains of the HRV indices were analyzed using fast Fourier transformation. The time domain of the HRV indices were computed from the IBI of the raw EKG signals. All alpha levels were set at 0.05. Demographic information is provided in Table 1. Bivariate Pearson correlations were computed in order to examine the association between aspects of sleep and resting ANS activity.

Table 1.

Descriptive Characteristics of Study Participants

Study Variables
Age (in years, M ± SD)	20.716 ± 3.256
Male (M; %)	52; 36.9%
Female (M; %)	89; 63.1%
White (M; %)	117; 83.0%
Non-White (M; %)	23; 16.3%
Av Sleep (in hours, M ± SD)	6.720 ± 0.993
Freshman (M; %)	51; 36.2%
Sophomore (M; %)	22; 15.6%
Junior (M; %)	43; 30.5%
Senior (M; %)	24; 17.0%

Results

Demographics Variables

Due to the disproportion in racial identity in the current sample (i.e. 83% Caucasian), we excluded racial differences from analysis in this study. There were no significant differences by sex or classification status on the dependent measures. For the current study, we collapsed across race, sex, and classification status and explored this college student population linearly as a whole.

Components of Sleep

Table 2 contains the means and standard deviations for our sleep measure. The average PSQI global score for our college sample was 7.10 (SD = 2.64), and a score greater than 5 was indicative of poor sleep quality. The sleep components yielding the highest scores were sleep latency (M = 1.33, SD = 0.66) and sleep duration (M=1.28, SD = 0.97). Collectively, these results revealed that college students in our sample reported difficulty falling asleep and staying asleep; therefore, indicating insufficient daily sleep. Additionally, results revealed no significant differences in terms of sex for the components of sleep (see Table 3).

Table 2.

Means and Standard Deviations for Components of PSQI

	Mean	Std. Dev.
Subjective Sleep Quality	1.225	0.600
Sleep Latency	1.333	0.662
Sleep Duration	1.284	0.973
Sleep Efficiency	0.567	0.839
Sleep Disturbance	1.183	0.455
Use of Sleep Medication	0.437	0.863
Daytime Dysfunction	1.078	0.773

Note. Scores range from 0–3 with higher scores representing worse sleep quality.

Table 3.

Independent Samples T-Test of Components of PSQI by Sex

	Male		Female
	Mean	SD	Mean	SD	p
Subjective Sleep Quality	1.289	0.605	1.191	0.601	0.356
Sleep Latency	1.269	0.564	1.371	0.713	0.381
Sleep Duration	1.327	1.004	1.258	0.960	0.688
Sleep Efficiency	0.539	0.917	0.584	0.795	0.756
Sleep Disturbance	1.192	0.487	1.180	0.441	0.876
Use of Sleep Medication	0.442	0.826	0.438	0.891	0.978
Daytime Dysfunction	1.058	0.850	1.079	0.727	0.877
Global PSQI	7.115	2.784	7.101	2.567	0.975

Note. PSQI = Pittsburgh Sleep Quality Index

Skin Conductance

Changes in SC are presented in Fig. 1. After the acclimation period, SC was assessed at rest for 5 minutes. The SC mean was 2.978 μS (SD = 2.612), and most participants displayed a slight decrease in SC during baseline. Correlational analyzes yielded significant associations between SC at rest and components of sleep. More specifically, poor sleep quality was positively associated with increased SC levels (r = 0.187, p = 0.027). Likewise, higher sleep latency scores were positively associated with increased SC levels (r = 0.173, p = 0.040). Reliance upon sleep medication was also positively associated with increased SC levels (r = 0.306, p < 0.001). As expected, Global PSQI scores were positively associated with increased SC levels (r = 0.188, p = 0.026) (see Table 4).

Fig 1.

Histogram of Skin Conductance level (SCL) at rest for all study participants. The mean SCL is 2.98 μS for the current college student sample. SCL reflects the tonic level of electrical conductivity of the skin.

Table 4.

Correlation between Skin Conductance and Components of PSQI

	r	p
Subjective Sleep Quality	0.187*	0.027
Sleep Latency	0.173*	0.040
Sleep Duration	0.009	0.913
Sleep Efficiency	−0.024	0.781
Sleep Disturbance	−0.053	0.530
Use of Sleep Medication	0.306**	0.000
Daytime Dysfunction	0.051	0.549
Global PSQI	0.188*	0.026

Note.

^*.Correlation is significant at the 0.05 level (2-tailed).

^**.Correlation is significant at the 0.01 level (2-tailed).

Heart Rate Variability

Heart rate and HRV were examined in the current study. Baseline measures were assessed at the same time SC was assessed. The means and standard deviations for the cardiovascular variables are presented in Table 5. To examine the relationship between components of HRV (i.e., time and frequency domains) and aspects of sleep, correlational analyses were performed. Explorations into the relationship between HRV and sleep components showed that the amount of time reported to fall asleep each night (sleep latency) was negatively associated with SDNN (r = −0.173, p = 0.041) (see Fig. 2) suggesting that the longer it takes one to go to sleep, the lower one’s HRV. In addition, sleep efficiency was negatively correlated with LF% (r = −0.262, p = 0.002). Based on our findings, we suggest that better sleep efficiency is associated with reduced sympathetic activation. No other relationships demonstrated significance in terms of HRV and sleep components (p > 0.05).

Table 5.

Means and Standard Deviations for components of HRV

	Mean	Std. Dev.
Heart Rate	79.904	11.339
VLF	22.328	9.836
LF	42.453	10.486
HF	34.777	12.992
LF/HF Ratio	4.604	36.292
SDNN	45.942	29.030

Note. VLF = Very Low Frequency spectrum (0.003–0.04Hz), LF = Low Frequency spectrum (0.04–0.15Hz), HF = High Frequency spectrum (0.15–0.40Hz), LF/HF ratio = the ratio between of Low Frequency to High Frequency Heart Rate Variability, SNDD = standard deviation of normal-normal intervals.

Fig 2.

The Relationship between time to fall asleep (in mins) and standard deviation of the normal-normal (SDNN) interval component of Heart Rate Variability. Lower SDNN is reflective of worse health. This figure clearly shows that lower SDNN (i.e. worse heart health) is associated with longer times to fall asleep. Thus, longer times to fall asleep is not conducive to health.

Comment

Sufficient sleep is vital to overall well-being, but for many college students, a good night’s sleep is elusive. For this reason, this study was designed to provide insight into the relationship between components of sleep and ANS function at rest in healthy college students. Investigations into both sympathetic and parasympathetic branches of the ANS have shown that autonomic imbalances are precursors to disease formation and other health-related risks. ^33,39,40 Findings from the current study indicate that both quality and quantity of sleep were associated with autonomic functioning at rest. Our principal findings revealed that: (a) college students routinely experienced difficulty in falling asleep and staying asleep which often led to the use of sleep medication, and (b) poor sleep quality and duration were associated with greater SC levels and lower HRV at rest. Collectively, our findings reveal that college students who are poor sleepers tend to display autonomic imbalances at rest. Moreover, they may be at increased risk for serious health problems later in life.

SC measures are routinely used to examine emotional reactivity ^28,30,41,42 and general arousal. ²⁶ In the current study, we found poor sleep quality, increased sleep medication use, and longer sleep latency times were associated with greater SC levels at rest. Our findings are similar to that of Liu and colleagues who reported sleep deprived college students to display higher SC reactivity to a stressful task compared to their well-rested counterparts. ³¹ Electrodermal activity such as SC is used to assess activity of the SNS in particular. ²⁷ SNS dominance leads to autonomic imbalance, and places one at risk for chronic health disorders. ^33,43,44 Therefore, the findings of the current study suggest self-reported that poor sleep quality and duration are associated with greater SNS dominance. Moreover, we argue that this autonomic imbalance interferes with sleep efficiency and the perceptions of a good night sleep. It is plausible that the strenuous life of a college student is reflective in a heightened stress response and the frequent use of sleep medication is an attempt to restore autonomic balance, and therefore sleep.

Another way to assess autonomic functioning is to examine HRV at rest. This approach has been widely used as a non-invasive tool to evaluate cardiovascular autonomic control in clinical and non-clinical populations ¹⁹ and can be defined as the physiological variation between successive beats of the heart, ⁴⁵ with greater variability being associated with better health. ⁴⁶ Researchers have shown that greater HRV is associated with more parasympathetic dominance in the ANS. ³⁵ Sleep is also parasympathetically driven as transitions from wakefulness to sleep have shown to shift autonomic balance to that of a more parasympathetic tone. ⁴⁷ Results of the current study indicated that increased sleep latency and decreased sleep efficiency were associated with lower HRV as measured by LF% and SDNN components. Both components of HRV reflect the relative dominance of each branch of the ANS. Researchers have shown that greater LF% is reflective of baroreceptors and primarily sympathetically driven, ^48,49 while SDNN is a time domain and reflects the overall influence of both divisions of the ANS. ⁵⁰ Our findings are consistent with findings of previous researchers who found reduced HRV associated with poor sleep quality. ⁵¹ Therefore, based upon our findings, we suggest that the longer it takes one to fall asleep, paired with less time in sleep, is associated with autonomic imbalance. Moreover, our findings lend support to previous researchers who suggested that sleep complications parlay into negative health outcomes. ^23,25,53

Sleep has protective properties that bolster health; ^23,25,52,53 however, insufficient sleep is a consequence to which many college students are subjected daily. ⁵⁴ Sleep difficulties can manifest in college students due to a variety of reasons such as: deficits in time and stress management, ⁵⁵ poor coping skills, ⁵⁶ and fear of being in academic peril. ⁵⁷ Based upon the results of the current study, we suggest that interventions focused on self-reported sleep difficulties in college students can be used as a non-invasive way to explore autonomic function and reduce the risk for health issues later in life.

Limitations

A few study limitations should be noted. First, we assessed sleep via self-report. Although the PSQI is a well-validated and often used measure of sleep quality, in both clinical and non-clinical populations ^58,59 it is still subjective in nature. Objective measures such as actigraphy in conjunction with self-report measures of sleep are warranted. In this study, we correlated subjective measures of sleep with objective measures of health (SC and HRV). We found that individuals who reported difficulty in sleeping to also displayed higher resting SC levels and lower HRV. Moreover, this study was correlational in nature and causal inferences cannot be determined. It is plausible that disturbances in autonomic functioning cause sleep problems. Further studies are warranted to ferret out these findings. Finally, the current study did not have a very diverse sample in terms of race or classification status, thus a more diverse sample is warranted in order to determine if the current study’s findings hold true.

Conclusions

The current study was designed to investigate the relationship between components of self-reported sleep and indices of the ANS in relation to college students. We found several components of sleep (subjective sleep quality, sleep latency, and the use of sleep medication) were associated with autonomic functioning. More specifically, poorer sleep quality was associated with greater sympathetic dominance in college students. Sleep difficulties may either be the cause of, or the result of, increased sympathetic activation. This increased arousal may have caused students to take longer to fall sleep each night, which in turn, led students to consume more sleep medication. The maintenance and promotion of health and wellness is critical for college student success. Sufficient sleep is a vehicle for improved memory, cognitive performance, and overall health. ^52,53,60 Although sleep medication may promote the process of going to sleep, students need to be educated about the pitfalls of reliance upon sleep medication. College administrators should focus on the implementation of brief psychoeducational courses ⁶¹ or other sleep hygiene intervention programs promoting sleep health and the benefits of sleep as a way of improving student well-being. ^62–64