Abstract
Aims.
Nondipping blood pressure (BP) is associated with higher risk for hypertension and advanced target organ damage. Insomnia is the most common sleep complaint in the general population. We sought to investigate the association between sleep quality and insomnia and BP nondipping cross-sectionally and longitudinally in a large, community-based sample.
Methods
A subset of the Wisconsin Sleep Cohort (n=502 for cross-sectional analysis and n=260 for longitudinal analysis) were enrolled in the analysis. Polysomnography measures were used to evaluate sleep quality. Insomnia symptoms were obtained by questionnaire. Blood pressure was measured by 24-hour ambulatory BP monitoring.Logistic regression models estimated cross-sectional associations of sleep quality and insomnia with BP nondipping. Poisson regression models estimated longitudinal associations between sleep quality and incident nondipping over a mean 7.4 years of follow-up. Systolic and diastolic nondipping were examined separately.
Results.
In cross-sectional analyses, difficulty falling asleep, longer waking after sleep onset, shorter and longer total sleep time, lower sleep efficiency and lower rapid eye movement stage sleep were associated with higher risk of systolic and diastolic BP nondipping. In longitudinal analyses, the adjusted relative risks (95% confidence interval) of incident systolic nondipping were 2.1 (1.3–3.5) for 1 hour longer waking after sleep onset, 2.1 (1.1–5.1) for 7–8h total sleep time, and 3.7 (1.3–10.7) for ≥ 8h total sleep time (compared to total sleep time 6–7 hours), and 1.9 (1.1–3.4) for sleep efficiency <0.8, respectively.
Conclusions.
Clinical features of insomnia and poor sleep quality are associated with nondipping BP. Our findings suggested nondipping might be one possible mechanism by which poor sleep quality was associated with worse cardiovascular outcommes.
Keywords: Blood pressure nondipping, hypertension, insomnia, sleep quality, Wisconsin sleep cohort
Condensed Abstracts.
The association between sleep quality and insomnia symptoms and BP nondipping was assessed cross-sectionally and longitudinally in a large, community-based sample. In cross-sectional analyses, difficulty falling asleep and poor objective sleep quality were associated with higher risk of systolic and diastolic BP nondipping. In longitudinal analyses, longer waking after sleep onset, too short and too long sleep time (compared to to total sleep time 6–7 hours) and lower sleep efficiency were associated with higher risk of SBP nondipping. In conclusion, clinical features of insomnia and poor sleep quality are associated with nondipping BP.
1. Introduction
Systemic blood pressure (BP) varies with different physiologic states and is typically 10% to 20% lower during sleep compared to wake [1]. However, some individuals do not experience this BP reduction (BP “dipping”) during sleep. A nocturnal BP that decreases less than 10% is defined as “nondipping”. Nondipping BP is associated with higher risk for hypertension in a normotensive population, and advanced target organ damage and poor cardiovascular prognosis among hypertensive patients [2–6].
Sleep has a ‘toning-down’ effect on a number of physiologic parameters, including the aforementioned effect of sleep on BP [7]. There is some evidence that both day-to-night and nighttime regulation of BP are associated with autonomic changes occurring during the wake-sleep cycle, suggesting that BP should be particularly sensitive to disturbances occurring during sleep, especially in the presence of hyperactivation of the sympathetic nervous system, such as that which may occur in people with insomnia [8].
Insomnia is the most common sleep complaint in the general population, with prevalence estimates ranging from 5% to 50% [9]. Insomnia refers to the subjective experience of insufficient restorative sleep due to difficulty falling asleep, difficulty maintaining sleep or early awakening [10]. There is emerging epidemiological evidence linking insomnia to higher cardiovascular morbidity and mortality [11–13]. In addition to subjective sleep complaints, objective measurements by polysomnography (PSG) have been used to assess sleep quality [14]. Studies have shown that PSG markers, including decreased sleep duration, poorer sleep continuity, total sleep time and longer sleep latency can be important for a variety of clinical outcomes [15,16].
There are only a few small cross-sectional studies[17–20], and no longitudinal studies that have assessed associations of insomnia and sleep quality with nighttime BP nondipping, which, as a precursor to cardiovascular disease [2,4,21], could be one mechanism linking insomnia and sleep quality with cardiovascular morbidity and mortality. In the present study we investigate whether self-reported insomnia symptoms and objective measures of sleep quality: 1) are cross-sectionally associated with BP nondipping, and 2) predict incident BP nondipping at an average 7.4-years (range: 3 to 15 years) follow-up in the Wisconsin Sleep Cohort study, an ongoing investigation of the natural history, causes, and consequences of common sleep disorders in a community-based sample of adults.
2. MATERIALS AND METHODS
2.1. Participants and data collection
Briefly, study participants composed a probability sample of employees of 4 Wisconsin state agencies, aged 30 to 60 years at the time of recruitment. Participants were invited to undergo overnight studies at baseline and follow-up at approximate 4-year intervals. All study volunteers provided signed informed consent prior to participating in study protocols. The Informed consent documents and study protocols for the ongoing Wisconsin Sleep Cohort Study, described in detail previously [22], were reviewed for compliance with the principles outlined in the Declaration of Helsinki and were approved by the University of Wisconsin-Madison Health Science Institutional Review Board.
Of 1,533 baseline cohort participants, 502 met the inclusion criteria for the cross-sectional analysis: (1) having an overnight polysomnography study, and (2) having ambulatory BP monitoring (ABPM) data collected within <5 months of PSG (mean=3.7 months) (Supplemental Figure 1A). Self-reported insomnia symptoms were assessed during the overnight PSG study visit. Participants taking antihypertensive medications were excluded; thus, the final cohort for the cross-sectional analysis included 399 participants.
Study participants with 2 or more ABPM studies were included in the longitudinal analyses. Among 702 participants with baseline ABPM studies, 368 men and women completed 2 or more 24‑hour ABPM follow-up studies at follow-up intervals ranging from 3–15 years (Supplemental Figure 1B). Reasons for lack of follow-up data included: not all baseline ABPM participants were invited for follow-up studies; refusal (mostly due to inconvenience); laboratory scheduling difficulties; and insufficient wear-time (e.g., a lack of sleeping ABPM observations). For analysis of incident nondipping blood pressure, we excluded participants who were nondipping at their baseline ABPM studies (assessed separately for systolic blood pressure (SBP) and diastolic blood pressure (DBP) analyses). We also excluded participants who were on antihypertensive medications at baseline. The remaining sample consisted of 260 participants to follow for development of systolic BP nondipping and 281 to follow for development of diastolic BP nondipping.
Sleep studies were conducted at the University of Wisconsin-Madison Clinical Research Unit, in a sleep-laboratory suite with home-like bedrooms.
2.2. Subjective sleep quality
Subjective sleep quality was assessed with responses to questions about insomnia symptoms and sleep satisfaction. The health questionnaire and mailed surveys included 4 items on insomnia: difficulty in getting to sleep (referred to as difficulty in falling asleep or initiating sleep), waking up repeatedly during the night (repeated nocturnal awakenings), waking up too early in the morning and being unable to get back to sleep (awakening too early), and waking up during the night and having a hard time getting back to sleep (difficulty getting back asleep). Response categories were never or rarely (≤once/month), sometimes (2–4 times/month), often (5–15 times/month), and almost always (16–30 times/month). Individual symptom severity was defined with a dichotomous variable: often/almost always (≥5 times/month) versus sometimes/less (<5 times/month). General insomnia (0, 1) was defined as having any of the insomnia symptoms at a frequency of “often or almost always”. Surveys also included questions about how often sleep was satisfactory (most of the time, some of the time, not usually or never), and the frequency of naps.
For the longitudinal analyses, we used two consecutive surveys to define insomnia. Participants were excluded from the longitudinal analyses of subjective sleep problems if they did not complete 2 or more surveys or if the self-report of insomnia or sleep complaints were not consistent on these surveys (n=79 excluded from the systolic BP cohort and n=87 excluded from the diastolic BP cohort).
2.3. Polysomnographically assessed markers
Objective data on sleep latency, waking after sleep onset (WASO), sleep efficiency, and total sleep time were obtained during the overnight protocol by means of full 18-channel polysomnography (Grass Heritage PSG Digital Sleep System; Grass Technologies, Warwick, Rhode Island). Sleep stage for each 30-second epoch was scored according to conventional criteria[23]. “Sleep latency” was defined as the amount of time (minutes) from “lights off” to the first of 3 consecutive epochs of stage 1 sleep or the first epoch of any other stage of sleep; “waking after sleep onset” as the amount of time (minutes) spent awake after first sleep onset; “total sleep time” as the total amount of time spent sleeping (minutes); and “sleep efficiency” (%) as the proportion of total sleep time out of total duration of time in bed from “lights out.” Total sleep time was analyzed as a continuous variable, as well as a categorical variable with <6 hours, 6–7 hours (reference), 7–8 hours and > 8 hours of sleep.
2.4. 24-Hour ABPM
The baseline 24-hour ABPM was performed within 4 months (range: 0.03 to 5 months) following the overnight sleep study, with the Accutracker II (Suntech Medical Instruments/ Eutectics Electronics, Raleigh, NC). Details of the study protocol and ABPM quality data have been previously published[24]. Briefly, ABPM readings were obtained at random intervals averaging every 15 to 20 minutes during wakefulness and every 30 minutes during sleep. Individual wake and sleep mean BPs were computed by averaging ABPM measurements during sleep and wake defined by participants’ recorded sleep and wake times. The time participants reported falling asleep and waking were used to establish the ABPM sleep period. Systolic BP nondipping was defined as having a ratio of mean systolic sleep BP to mean systolic wake BP ratio >0.9. Diastolic BP nondipping was defined analogously.
For longitudinal analyses, incidence of BP nondipping status was prospectively examined among 260 participants for systolic BP and 281 participants for diastolic BP over the entire follow-up time period. An outcome variable was created to represent the change in dipping status: no development of nondipping over the entire follow-up or incident nondipping (development of nondipping during any subsequent follow-up visit). Those individuals who fluctuated between dipping, nondipping, and dipping again were excluded (n = 11 for systolic BP and n = 8 for diastolic BP). The mean follow-up period was 7.4 years. A variable for the length of time of participants’ follow-up intervals was used in all analyses to account for individual differences.
2.5. Other Data Collection
Body mass index (BMI) was calculated from measured body weight and height (kg/m2). Participants completed a questionnaire on the night of the sleep study regarding the following covariates: current smoking, alcohol consumption (number of alcoholic drinks per week), caffeine consumption (number of caffeinated drinks per week), self-reported hours of exercise, and reported usual sleep duration on weekdays and weekends. Habitual sleep time in hours was calculated as the weighted average of weekday and weekend sleep time (i.e., [5 × weekday + 2 × weekend]/ 7). Alcohol and caffeine consumption were categorized into 0, 1–2 and >2 drinks/day. Self-reported physician-diagnosed chronic illness (coronary heart disease, hypertension and diabetes) and medications, including sedatives and antihypertensives, were obtained. On the same evening as the overnight studies, participants completed the Zung Self-Rating Depression Scale[25] and provided information on regularly taken medications for depression. As previously reported[16], we used a “modified” Zung scale that excluded 2 items (“I have trouble sleeping through the night” and “I get tired for no reason”) which may indicate insomnia and result in a built-in association between insomnia and depression.
Sleep-disordered breathing status was assessed by polysomnography and summarized as the apnea-hypopnea index (AHI). Using conventional cut points [23], the AHI was categorized as <5 events/h (little or no sleep-disordered breathing), 5–<15 events/h (mild), or ≥15 events/h (moderate or worse). Individual who used continuous positive airway pressure (CPAP) therapy on the overnight PSG were categorized with the moderate or worse sleep-disordered breathing group. Counts of sleep leg movements per hour of sleep were recorded. In 2014, a question about times of nocturia was added to the sleep questionnaire and nocturia frequency (≤1 vs >1 times per night) was collected. On average, nocturia measurement was collected 10 years after BP measurement.
2.6. Data Analysis
All analyses were performed with SAS software, release 9.4. Between-group comparisons were performed by two sample t-test or by Wilcoxon signed rank tests (for non-normally distributed variables). Assessment of normality was performed using the Shapiro-Wilk test. In cross-sectional analyses, linear regression was performed to estimate the association between these markers and the sleep/awake BP ratio, for systolic and diastolic separately; logistic regression was performed to estimate the association between sleep quality markers and BP nondipping. Multivariate models also included the following covariates: age; sex; BMI; current smoking; alcohol and caffeine consumption; chronic illnesses, including coronary heart disease, hypertension, diabetes, depression; nap (≤1 time vs >1 times per day); self-reported hours of weekly physical exercise; sleep apnea; use of sedative medications; leg movements; and the interval between the ABPM data collection and the overnight study visit, when the sleep data were collected.
For longitudinal analyses, Poisson regression models were performed to estimate the associations between sleep quality and incident nondipping over a mean (SD) 7.4 (3.1) years follow-up, for systolic nondipping and diastolic nondipping separately. In addition to baseline insomnia markers, the models included the following covariates: age, sex, BMI at baseline, BMI at follow-up, current smoking, alcohol and caffeine consumption, hypertension based on data from baseline visit, duration of follow-up, AHI, nap, physical exercise, use of sedative medications, leg movements, and whether or not an individual was on antihypertensive medication at any time during the follow-up period.
3. Results
3.1. Characteristics of participants
Demographic and clinical characteristics of the study participants are shown in Table 1. In the cross-sectional study, 21% of participants were defined as having nondipping SBP, and 38% defined as having insomnia symptoms. Participants with insomnia symptoms were on average older, had a higher prevalence of depression, and were more likely to be female, compared to participants without insomnia. No significant differences were noted between participants with insomnia and those without with respect to BMI, smoking, alcohol and coffee consumption, self-reported physician-diagnosed chronic illness or AHI. Participants with insomnia also had higher SBP and DBP sleep/awake ratios, a higher sleeping heart rate, and were more likely to have nondipping DBP.
Table 1.
Study Sample Demographic and Characteristics Stratified by Insomnia (n=399).
| Total | Not Insomnia | Insomnia | |
|---|---|---|---|
| N (%) | 399 | 247 (61.9) | 152(38.1) |
| Gender, female, N (%) | 167(41.9) | 94(38.1) | 73(48.0) |
| Age, years | 51.7±8.3 | 51.1±8.0* | 52.8±8.6 |
| BMI (kg/m2) | 30.2±6.3 | 30.4±6.4 | 29.8±6.2 |
| Sedative medication, N (%) | 13(3.3) | 7(2.8) | 6(4.0) |
| Current smoker, N (%) | 71(17.8) | 42(17.0) | 29(19.2) |
| Alcohol, >2 drinks/week, N (%) | 162(40.7) | 100(40.5) | 62(41.1) |
| Caffeine, >2 drinks/week, N (%) | 212(53.3) | 130(52.6) | 82(54.3) |
| Hypertension, N (%) | 32(8.0) | 20(8.1) | 12(7.9) |
| Diabetes, N (%) | 15(3.8) | 10(4.1) | 5(3.3) |
| Coronary heart disease, N (%) | 11(2.8) | 4(1.6) | 7(4.6) |
| Depression, N (%) | 87(22.5) | 42(17.6)** | 45(30.4) |
| Sleep SBP, mmHg | 108.0±12.4 | 107.5±11.5 | 108.7±13.7 |
| Sleep DBP, mmHg | 62.3±7.8 | 61.9±7.4 | 63.0±8.3 |
| Awake SBP, mmHg | 126.7±12.2 | 127.2±11.6 | 126.0±13.1 |
| Awake DBP, mmHg | 76.9±7.5 | 77.0±7.1 | 76.7±8.1 |
| Ratio of SBP sleep/awake | 0.85±0.06 | 0.85±0.06* | 0.86±0.06 |
| Ratio of DBP sleep/awake | 0.81±0.07 | 0.80±0.07* | 0.82±0.08 |
| SBP nondipper, N (%) | 82(20.6) | 46(18.6) | 36(23.7) |
| DBP nondipper, N (%) | 46(11.5) | 20(8.1)** | 26(17.1) |
| Sleeping heart rate, bpm | 65.3±9.1 | 64.7±9.3* | 66.3±8.6 |
| Waking heart rate, bpm | 76.8±9.7 | 76.8±9.5 | 76.8±10.2 |
| AHI <5 | 174(43.6) | 112(45.3) | 62(40.8) |
| 5≤ AHI <15 | 132(33.1) | 81(32.8) | 51(33.5) |
| AHI≥15 | 93(23.3) | 54(21.9) | 39(25.7) |
Continuous variables were presented as mean± standard deviation.
AHI, apnea-hypopnea index; DBP, diastolic blood pressure; SBP, systolic blood pressure.
P <0.05 for comparison between not-insomnia and insomnia groups.
P <0.01 for comparison between not-insomnia and insomnia groups.
Participants with insomnia symptoms had shorter self-reported sleep duration and less satisfaction with their usual sleep than participants who did not report insomnia symptoms (Table 2). Repeated nocturnal awakening was the most frequently reported insomnia symptom. Participants with insomnia symptoms had shorter PSG-measured sleep duration, slightly greater WASO and significantly less sleep efficiency. No significant differences were noted in different percent of time in sleep stages between participants with and without insomnia.
Table 2.
Study Sample Sleep Characteristics Stratified by Insomnia (n=399).
| Total | No Insomnia | Insomnia | |
|---|---|---|---|
| N (%) | 399 | 247 (61.9) | 152(38.1) |
| Habitual sleep time, hours | 7.1±0.9 | 7.2±0.8* | 6.9±0.9 |
| Most of time satisfied with sleep, N (%) | 269(67.4) | 203(82.2)** | 66(43.4) |
| Difficulty getting back to sleep, N (%) | 53(13.3) | 0 | 53(34.9) |
| Difficulty falling sleep, N (%) | 46(11.5) | 0 | 46(30.3) |
| Early morning awakenings, N (%) | 59(14.8) | 0 | 59(38.8) |
| Repeated nocturnal awakenings, N (%) | 111(27.8) | 0 | 111(73.0) |
| Nap >1 time per day | 239 (59.9) | 141 (57.1) | 98 (64.5) |
| PSG assessed markers of insomnia | |||
| Sleep latency, hour | 0.2±0.2 | 0.2±0.2 | 20.6±19.7 |
| Waking after sleep onset, hour | 1.0±0.7 | 1.0±0.7 | 1.0±0.6 |
| Total sleep time, hour | 6.5±1.0 | 6.6±1.0* | 6.3±0.9 |
| Sleep efficiency, % | 83.9±9.6 | 84.6±9.8* | 82.8±9.0 |
| % of Stage REM Sleep | 18.1±6.1 | 18.2±5.8 | 17.8±6.5 |
| % of stage 1 sleep | 9.9±6.3 | 9.9±6.1 | 10.0±6.7 |
| % of stage 2 sleep | 64.2±9.0 | 64.0±8.41 | 64.6±9.9 |
| % of stage3/4 sleep | 7.7±7.2 | 7.9±7.1 | 7.5±7.3 |
Continuous variables were presented as Mean± standard deviation.
REM, rapid eye movement.
P <0.05 for comparison between not-insomnia and insomnia groups.
No. of symptoms was treated as continuous variable.
Symptom-days per month can range from 0 (no symptoms reported) to 100 (=4*25, i.e., all four insomnia symptoms reported as “almost always”)
P <0.01 for comparison between not-insomnia and insomnia groups.
3.2. Cross-sectional associations of sleep characteristics and BP nondipping
The unadjusted and adjusted ORs for sleep characteristics and SBP and DBP nondipping were shown in Table 3. Difficulty falling asleep was associated with both SBP nondipping after adjusting for possible cofounders (OR= 2.6, 95% CI, 1.2–5.5) and DBP nondipping (OR=4.5, 95% CI, 1.9–10.8). Insomnia was associated with DBP nondipping (OR=2.4, 95% CI, 1.2– 4.7) but not significantly with SBP nondipping.
Table 3.
Cross-sectional Associations between Systolic and Diastolic BP Nondipping and Sleep Characteristics
| Systolic nondipping | Diastolic nondipping | |||
|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | |||
| Unadjusted | Adjusted | Unadjusted | Adjusted | |
| Self-report symptoms | ||||
| Difficulty getting back to sleep | 1.3 (0.7, 2.6) | 1.5 (0.7, 3.0) | 2.0 (0.9, 4.4) | 1.9 (0.8, 4.4) |
| Difficulty falling sleep | 2.3 (1.2, 4.5)* | 2.6 (1.2, 5.5)* | 3.8 (1.8, 7.9)** | 4.5 (1.9, 10.8)** |
| Early morning awakenings | 1.0 (0.5, 2.0) | 1.0 (0.5, 2.4) | 1.1 (0.4, 2.5) | 1.2 (0.5, 3.0) |
| Repeated nocturnal awakenings | 1.1(0.6, 1.9) | 1.1 (0.6, 2.0) | 1.6 (0.9, 3.1) | 1.8 (0.9, 3.7) |
| Insomnia (# of symptoms) | ||||
| 0 | Ref | Ref | Ref | Ref |
| 1 | 1.1 (0.6, 2.1) | 1.3 (0.6, 2.6) | 1.7 (0.8, 3.8) | 1.7 (0.7, 4.2) |
| 2 | 2.5 (1.1, 5.4)* | 2.9 (1.2, 6.8)* | 4.8 (2.0, 11.4)* | 5.6 (2.1, 15.3)** |
| 3 | 0.9 (0.3, 2.8) | 0.9 (0.3, 3.2) | 0.5 (0.1, 3.9) | 0.5 (0.06, 4.2) |
| 4 | 1.5 (0.4, 5.6) | 1.7 (0.4, 6.9) | 3.7 (0.9, 14.6) | 5.5 (1.2, 25.5)* |
| Habitual sleep time, hours | 1.0 (0.95, 1.12) | 0.8 (0.6, 1.1) | 0.9 (0.7, 1.4) | 0.9 (0.6, 1.3) |
| PSG assessed markers | ||||
| Sleep latency, hours | 2.6 (0.9, 7.8) | 2.5 (0.7, 8.7) | 1.3 (0.3, 5.8) | 1.4 (0.3, 7.0) |
| Waking after sleep onset, hours | 1.5 (1.1, 2.1)* | 1.6 (1.1, 2.3)* | 1.5 (1.02, 2.3)* | 1.6 (0.98, 2.5) |
| TST <6 h | 1.9 (1.6, 3.4)* | 2.0 (1.1, 3.9)* | 1.3 (0.7, 2.6) | 1.3 (0.6, 2.7) |
| 6 h ≤TST <7h | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| 7 h ≤TST <8h | 0.4 (0.2, 0.9)* | 0.3 (0.1, 0.9)* | 0.4 (0.2, 1.2) | 0.3 (0.1, 1.01) |
| TST≥ 8 h | 1.6 (0.6, 4.0) | 1.7 (0.6, 4.6) | 0.3 (0.054, 2.1) | 0.2 (0.03, 2.1) |
| Sleep efficiency, per 10% increment | 0.7 (0.5, 0.9)** | 0.7 (0.5, 0.9)** | 0.7(0.6, 0.9)* | 0.7 (0.5, 1.0) |
| % of REM sleep, per 10% increment | 0.6 (0.4, 0.9)* | 0.7 (0.4, 1.0) | 0.5 (0.3, 0.8)** | 0.4 (0.2, 0.8)** |
AHI, apnea-hypopnea index; CI, confidence interval; OR, odds ratio; REM, rapid eye movement; TST: total sleep time.
Analysis adjusted for age, sex, BMI, smoke, caffeine and alcohol consumption, diabetes, coronary heart disease, hypertension, depression, use of sedative medication, nap frequency, self-reported hours of weekly physical exercise, AHI, leg movements, and time between ABPM and insomnia survey.
P <0.05.
P <0.01.
PSG-measured wake after sleep onset, sleep efficiency and % time in rapid eye movement (REM) sleep all were associated with SBP and DBP nondipping in unadjusted models. After full adjustment, the association of waking after sleep onset and sleep efficiency and SBP nondipping remained (OR(95% CI) = 1.6 (1.1, 2.3) and 0.7 (0.5, 0.9) for WASO and sleep efficiency, respectively). Additionally, the fully adjusted associations of WASO and % time in REM sleep and DBP nondipping were not attenuated (OR (95% CI) = 1.6 (0.98, 2.5) and 0.4 (0.2, 0.8), respectively). Compared to participants with total sleep time of 6 to 7 hours, those sleeping less than 6 hours per night had adjusted OR for SBP nondipping of 2.0 (95% CI, 1.1–3.9), and those sleeping 7–8 hours had adjusted OR of 0.3 (95% CI, 0.1–0.9)
SBP and DBP nondipping were not significantly associated with difficulty getting back to sleep, early morning awakening, repeated nocturnal awakening, sleep latency and other sleep stages. When further adjusting for nocturia, most of the associations remained similar (data not shown). However, difficulty falling asleep was more strongly associated with both SBP nondipping (OR=3.6, 95% CI,1.6–8.3) and DBP nondipping (OR 5.7, 95% CI, 2.2–14.7) with adjument for nocturia.
3.3. Longitudinal association of sleep quality and blood pressure nondipping.
Those with longitudinal 24-hour ABPM data (≥2 data points) had a higher average BMI and higher prevalence of SBP and DBP nondipping than the total sample with ABPM (≥1 data point) (Supplemental table 1). No significant differences in baseline characteristics were found between systolic cohort participants included in the subjective insomnia analysis and those excluded (Supplemental table 2). Participants excluded from the subjective insomnia analyses (due to missing insomnia data) in the diastolic cohort were less liklely to have ever used antihypertensive medication than those who remained in the analyses. (Supplemental table 3.)
The mean (range) follow-up period was 7.4 (3–15) years, during which 49 (20%) developed systolic nondipping and 36 (13%) developed diastolic nondipping. General insomnia (having any of the 4 symptoms) and the specific symptom of difficulty falling asleep at baseline were both associated with higher risk (though non-significant) of incident SBP and DBP nondipping (Table 4 and Table 5). No significant associations were found between other sleep complaints and BP nondipping (data not shown). Longer duration of waking after sleep onset was associated with greater incident SBP nondipping risk (OR=1.8, 95% CI, 0.9–4.1). Adjusted SBP nondipping risk was 2- to 3- fold higher for participants with total sleep time >8 hours versus those with total sleep time 6–7 hours. Higher risk of SBP nondipping was observed for lower sleep efficiency (<0.8) in adjusted analyses (OR=1.9, 95% CI, 1.1–3.4). The associations remained similar after adjusting for nocturia. There were no significant associations between sleep quality markers and DBP nondipping in adjusted analyses.
Table 4.
Relative Risk of SBP Nondipping at 7.4-Year Follow-up According to Sleep Quality.
| Characteristics | Total sample, n | Incident nondipping, n(%) | RR (95% CI) |
|
|---|---|---|---|---|
| Unadjusted | Adjusted | |||
| Insomnia | ||||
| Yes | 55 | 14 (25.5) | 1.5 (0.8, 3.1) | 1.4 (0.5, 3.0) |
| No | 115 | 19 (16.5) | 1.0 (Ref) | 1.0 (Ref) |
| Missing | 79 | |||
| Difficulty falling asleep | ||||
| Often/almost always | 15 | 4 (26.7) | 1.4 (0.5, 4.1) | 1.3 (0.4, 3.9) |
| Sometimes/less | 184 | 26(18.5) | 1.0 (Ref) | 1.0 (Ref) |
| Missing | 50 | |||
| Sleep latency, hours | 249 | 49 (19.7) | 1.1 (0.4, 3.7) | 1.1 (0.3, 3.5) |
| Waking after sleep onset, per hour | 249 | 48 (19.7) | 1.7 (1.1, 2.6)* | 2.1 (1.3, 3.5)* |
| Total sleep time | ||||
| Tst< 6 h | 85 | 18 (21.2) | 1.9 (0.9, 4.0) | 1.8 (0.9, 4.1) |
| 6 h ≤TST <7h | 96 | 11 (11.5) | 1.0 (Ref) | 1.0 (Ref) |
| 7 h ≤TST <8h | 53 | 14 (26.4) | 2.3 (1.0, 5.1) § | 2.1 (1.1, 5.1)* |
| TST≥ 8 h | 15 | 6 (40.0) | 3.5 (1.3, 9.4)* | 3.7 (1.3, 10.7)* |
| Sleep efficiency | ||||
| Sleep efficiency ≥ 0.8 | 191 | 32 (16.8) | 1.0 (Ref) | 1.0 (Ref) |
| Sleep efficiency < 0.8 | 58 | 17 (29.3) | 1.8 (1.0, 3.3)* | 1.9 (1.1, 3.4)* |
Abbreviations: CI, confidence interval; RR, relative risk; tst, total sleep time, SBP: systolic blood pressure.
Analysis adjusted for baseline age, sex, BMI at baseline and follow-up, current smoking, hypertension, alcohol consumption from the baseline visit; duration of follow-up; self-reported hours of weekly physical exercise, use of sedative medication, and whether or not an individual was on antihypertensive medication at any time during the follow-up time period, nap frequency, leg movements, and sleep disordered breathing (AHI).
P <0.05.
P=0.05.
Table 5.
Relative Risk of DBP Nondipping at 7.4-Year Follow-up According to Sleep Quality.
| Characteristics | Total sample, n | Incident nondipping, n(%) | RR (95% CI) |
|
|---|---|---|---|---|
| Unadjusted | Adjusted | |||
| Insomnia | ||||
| Yes | 62 | 11(17.7) | 1.5 (0.7, 3.2) | 1.3 (0.5, 3.2) |
| No | 124 | 15(12.1) | 1.0 (Ref) | 1.0 (Ref) |
| Missing | 87 | |||
| Difficulty falling asleep | ||||
| Often/almost always | 18 | 4(22.2) | 1.9 (0.7, 5.5) | 1.6 (0.5, 5.4) |
| Sometimes/less | 198 | 23 (11.6) | 1.0 (Ref) | 1.0 (Ref) |
| Missing | 57 | |||
| Sleep latency, hours | 273 | 36 (13.2) | 0.5 (0.05, 3.7) | 0.5 (0.05, 4.0) |
| Waking after sleep onset, per hour | 273 | 36 (13.2) | 0.9 (0.5, 1.7) | 1.00 (0.5, 2.2) |
| Total sleep time | ||||
| Tst< 6 h | 91 | 9 (9.9) | 0.7 (0.3, 1.7) | 0.8 (0.3, 1.9) |
| 6 h ≤TST <7h | 104 | 14 (13.5) | 1.0 (Ref) | 1.0 (Ref) |
| 7 h ≤TST <8h | 62 | 8 (12.9) | 1.0 (0.4, 2.3) | 1.0 (0.4, 2.4) |
| TST≥ 8 h | 16 | 5 (31.3) | 2.3(0.8, 6.4) | 2.1 (0.7, 6.7) |
| Sleep efficiency | ||||
| Sleep efficiency ≥ 0.8 | 213 | 28 (13.2) | 1.0 (Ref) | 1.0 (Ref) |
| Sleep efficiency < 0.8 | 60 | 8 (13.3) | 1.0 (0.5, 2.3) | 1.2 (0.5, 2.8) |
Abbreviations: CI, confidence interval; DBP: diastolic blood pressure RR, relative risk; tst, total sleep time.
Analysis adjusted for baseline age, sex, BMI at baseline and follow-up, current smoking, hypertension, alcohol consumption from the baseline visit; duration of follow-up; self-reported hours of weekly physical exercise, use of sedative medication, and whether or not an individual was on antihypertensive medication at any time during the follow-up time period, nap frequency, leg movements, and sleep disordered breathing (AHI).
3.4. Sleep BP to Wake BP ratio
In addition to testing the dichotomous dipping outcomes, we also evaluated the associations of sleep quality markers and the BP sleep/wake ratios as continuous variables (Supplemental Table 4). The results of these analyses were consistent with the results of the main analyses of SBP and DBP nondipping.
4. Discussion
In this population-based study, we documented significant cross-sectional associations of subjective and objective sleep quality markers with BP nondipping. In cross-sectional analyses, the blunted BP fall occurred in association with the symptom of difficulty falling asleep, longer duration of waking after sleep onset and shorter total sleep time. Lower sleep efficiency was associated with risk of SBP nondipping while lower %REM sleep was associated with DBP nondipping. In longitudinal analyses, longer waking after sleep onset, longer total sleep time and lower sleep efficiency were associated with higher risk of incident SBP nondipping. These findings suggest that sleep quality – measured both subjectively and objectively – may have a role in the development of BP nondipping.
Our results are consistent with other published cross-sectional studies. Loredo et al[26] observed that nocturnal BP nondipping correlated with polysomnographic indices of poor sleep, including less stage 4 sleep and longer duration of waking after sleep onset. Hypertensive non-dippers also had less stage 4 sleep and higher microarousal rates [27]. Yilmaz et al [17] found that in younger hypertensive participants, sleep quality, sleep efficiency and sleep disturbance scores measured by the Pittsburgh Sleep Quality Index were significantly higher in non-dippers. Blunted SBP dipping also has been observed in normotensive participants with chronic insomnia, and is associated with higher activity in EEG β frequency [18]. Matthews et al [19] found that greater sleep/awake ratios of BP were associated with more fragmented sleep, more light sleep and less REM sleep. Another study found that poor sleep quality, stressful status and higher activation of the sympathetic nervous system were predictors of nondipping hypertension [20].
We also found that, compared to sleep time that was between 6 and 7 hours, shorter and longer sleep time both were associated with higher risk of SBP nondipping in cross-sectional and longitudinal analyses. Studies have shown that short and long self-reported sleep durations (compared to 6–8 hours) are associated with an increased risk of cardiovascular events and all-cause mortality[28,29]. Our study and the results from these other published studies are consistent with a U-shaped association of sleep duration with SBP nondipping and mortality, with worse outcomes at both ends of the distribution.
Our findings, as well as prior findings, of an association between sleep quality and nocturnal nondipping are important because of demonstrated associations between nondipping and advanced target organ damage and poor cardiovascular prognosis [4,5,21]. Studies have underscored a significant association between symptoms of insomnia and long-term cardiovascular mortality, independent of identified risk factors [12,30,31]. Difficulty falling asleep was found to predict cardiovascular mortality in a middle-aged population[31,32].These data suggest that, as with other cardiovascular risk factors, poor sleep quality may exert negative effects on the cardiovascular system over a long time before onset of overt disease.
Our data suggest that nondipping may be one of several possible mechanisms by which such outcomes may be initiated or exacerbated among people with poor sleep quality. Studies have shown that individuals with chronic insomnia suffer from increased arousal and sympathetic activation, implying that they could be at increased risk for development of coronary artery disease [13,33]. The activity of the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system have been found to relate positively to the degree of objectively-assessed sleep disturbance [34]. It is possible the presence of higher nocturnal blood pressure and lack of dipping could induce a persistently higher cardiovascular burden, and lead to cardiovascular damage and increase long-term cardiovascular risk. Examing dipping patterns has been suggested as part of hypertension treatment guidelines [35]. Evidence suggests that chronotherapy to reverse a non-dipping pattern back to a dipping pattern is feasible and useful [36,37]. An alternative (though not mutually exclusive) interpretation of our findings is that BP nondipping and insomnia may be parallell epiphenomena of an underlying derangement of autonomic nervous system function (in which case, e.g., treatment of insomnia or sleep disturbance may not affect nondipping status if the underlying cause of both remains unaltered).
Our study has several strengths. First, to our knowledge, this study is the first to investigate the longitudinal association between sleep quality and BP nondipping, with sleep quality measured before the emergence of nondipping status. This strengthens the evidence that sleep quality might play a causal role in the development of systolic nondipping. Second, our study was performed in a community-based cohort, enabling us to measure the association of sleep quality and nondipping in a nonclinical population. Third, our data were based on ABPM using actual sleep time and wake time recorded by participants and not arbitrary preset times. This allowed for better accuracy in defining sleep and wake BP. Lastly, we were able to control for confounding effects of many potentially important covariates, such as sleep-disordered breathing, hypertension, smoking, alcohol consumption, physical activity, BMI and change in BMI over time.
Our study has limitations. One important limitation is the small sample size in our longitudinal analysis. Only 49 participants developed SBP nondipping. Larger study samples yielding greater numbers of incident nondipping events are warranted for confirmation of our findings. Second, we did not follow up all participants who had a baseline 24-hour BP study. Our analysis of the population characteristics in the total eligible follow-up sample versus the actual follow-up sample showed a higher BMI and higher percentage of SBP and DBP nondipping in the total sample. However, we adjusted for baseline and follow-up BMI, and all the baseline nondippers were excluded for the longitudinal analysis. It is possible that the difference reflects better health among those who underwent multiple follow-up ABPM assessments and may have led to fewer cases of nondipping. Third, our objective measures of sleep quality are based on a single night of polysomnography. If participants with nondipping BP, compared with participants with dipping BP, had more trouble sleeping in the laboratory, their sleep quality could be relatively underestimated. However, participants were asked to compare their sleep during the polysomnography night with their usual night’s sleep (much worse, worse, about the same, better, much better), and we found no difference by BP dipping status. Fourth, despite adjusting for many variables, there may have been other potential cofounding factors associated with both sleep quality and nondipping that were not accounted (or fully accounted) for, such as nocturia and periodic limb movements. Participants with nocturia had a lesser drop in BP during sleep compared with those without nocturia [38,39], and generally have lower sleep quality [40]. Though associations between sleep quality and nondipping remained similar after we further adjusted for nocturia, we might not be able to fully account for the confounding given the time gap between BP and nocturia assessments.
Our findings of the longitudinal association of sleep quality with nocturnal SBP nondipping may have clinical and public health relevance, since insomnia, poor sleep quality, and hypertension are prevalent in the general population[10,41]. Our findings are consistent with a potential cause-effect relationship between poor sleep quality and SBP nondipping; SBP nondipping may be one mechanism by which sleep quality is associated with poor cardiovascular outcomes; and sleep quality may be relevant for both quality of life and cardiovascular health.
Supplementary Material
Acknowledgments
We thank the staff and participants of the Wisconsin Sleep Cohort for their important contributions and Drs. K. Mae Hla and Terry Young for oversight of the ABPM data collection.
Sources of Funding
This work was supported by the National Heart, Lung, and Blood Institute (R01HL62252), National Institute on Aging (1R01AG036838) and the National Center for Research Resources (1UL1RR025011) at the US National Institutes of Health.
Footnotes
Disclosures
There are no relevant conflicts of interest on the part of any study authors. There are no relevant financial, personal or professional relationships with other people or organizations to disclose.
References
- 1.Parati G, Ochoa JE, Lombardi C, Bilo G. Assessment and management of blood-pressure variability. Nature Reviews Cardiology 2013; 10 (3):143. [DOI] [PubMed] [Google Scholar]
- 2.Fagard RH, Celis H, Thijs L, Staessen JA, Clement DL, De Buyzere ML, et al. Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension. Hypertension 2008; 51 (1):55–61. [DOI] [PubMed] [Google Scholar]
- 3.Boggia J, Li Y, Thijs L, Hansen TW, Kikuya M, Björklund-Bodegård K, et al. Prognostic accuracy of day versus night ambulatory blood pressure: a cohort study. The Lancet 2007; 370 (9594):1219–1229. [DOI] [PubMed] [Google Scholar]
- 4.Routledge FS, McFetridge-Durdle JA, Dean C. Night-time blood pressure patterns and target organ damage: a review. Canadian Journal of Cardiology 2007; 23 (2):132–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Udani S, Lazich I, Bakris GL. Epidemiology of hypertensive kidney disease. Nature Reviews Nephrology 2011; 7 (1):11. [DOI] [PubMed] [Google Scholar]
- 6.Ohkubo T, Hozawa A, Yamaguchi J, Kikuya M, Ohmori K, Michimata M, et al. Prognostic significance of the nocturnal decline in blood pressure in individuals with and without high 24-h blood pressure: the Ohasama study. Journal of hypertension 2002; 20 (11):2183–2189. [DOI] [PubMed] [Google Scholar]
- 7.Tobaldini E, Costantino G, Solbiati M, Cogliati C, Kara T, Nobili L, et al. Sleep, sleep deprivation, autonomic nervous system and cardiovascular diseases. Neuroscience & Biobehavioral Reviews 2017; 74:321–329. [DOI] [PubMed] [Google Scholar]
- 8.Thomas SJ, Calhoun D. Sleep, insomnia, and hypertension: current findings and future directions. Journal of the American Society of Hypertension 2017; 11 (2):122–129. [DOI] [PubMed] [Google Scholar]
- 9.Morin CM, Benca R. Chronic insomnia. The Lancet 2012; 379 (9821):1129–1141. [DOI] [PubMed] [Google Scholar]
- 10.Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep medicine reviews 2002; 6 (2):97–111. [DOI] [PubMed] [Google Scholar]
- 11.Chien K-L, Chen P-C, Hsu H-C, Su T-C, Sung F-C, Chen M-F, et al. Habitual sleep duration and insomnia and the risk of cardiovascular events and all-cause death: report from a community-based cohort. Sleep 2010; 33 (2):177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sofi F, Cesari F, Casini A, Macchi C, Abbate R, Gensini GF. Insomnia and risk of cardiovascular disease: a meta-analysis. European journal of preventive cardiology 2014; 21 (1):57–64. [DOI] [PubMed] [Google Scholar]
- 13.Javaheri S, Redline S. Insomnia and risk of cardiovascular disease. Chest 2017; 152 (2):435–444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Young T, Rabago D, Zgierska A, Austin D, Laurel F. Objective and subjective sleep quality in premenopausal, perimenopausal, and postmenopausal women in the Wisconsin Sleep Cohort Study. Sleep 2003; 26 (6):667–672. [DOI] [PubMed] [Google Scholar]
- 15.Terzano MG, Parrino L, Spaggiari MC, Palomba V, Rossi M, Smerieri A. CAP variables and arousals as sleep electroencephalogram markers for primary insomnia. Clinical Neurophysiology 2003; 114 (9):1715–1723. [DOI] [PubMed] [Google Scholar]
- 16.Szklo-Coxe M, Young T, Peppard PE, Finn LA, Benca RM. Prospective associations of insomnia markers and symptoms with depression. American journal of epidemiology 2010:kwp454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yilmaz MB, Yalta K, Turgut OO, Yilmaz A, Yucel O, Bektasoglu G, et al. Sleep quality among relatively younger patients with initial diagnosis of hypertension: dippers versus non-dippers. Blood Press 2007; 16 (2):101–105. [DOI] [PubMed] [Google Scholar]
- 18.Lanfranchi PA, Pennestri M-H, Fradette L, Dumont M, Morin CM, Montplaisir J. Nighttime blood pressure in normotensive subjects with chronic insomnia: implications for cardiovascular risk. Sleep 2009; 32 (6):760–766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Matthews KA, Kamarck TW, M HH, Strollo PJ, Owens JF, Buysse DJ, et al. Blood pressure dipping and sleep disturbance in African-American and Caucasian men and women. Am J Hypertens 2008; 21 (7):826–831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Huang Y, Mai W, Hu Y, Wu Y, Song Y, Qiu R, et al. Poor sleep quality, stress status, and sympathetic nervous system activation in nondipping hypertension. Blood pressure monitoring 2011; 16 (3):117–123. [DOI] [PubMed] [Google Scholar]
- 21.Staessen JA, Thijs L, Fagard R, O’brien ET, Clement D, de Leeuw PW, et al. Predicting cardiovascular risk using conventional vs ambulatory blood pressure in older patients with systolic hypertension. Jama 1999; 282 (6):539–546. [DOI] [PubMed] [Google Scholar]
- 22.Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. New England Journal of Medicine 1993; 328 (17):1230–1235. [DOI] [PubMed] [Google Scholar]
- 23.Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. 1968. [DOI] [PubMed]
- 24.Hla KM, Young TB, Bidwell T, Palta M, Skatrud JB, Dempsey J. Sleep apnea and hypertension: a population-based study. Annals of Internal Medicine 1994; 120 (5):382–388. [DOI] [PubMed] [Google Scholar]
- 25.Zung WW. A self-rating depression scale. Archives of general psychiatry 1965; 12 (1):63–70. [DOI] [PubMed] [Google Scholar]
- 26.Loredo JS, Nelesen R, Ancoli-Israel S, Dimsdale JE. Sleep quality and blood pressure dipping in normal adults. Sleep: Journal of Sleep and Sleep Disorders Research 2004. [DOI] [PubMed] [Google Scholar]
- 27.Routledge F, McFetridge-Durdle J. Nondipping blood pressure patterns among individuals with essential hypertension: a review of the literature. European Journal of Cardiovascular Nursing 2007; 6 (1):9–26. [DOI] [PubMed] [Google Scholar]
- 28.Cappuccio FP, Cooper D, D’elia L, Strazzullo P, Miller MA. Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies. European heart journal 2011; 32 (12):1484–1492. [DOI] [PubMed] [Google Scholar]
- 29.Sabanayagam C, Shankar A. Sleep duration and cardiovascular disease: results from the National Health Interview Survey. Sleep 2010; 33 (8):1037–1042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sofi F, Cesari F, Casini A, Macchi C, Abbate R, Gensini GF. Insomnia and risk of cardiovascular disease: a meta-analysis. European journal of preventive cardiology 2014; 21 (1):57–64. [DOI] [PubMed] [Google Scholar]
- 31.Mallon L, Broman JE, Hetta J. Sleep complaints predict coronary artery disease mortality in males: a 12‐year follow‐up study of a middle‐aged Swedish population. Journal of internal medicine 2002; 251 (3):207–216. [DOI] [PubMed] [Google Scholar]
- 32.Grandner MA, Jackson NJ, Pak VM, Gehrman PR. Sleep disturbance is associated with cardiovascular and metabolic disorders. Journal of sleep research 2012; 21 (4):427–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Li Y, Vgontzas AN, Fernandez-Mendoza J, Bixler EO, Sun Y, Zhou J, et al. Insomnia with physiological hyperarousal is associated with hypertension. Hypertension (Dallas, Tex : 1979) 2015; 65 (3):644–650. [DOI] [PubMed] [Google Scholar]
- 34.Vgontzas AN, Bixler EO, Lin H-M, Prolo P, Mastorakos G, Vela-Bueno A, et al. Chronic insomnia is associated with nyctohemeral activation of the hypothalamic-pituitary-adrenal axis: clinical implications. The Journal of Clinical Endocrinology & Metabolism 2001; 86 (8):3787–3794. [DOI] [PubMed] [Google Scholar]
- 35.Nerenberg KA, Zarnke KB, Leung AA, Dasgupta K, Butalia S, McBrien K, et al. Hypertension Canada’s 2018 guidelines for diagnosis, risk assessment, prevention, and treatment of hypertension in adults and children. Canadian Journal of Cardiology 2018; 34 (5):506–525. [DOI] [PubMed] [Google Scholar]
- 36.Sleight P, Yusuf S, Pogue J, Tsuyuki R, Diaz R, Probstfield J, et al. Blood-pressure reduction and cardiovascular risk in HOPE study. The Lancet 2001; 358 (9299):2130–2131. [DOI] [PubMed] [Google Scholar]
- 37.Svensson P, de Faire U, Sleight P, Yusuf S, Östergren J Comparative effects of ramipril on ambulatory and office blood pressures: a HOPE Substudy. Hypertension (Dallas, Tex : 1979) 2001; 38 (6):e28–e32. [DOI] [PubMed] [Google Scholar]
- 38.Agarwal R, Light RP, Bills JE, Hummel LA. Nocturia, nocturnal activity, and nondipping. Hypertension (Dallas, Tex : 1979) 2009; 54 (3):646–651. [DOI] [PubMed] [Google Scholar]
- 39.Matsumoto T, Tabara Y, Murase K, Setoh K, Kawaguchi T, Nagashima S, et al. Nocturia and increase in nocturnal blood pressure: the Nagahama study. Journal of hypertension 2018; 36 (11):2185–2192. [DOI] [PubMed] [Google Scholar]
- 40.Zeitzer JM, Bliwise DL, Hernandez B, Friedman L, Yesavage JA. Nocturia compounds nocturnal wakefulness in older individuals with insomnia. Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine 2013; 9 (3):259–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ayas NT, White DP, Manson JE, Stampfer MJ, Speizer FE, Malhotra A, et al. A prospective study of sleep duration and coronary heart disease in women. Archives of internal medicine 2003; 163 (2):205–209. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.