Skip main navigation

Decreased Slow Wave Sleep Increases Risk of Developing Hypertension in Elderly Men

Originally publishedhttps://doi.org/10.1161/HYPERTENSIONAHA.111.174409Hypertension. 2011;58:596–603

Abstract

The importance of sleep to health and cardiovascular disease has become increasingly apparent. Sleep-disordered breathing, sleep duration, and sleep architecture may all influence metabolism and neurohormonal systems, yet no previous study has evaluated these sleep characteristics concurrently in relation to incident hypertension. Our objective was to determine whether incident hypertension is associated with polysomnography measures of sleep-disordered breathing, sleep duration, and sleep architecture in older men. Participants were 784 community-dwelling, ambulatory men ≥65 years of age (mean age: 75.1±4.9 years) from the Outcomes of Sleep Disorders in Older Men Study who did not have hypertension at the time of their in-home polysomnography sleep studies (2003–2005) and who returned for follow-up (2007–2009). Of 784 older men included in this report, 243 met criteria for incident hypertension after a mean follow-up of 3.4 years. In unadjusted analyses, incident hypertension was associated with increased hypoxemia, increased sleep stages N1 and N2, and decreased stage N3 (slow wave sleep [SWS]). After adjustment for age, nonwhite race, study site, and body mass index, the only sleep index to remain significantly associated with incident hypertension was SWS percentage (odds ratio for lowest to highest quartile of SWS: 1.83 [95% CI: 1.18 to 2.85]). No attenuation of this association was seen after accounting for sleep duration, sleep fragmentation, and indices of sleep-disordered breathing. Percentage time in SWS was inversely associated with incident hypertension, independent of sleep duration and fragmentation, and sleep-disordered breathing. Selective deprivation of SWS may contribute to adverse blood pressure in older men.

Introduction

The importance of sleep to health, including hypertension (HTN) and cardiovascular disease, continues to be elucidated. HTN is present in >30% of the US adult population, with a much higher risk in the elderly: >65% of Americans age >60 years have been diagnosed with HTN.1 Sleep disturbances and disorders are also exceedingly common in older adults. Sleep disorders such as sleep-disordered breathing (SDB), sleep duration, and sleep architecture may affect neurohormonal axes, including the sympathetic nervous system, which contribute to elevated blood pressure and HTN. To our knowledge, a comprehensive evaluation of sleep characteristics and incident HTN in an elderly cohort has not been reported.

SDB, which includes obstructive sleep apnea, is strongly associated with HTN.2 In a study of veterans, 60% of obstructive sleep apnea patients had HTN,3 which is often refractory to treatment.4 Conversely, OSA may be present in ≥30% of adults with HTN.5 Previous studies have suggested that SDB, as determined by the respiratory disturbance index (RDI; number of respiratory events per hour of sleep), precedes incident HTN,6 although these associations may be partly attributable to obesity.7

Epidemiological studies of self-reported sleep have also found that sleep deprivation and/or short sleep duration are associated with both prevalent and incident HTN8,9 and that such associations may be most pronounced among premenopausal women.10 A case-control polysomnographic study showed that severity of HTN was associated with short sleep duration, lower sleep efficiency, and less rapid eye movement (REM) sleep.11 Sleep disruption may also lead to increased blood pressure.12

Human sleep is divided into REM and non-REM (NREM) sleep. NREM is further divided into the stages of N1 (previously called stage 1), N2 (previously stage 2), and N3 (previously stage 3 and 4; also called slow wave sleep [SWS]).13 Both NREM and REM sleep are assessed by the brain wave activity (frequency and amplitude) recorded with polysomnography (PSG). SWS is considered to be “restorative” and is the sleep stage associated with the highest arousal threshold. The importance of SWS continues to be elucidated; it has been implicated in memory14 and overnight improvements in perceptual and visuomotor performance15 and learning.16 SWS has also been associated with transient metabolic, hormonal, and physiological changes, which affect glucose metabolism, and is associated with decreased sympathetic nervous activity and increased vagal tone.17 These, in turn, are associated with decreased heart rate and blood pressure, which may influence nocturnal blood pressure profiles. The disappearance of a nocturnal “dipping” blood pressure pattern is known to increase the risk of HTN and cardiovascular disease.18

We recently reported from the Outcomes of Sleep Disorders in Older Men Study (Osteoporotic Fractures in Men [MrOS] Sleep Study) that decreased percentage of time in SWS assessed by PSG was related to increased odds of obesity.19 Obesity and high blood pressure are very highly associated conditions with important public health consequences as major risk factors for both cardiovascular disease and death.20 To our knowledge, assessment of the association between SWS or other markers of sleep architecture with HTN has not yet been reported in prospective epidemiological studies.

The aim of the present study was to determine whether sleep characteristics, including SDB, determined by RDI or hypoxemia, decreased sleep duration, or indices of sleep architecture (which include overall arousal index, time in each sleep stage, sleep efficiency, or sleep fragmentation), predict incident HTN among elderly community-dwelling men.

Methods

Study Subjects

Study subjects were participants in the MrOS Sleep Study conducted in 2003–2005, an ancillary study of the parent MrOS study, a cohort of 5994 community-dwelling men, aged ≥65 years, described previously.21,22 A total of 3135 participants in the MrOS Sleep Study had full in-home PSG. See the online Data Supplement at for details of MrOS participants who were enrolled and excluded.

For the present study, 2911 older men had PSG data, of whom 1986 (68.2%) had a history of HTN, were taking antihypertensive medications, or had an elevated systolic blood pressure (SBP) ≥140 mm Hg or diastolic blood pressure (DBP) ≥90 mm Hg at the initial sleep visit and were excluded from the current analyses; 49 participants had missing data for either history of HTN or SWS and were also excluded. Of the 876 remaining, 784 participants attended the follow-up visit (2007–2009) and had complete HTN data (blood pressure measurements, response to questionnaires, and antihypertensive medication data) and are included in this report. In general, older men who regularly used mechanical sleep devices were excluded from the MrOS Sleep Study. However, our analysis sample included 1 man who reported intermittent use of a sleep device (but did not use the device on the night of the PSG measurement). The mean follow-up time was 3.40 years (SD 0.45 years). All of the protocols were approved by the institutional review boards at the respective enrollment sites, and participants signed informed consent to participate in the MrOS Sleep Study.

Classification and Measurement of HTN

As noted, HTN was defined as those who self-reported HTN (by affirmative answer to the question, “Have you ever had hypertension?”) or use of antihypertensive medications or who had measured SBP ≥140 mm Hg or DBP ≥90 mm Hg. Pre-HTN was classified as an SBP 120 to <140 mm Hg or DBP 80 to <90 mm Hg.23 Two seated resting blood pressure measurements were performed at the time of the initial sleep study using a conventional mercury sphygmomanometer on the right arm and an appropriate cuff size. Participants rested with their feet flat on the floor for ≥5 minutes before measurements, which were averaged. SBP was determined by phase I and diastolic as phase V Korotkoff values. At the follow-up visit, seated blood pressures were measured using the same protocol, but a BP Tru automated blood pressure monitor (model BMP-300) was used because of prohibition of mercury sphygmomanometers for safety reasons.

Other Measures

Self-administered questionnaires to assess participant demographics, including race; lifestyle factors such as physical activity, depression, and alcohol and smoking; and classification of medication use are described in the online Data Supplement. Height (in centimeters) was measured on regularly calibrated Harpenden stadiometers and weight (in kilograms) on calibrated standard balance beam or digital scales. Body mass index (BMI) was calculated as kilograms per meter squared. Waist, hip, and neck circumferences were measured using a tape measure and standard techniques.24

Polysomnography

In-home, single-night sleep studies using unattended PSG (Safiro, Compumedics, Inc, Melbourne, Australia) were performed. See the online Data Supplement for more details. Sleep stages (REM and stages 1 to 4 NREM, now N1 to N3) were scored using the standard criteria at the time of the studies.25 SWS was defined as the total sleep time scored as stage 3 plus stage 4, now N3, and expressed as percentage of total sleep time. RDI, an indication of the severity of SDB, was defined as the number of apneas and hypopneas (based on the Sleep Heart Health Study criteria26) per hour of sleep, each associated with an oxygen desaturation of ≥4%. Hypoxemia was determined by sleep time spent with oxygen desaturation to <80% (Pao2 <80%). Central apnea index was defined as the number of apneas per hour of sleep with no displacement of either chest or abdominal inductance channels, regardless of oxygen desaturation. Total sleep duration at night and sleep efficiency, defined as the percentage of the sleep period (time in bed) spent asleep, were also obtained from PSG. Overall arousal index was determined as number of arousals per hour of sleep. Wake after sleep onset (the minutes awake after having been asleep) was used as a marker of sleep fragmentation.

Statistical Analyses

Initial associations of sleep variables with incident HTN were examined using ANOVA, Kruskal-Wallis, or χ2 tests. Sleep variables considered included indices of SDB (by RDI and percentage of sleep time with oxygen saturation <80%), nighttime sleep duration, and sleep architecture (sleep stage distributions, wake after sleep onset, sleep efficiency, and overall arousal index). Descriptive and inferential statistics were performed using SAS version 9.2 (SAS Institute, Inc, Cary, NC). To account for multiple end points that are highly correlated, we used the Dubey and Armitage-Parmar procedure.27

Potential confounders were identified a priori. Summary statistics for variables believed to be related to the polysomnographic variables and/or incident HTN were performed. ANOVA, Kruskal-Wallis, or χ2 tests were used to examine associations with the sleep characteristics and incident HTN for normal, nonparametric, and categorized variables, respectively. Variables that were significant at P<0.10 for both outcomes were included in a multivariable model. A list of potential confounders and sleep variables considered are shown in Tables 1 and 2, respectively.

Table 1. Demographic Characteristics of the Study Cohort: Overall and by Follow-Up HTN Status

CharacteristicOverall (N=784)Incident HTN (N=243)No Incident HTN (N=541)P
Age, y75.1±4.975.7±5.074.9±4.80.032
BMI, kg/m226.4±3.426.7±3.726.3±3.30.088
Neck circumference, cm39.0±2.739.1±2.838.9±2.60.209
Waist circumference, cm97.3±10.098.3±10.396.9±9.80.060
Hip circumference, cm101.9±8.3102.1±8.0101.8±8.40.593
PASE score160.8±71.3159.6±72.3161.4±70.90.75
Baseline SBP, mm Hg121.0±10.6125.4±9.6119.0±10.4<0.0001
Baseline DBP, mm Hg67.5±7.768.7±7.566.9±7.80.003
Follow-up SBP, mm Hg118.7±16.3125.0±21.2116.0±12.7<0.0001
Follow-up DBP, mm Hg68.1±9.769.2±11.667.6±8.80.064
Nonwhite race, n (%)68 (8.7)20 (8.2)48 (8.9)0.768
Self-rating of overall health “excellent or good,” n (%)501 (92.8)222 (92.3)723 (92.2)0.19
Current smoker, n (%)22 (2.8) 4 (1.6)18 (3.3)0.188
Has ≥1 alcoholic drink per wk, n (%)*432 (55.4)127 (52.9)305 (56.5)0.355
Diabetes mellitus, n (%)45 (5.7)19 (7.8)26 (4.8)0.094
History of cardiovascular disease, n (%)154 (19.7)58 (24.1)96 (17.7)0.040
GDS score ≥634 (4.3) 9 (3.7)25 (4.6)0.560
Prehypertensive at baseline, n (%)477 (60.8)190 (78.2)287 (53.1)<0.0001

Data are mean±SD unless otherwise specified. HTN indicates hypertension; BMI, body mass index; PASE, physical activity score in the elderly; SBP, systolic blood pressure; DBP, diastolic blood pressure; GDS, geriatric depression scale. Prehypertension is defined as SBP 120 to 139 mm Hg or DBP 80 to 89 mm Hg. P value compares those with and without incident hypertension.

*Further categorization of alcoholic drinks per week 0, <1, 1 to 2, 3 to 5, 6 to 13, or ≥14 did not significantly change results.

P<0.05.

Table 2. Sleep Characteristics Of Study Cohort: Overall and by Follow-Up HTN Status

CharacteristicOverall (N=784)Incident HTN (N=243)No Incident HTN (N=541)P
Sleep-disordered breathing, mean±SD
    RDI10.0±12.010.8±13.59.7±11.30.256*
    Hypoxemia (% sleep time Pao2 <80%)0.05±0.30.09±0.40.03±0.20.025*
    Central apnea index1.5±4.51.7±5.61.4±4.00.835*
    RDI ≥15, n (%)164 (20.9)51 (21.0)113 (20.9)0.975
    RDI ≥5, n (%)423 (54.0)133 (54.7)290 (53.6)0.770
Sleep duration, mean±SD
    Total sleep duration, min364.1±66.4365.5±72.3363.5±63.70.715
Sleep architecture, mean±SD
    Overall arousal index22.9±11.323.1±12.322.8±10.80.724*
    Sleep efficiency, %75.9±11.774.9±12.876.4±11.20.116
    Wake after sleep onset, min106.3±63.0112.2±63.7103.6±62.60.079
    % time in stage 1 (N1) sleep6.5±4.27.2±5.36.2±3.60.013*
    % time in stage 2 (N2) sleep62.0±9.263.1±9.261.6±9.20.033
    % time in slow wave (N3) sleep11.2±8.59.8±8.111.8±8.50.002*
    % time in REM20.2±6.220.0±6.420.3±6.10.479

RDI indicates respiratory disturbance index; HTN, hypertension; REM, rapid eye movement sleep.

*Data show the use of Kruskal-Wallis statistic because of nonnormal distributions.

P<0.05.

To account for nonlinear associations, sleep variables were categorized as quartiles. Parameters with a high number of 0 values (central apnea index and hypoxemia [percentage of time with Pao2 <80%]) were expressed as present or absent. Logistic regression was used to determine the odds of incident HTN for each quartile of sleep exposure compared with the highest quartile. Unadjusted models were run first, followed by models that accounted for age and nonwhite race, and then study site and BMI were added.

In adjusted analyses, incident HTN was only associated with SWS. Additional models were developed to better assess the influence of SWS after considering other sleep exposures. This included adjustment for RDI, percentage of time with oxygen saturation Pao2 <80%, total sleep duration, sleep efficiency, and the overall arousal index. In a subgroup analysis (shown in the online Data Supplement), we determined whether the association between SWS and incident HTN persisted after excluding men who were prehypertensive at baseline.

Results

Baseline Characteristics and Univariate Analyses

The 784 men included in this report had a mean age of 75.1 years (SD 4.9 years) and mean BMI of 26.4 kg/m2 (SD 3.4 kg/m2); ≈90% were white. Demographic characteristics are shown in Table 1. The mean baseline SBP was 121.0 mm Hg (SD 10.6 mm Hg) and DBP was 67.5 mm Hg (SD 7.7 mm Hg). A majority of the participants (61%) were prehypertensive, classified as a SBP 120 to <140 mm Hg or DBP 80 to <90 mm Hg23; 19.7% reported a history of cardiovascular disease, and 5.7% reported a history of diabetes mellitus. Almost all (92.3%) rated their health as “excellent” or “good.”

At baseline, participants had a mean RDI of 10.0 (SD 12.0; median of 5.7), but 54.0% of men had an RDI >5, indicating at least mild SDB, shown with other sleep characteristics in Table 2. On average, the total sleep duration was 364 minutes (SD 66.4 minutes) or 6.1 hours. Average percentage of time in REM was 20.2% (SD 6.2%), stage 1 (N1) 6.5% (SD 4.2%), stage 2 (N2) 62.0% (SD 9.2%), and SWS (stages 3 and 4, N3) 11.2% (SD 8.5%).

Incident HTN Was Associated With Sleep Architecture

Over the 3.4-year follow-up interval, 243 men developed incident HTN; 59 (24.3%) were diagnosed by blood pressure criteria alone, 9 (3.7%) by self-report of HTN alone, 5 (2.1%) had both self-reported HTN and met blood pressure criteria, and 170 (70%) were taking ≥1 antihypertensive medication. Those with incident HTN were older (75.7 versus 74.9 years; P=0.03) and were more likely to report cardiovascular disease (24.1% versus 17.7%; P=0.04), shown in Table 1.

There was a nonsignificant trend toward higher mean RDI with incident HTN, shown in Table 2. Percentage of time in sleep stages 1 (N1), 2 (N2), and SWS (N3) were associated with incident HTN, shown in Table 2. Those with incident HTN had poorer sleep architecture, as evidenced by significantly less SWS (mean: 9.8% versus 11.2%; P=0.002) and correspondingly more stage 1 (N1) and stage 2 (N2) sleep (P=0.013 and P=0.033, respectively). After adjusting for age, race, BMI, and study site, only the association between SWS and incident HTN remained statistically significant (Table S1, available in the online Data Supplement).

SWS and Incident HTN

To further explore the associations between SWS and incident HTN, we first assessed the associations of covariates and other sleep variables with quartiles of time in SWS (Table 3). The lowest quartile of percentage of sleep time in SWS (<4.1%) was associated with increasing age (P<0.001), BMI (P=0.015), and neck circumference (P=0.012) but not baseline SBP and DBP. In those in the lowest quartile of SWS, we observed a higher mean RDI (P<0.001), a higher mean overall arousal index (P<0.001), lower mean total sleep duration (P=0.009), and lower mean sleep efficiency (P<0.001). Those in the lowest quartile of percentage of sleep time in SWS were also more likely to have any SDB (RDI ≥5) or moderate-severe SDB (RDI ≥15; P=0.048 and P=0.0005, respectively). Central apnea index was not associated with quartile of SWS. Men taking either an angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker at follow-up (9.6% total) were more likely to be in the lowest quartile of SWS (P=0.034).

Table 3. Baseline Demographic Characteristics of Cohort, Overall and by Percent Time in Slow Wave Sleep (SWS) Quartiles (Q1–Q4)

CharacteristicOverall (N=784)Q1: <4.1 (N=195)Q2: 4.1 to 10.2 (N=197)Q3: 10.2 to 16.9 (N=192)Q4: >16.9 (N=200)P
Age, mean±SD, y75.1±4.976.5±5.574.6±4.974.6±4.374.8±4.50.0001
BMI, kg/m226.4±3.427.0±3.726.3±3.026.5±3.625.9±3.30.015
Neck circumference, cm39.0±2.739.4±2.638.9±2.638.9±2.738.5±2.70.012
Waist circumference, cm97.3±10.098.8±10.397.1±8.797.0±10.496.4±10.20.096
Hip circumference, cm101.9±8.3102.9±7.9101.6±7.2101.9±10.2101.1±7.50.162
Baseline SBP, mm Hg121.0±10.6121.8±10.5120.9±10.1120.9±10.9120.3±10.90.569
Baseline DBP, mm Hg67.5±7.767.3±7.867.7±7.567.5±7.867.5±7.90.962
Follow-up SBP, mm Hg118.7±16.3121.0±16.5118.2±14.8116.6±16.6118.9±16.80.061
Follow-up DBP, mm Hg68.1±9.768.1±9.868.6±9.567.0±9.568.4±10.00.371
PASE score160.8±71.3162.6±71.9152.5±69.6164.3±71.6163.9±72.20.305
RDI10.0±12.013.3±15.39.8±10.99.4±11.67.7±8.80.006
Overall arousal index22.9±11.326.5±13.522.7±11.121.8±9.520.6±9.8<0.0001
Central apnea index1.5±4.51.7±6.11.5±4.01.2±2.91.4±4.50.813
Sleep efficiency, %75.9±11.772.4±12.576.1±11.676.3±11.978.7±10.1<0.0001
Total sleep duration, min364.2±66.4350.2±77.0370.0±58.7368.6±66.8367.7±60.40.009
Wake after sleep onset, min106.3±63.1122.2±64.0109.5±70.6102.1±57.591.6±55.4<0.0001
Time in SWS, min41.4±31.86.0±5.025.7±7.849.4±11.283.7±20.7<0.0001
Nonwhite race, n (%)68 (8.7)22 (11.3)16 (8.1)22 (11.5) 8 (4)0.027
Current smoker, n (%)22 (2.8) 9 (4.6) 4 (2.0) 1 (0.5) 8 (4)0.058
Has ≥1 alcoholic drink per wk, n (%)*432 (55.4)104 (53.9)108 (55.4)105 (54.7)115 (57.5)0.902
Use of sleep medication, n (%)72 (9.2)19 (9.7)21 (10.7)18 (9.4)14 (7)0.628
Diabetes mellitus, n (%)45 (5.7)13 (6.7) 9 (4.6)13 (6.8)10 (5.0)0.708
History of cardiovascular disease, n (%)154 (19.7)39 (20.1)40 (20.3)33 (17.3)42 (21.0)0.80
Prehypertensive at baseline, n (%)477 (60.8)127 (65.1)118 (59.9)112 (58.3)120 (60.0)0.543
Self-rating of health: excellent or good, n (%)723 (932.3)188 (96.4)183 (92.9)168 (88.0)184 (92.0)0.0197
GDS score ≥6, n (%)34 (4.3) 9 (4.6) 6 (3.0) 8 (4.17)11 (5.5)0.684
RDI ≥5, n (%)423 (54.0)119 (61.0)111 (56.3)94 (49.0)99 (49.5)0.048
Sleep time Pao2 <80%, n (%)0.05±0.30.1±0.50.02±0.150.04±0.210.04±0.240.013
ACEI/ARB use at follow-up, n (%)75 (9.6)29 (14.9)14 (7.1)15 (7.8)17 (8.5)0.034
Diuretic use at follow-up, n (%)48 (6.1)16 (8.2) 9 (4.6)16 (8.3) 7 (3.5)0.097
β-Blocker use at follow-up, n (%)79 (10.1)21 (10.8)17 (8.6)23 (12.0)18 (9)0.66
Calcium channel blocker use at follow-up, n (%)17 (2.2) 5 (2.6) 6 (3.1) 2 (1.0) 4 (2.0)0.57

BMI indicates body mass index; PASE, physical activity score in the elderly; REM, rapid eye movement sleep; SBP, systolic blood pressure; DBP, diastolic blood pressure; RDI, respiratory disturbance index; GDS, geriatric depression scale; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker.

*Further categorization of alcoholic drinks per week 0, <1, 1 to 2, 3 to 5, 6 to 13, or ≥14 did not significantly change results.

Data show the use of Kruskal-Wallis statistic because of nonnormal distributions.

P<0.05.

In unadjusted analyses, 41% of those in quartile 1 of SWS (Q1; least SWS) developed incident HTN compared with <30% in each of the other quartiles (Figure). We further examined the association of SWS and incident HTN in multivariate models (Table 4). After adjusting for age and BMI, those in Q1 of SWS experienced an ≈80% increase in risk of incident HTN as compared with those in Q4 (odds ratio: 1.81 [95% CI: 1.18 to 2.80]). These results were unchanged after further adjustment for study site, race, history of cardiovascular disease, the polysomnographic sleep variables of overall arousal index, sleep time with Pao2 <80% saturation, RDI, sleep efficiency, total sleep duration, and smoking history and alcohol use. After adjusting for multiple correlated comparisons using the Dubey and Armitage-Parmar procedure,27 results were not significantly changed. In addition, substituting height and weight, waist circumference, hip circumference, or hip:waist ratio for BMI did not significantly change the results.

Figure.

Figure. Percentage of participants in each quartile of slow wave sleep (SWS) who developed incident hypertension.

Table 4. Adjusted Odds Ratios for Incident Hypertension by Slow Wave Sleep Quartile

Adjusted CharacteristicsSWS Quartile 1SWS Quartile 2SWS Quartile 3SWS Quartile 4P for Trend
Age and BMI1.81 (1.18 to 2.80)1.09 (0.70 to 1.70)1.15 (0.74 to 1.80)1.0 (reference)0.012
+Race, study site1.83 (1.18 to 2.85)1.14 (0.73 to 1.78)1.18 (0.75 to 1.85)1.00.012
+History of CVD1.84 (1.18 to 2.87)1.14 (0.73 to 1.79)1.17 (0.75 to 1.84)1.00.012
+RDI1.86 (1.19 to 2.91)1.14 (0.73 to 1.80)1.18 (0.75 to 1.85)1.00.011
+Total sleep duration, % sleep time, Pao2 <80%, overall arousal index and sleep efficiency1.87 (1.18 to 2.95)1.12 (0.71 to 1.77)1.15 (0.73 to 1.82)1.00.013
+Alcohol use* and current smoking1.82 (1.15 to 2.88)1.11 (0.70 to 1.75)1.15 (0.73 to 1.82)1.00.019

BMI indicates body mass index; SWS, slow wave sleep; CVD, cardiovascular disease; RDI, respiratory disturbance index (No. of apneas and hypopneas per h of sleep, each associated with an oxygen desaturation of ≥4%).

*Alcohol use denotes ≥1 drink per wk. Further categorization of alcoholic drinks per week 0, <1, 1 to 2, 3 to 5, 6 to 13, or ≥14 did not significantly change results.

Subgroup analysis of sleep architecture in normotensive men who progressed to pre-HTN or HTN, described in the online Data Supplement and Figure S1, revealed an association with decreased percentage of time in SWS (P=0.004) and percentage of increased time in stage 2 (N2; P=0.042) sleep as subjects either stayed normotensive or progressed to pre-HTN or HTN.

Discussion

Sleep disturbances are exceedingly common in older adults and may contribute to a number of adverse health outcomes, including HTN and cardiovascular disease through perturbations in a number of physiological systems. Most previous research on the relationship between incident HTN and sleep disturbances has focused on SDB, with some studies also addressing sleep duration. To our knowledge, this is the first large-scale analysis of comprehensive PSG data that assesses the association of indices of 3 key sleep domains, SDB, sleep duration, and sleep architecture, to incident HTN in an elderly cohort. In these community-dwelling older men, we found that men with SWS in the lowest quartile had an ≈1.8-fold increased incidence of HTN compared with men with the highest levels of SWS. This association persisted after adjusting for several covariates and after considering other sleep exposures, such as RDI, the overall arousal index, and total sleep duration. This study adds to the growing important body of literature regarding sleep and cardiovascular risks. Rather than noting associations alone, it describes more detailed associations of incident disease with sleep in particular stages.

Decreased SWS is increasingly recognized as a marker, possibly etiologically associated with several adverse health outcomes. Sleep architecture, particularly SWS, has been implicated in neurocognition1416 and, more recently, in influencing endocrine function.17 Human metabolism is affected by changes in circadian rhythm, which includes the sleep-wake cycle.28 Low sleep quality and reduced SWS have been associated with insulin resistance17 and associated with the presence of diabetes mellitus,29 findings that may relate to the close interactions between the hypothalamic-pituitary-adrenal axis and sleep homeostasis. SWS may potentially be associated with the metabolic syndrome, a notion consistent with our previous observations of an inverse association of SWS with obesity19 in this cohort. The findings in this study provide new evidence that decreased SWS is associated with the development of HTN. Whether decreased SWS is also associated with altered glucose metabolism or with other components of metabolic syndrome could not be assessed because of the unavailability of biochemical measures. A limitation of this study is the lack of fasting glucose or insulin or hemoglobin A1C measurements to investigate these relationships further. However, this sample included only a few diabetics (5.7%), and adjusting for diabetes mellitus history did not attenuate the association.

Our results are consistent with a positive effect of SWS on blood pressure regulation. Studies of cardiac hemodynamics during sleep have noted decreased blood pressure and heart rate during NREM, particularly SWS,30 suggesting that the decline in blood pressure is attributable to a decline in cardiac output, because peripheral vascular resistance and stroke volume are unaffected. These adaptations are primarily mediated by the autonomic nervous system with decreased sympathetic nervous system activity and increased parasympathetic activity.30 Sympathetic nervous system activity continuously decreases with the deepening of NREM toward SWS, coinciding with the increase of the sympathovagal baroreflex. This baroreflex may account for the benefits of “dipping” of blood pressure on cardiovascular disease, which is impaired in hypertensive subjects who have increased sympathetic nervous system activity.31 This “nondipping” nocturnal blood pressure is a better predictor of all-cause and cardiovascular mortality, coronary heart disease, and stroke than daytime pressure.18 Experimental SWS deprivation in healthy subjects resulted in an attenuation of the dipping of blood pressure during SWS.32 Several studies have demonstrated that SWS is lower in older men than women and declines with age.33 Our data suggest that reduced SWS with aging may contribute to the high rates of HTN in older men. Our study found no independent associations between incident HTN and RDI (as a marker of SDB), arousal index (a marker for both SDB and disturbed sleep), and total sleep duration.

Two previous studies have evaluated whether SDB predicts incident HTN. First, the Wisconsin Sleep Cohort Study, with a younger mean age of 47 years, reported a dose-response association between SDB at baseline and incident HTN 4 years later, which was independent of known confounding factors.6 The Sleep Heart Health Study, with a mean age of 60 years, found that much of the longitudinal association was accounted for by obesity.7 The disparate results may be because of the difference in mean ages or exclusions, given that the incidence of HTN increases with age,1 and at our baseline sleep visit, 68% of the subjects were excluded at baseline because of prevalent HTN.

Consistent with some previous studies, total sleep duration measured by PSG was not associated with incident HTN in this older cohort.34,35 The literature on incident HTN and sleep duration examined this association in younger populations8,36 or with use of self-reported sleep duration.9 Our study suggests differences across the age span in the associations between sleep characteristics and health outcomes.

Poor sleep quality may be measured by several different metrics, including an increased arousal index and increased wake after sleep onset, decreased sleep efficiency, and increased stages 1 and 2 (N1 and N2) sleep. Previous studies have focused on the arousal index as a metric for increased sympathetic activation and as a potential predictor of HTN. Our study suggests that a specific reduction of SWS, rather than nonspecific sleep disruption, may be a more critical factor that influences blood pressure.

The strengths of this study include the use of PSG for objectively measuring sleep characteristics and assessment in a large community-dwelling cohort of older men. However, given that the parent study (MrOS) was originally designed to evaluate the risks of fractures in older men, this study must be considered exploratory in nature. We acknowledge that this observational study is challenged by many factors that influence both blood pressure and sleep, which also change over time. Additional study to confirm this association is necessary to determine the direction of the association after longer follow-up. Other limitations include the lack of 24-hour ambulatory blood pressure monitoring for HTN and the precise time of the blood pressure readings. Almost a quarter of those with incident HTN (24.3%) were classified based on a single blood pressure measurement; however, sensitivity analysis that excluded these participants showed similar results (data not shown).

In an exploratory study, we evaluated 11 sleep variables (Table s1) that raise questions about multiple comparisons. Because many of these indices are highly correlated (Pearson coefficients range from −0.86 to 0.64), we used the Dubey and Armitage-Parmar procedure27 to address the influence of testing for multiple correlated outcomes, and the results were not appreciably changed. These results in older men may not be generalizable to women or younger adults. It is uncertain whether there is a true association between the use of angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker and SWS (Table 3), but <10% of participants were taking these medications at follow-up (and none at baseline). Residual confounding because of other medications, activity, or comorbid conditions was not captured in this study. Also, use of PSG to determine sleep duration may not be optimal, because PSG monitoring, with possible disruptions of leads and electrodes, may not represent typical sleep patterns. However, previous work has shown that SWS has high night-to-night reproducibility.26

Perspectives

This prospective study suggests that older community-dwelling men with a lower percentage of sleep time in SWS have an increased risk of incident HTN. In contrast, indices of breathing disturbances, level of hypoxemia, sleep duration, and arousal index were not independently associated with an increased risk of HTN after considering confounders. This article adds to the growing body of literature that associates sleep architecture with metabolic and physiological changes that may reflect altered neurohormones and inflammatory markers. Further studies are necessary to confirm these observations, elucidate the causal pathways, and determine whether modifications in SWS improve HTN.

Sources of Funding

The Osteoporotic Fractures in Men Study is supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institute on Aging, National Center for Research Resources, and National Institutes of Health Roadmap for Medical Research under the following grants: U01 AR45580, U01 AR45614, U01 AR45632, U01 AR45647, U01 AR45654, U01 AR45583, U01 AG18197, U01-AG027810, and UL1 RR024140. This work was also supported by the Veterans Affairs Healthcare System (to M.M.F.) and AG08415 (to S.A.-I.). The authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The National Heart, Lung, and Blood Insitute provides funding for the Osteoporotic Fractures in Men Study sleep ancillary study “Outcomes of Sleep Disorders in Older Men” under the following grants: R01 HL071194, R01 HL070848, R01HL070847, R01 HL070842, R01 HL070841, R01 HL070837, R01 HL070838, and R01 HL070839.

Disclosures

M.M.F. discloses previous grant support from Forest Laboratories and currently owns stock and is employed by Amgen. S.A.I. is a consultant for Ferring Pharmaceuticals, Inc, GlaxoSmithKline, Johnson & Johnson, Merck, NeuroVigil, Inc, Pfizer, Philips, Purdue Pharma LP, sanofi-aventis, and Somaxon.

Footnotes

Correspondence to Maple M. Fung,
VA San Diego Healthcare System, 3350 La Jolla Village Dr, San Diego, CA 92161
(E-mail ); or Katie L. Stone,
San Francisco Coordinating Center, California Pacific Medical Center Research Institute, 185 Berry St, Lobby 5, 5th Floor, Suite 5700, San Francisco, CA 94107
(E-mail ).

References

  • 1. Ong KL, Cheung BM, Man YB, Lau CP, Lam KS. Prevalence, awareness, treatment, and control of hypertension among United States adults 1999–2004. Hypertension. 2007; 49: 69– 75. LinkGoogle Scholar
  • 2. Nieto FJ, Young TB, Lind BK, Shahar E, Samet JM, Redline S, D'Agostino RB, Newman AB, Lebowitz MD, Pickering TG. Association of sleep-disordered breathing, sleep apnea, and hypertension in a large community-based study: Sleep Heart Health Study. JAMA. 2000; 283: 1829– 1836. CrossrefMedlineGoogle Scholar
  • 3. Sharafkhaneh A, Richardson P, Hirshkowitz M. Sleep apnea in a high risk population: a study of Veterans Health Administration beneficiaries. Sleep Med. 2004; 5: 345– 350. CrossrefMedlineGoogle Scholar
  • 4. Logan AG, Perlikowski SM, Mente A, Tisler A, Tkacova R, Niroumand M, Leung RS, Bradley TD. High prevalence of unrecognized sleep apnoea in drug-resistant hypertension. J Hypertens. 2001; 19: 2271– 2277. CrossrefMedlineGoogle Scholar
  • 5. Friedman O, Logan AG. The price of obstructive sleep apnea-hypopnea: hypertension and other ill effects. Am J Hypertens. 2009; 22: 474– 483. CrossrefMedlineGoogle Scholar
  • 6. Peppard PE, Young T, Palta M, Skatrud J. Prospective study of the association between sleep-disordered breathing and hypertension. N Engl J Med. 2000; 342: 1378– 1384. CrossrefMedlineGoogle Scholar
  • 7. O'Connor GT, Caffo B, Newman AB, Quan SF, Rapoport DM, Redline S, Resnick HE, Samet J, Shahar E. Prospective study of sleep-disordered breathing and hypertension: the Sleep Heart Health Study. Am J Respir Crit Care Med. 2009; 179: 1159– 1164. CrossrefMedlineGoogle Scholar
  • 8. Knutson KL, Van Cauter E, Rathouz PJ, Yan LL, Hulley SB, Liu K, Lauderdale DS. Association between sleep and blood pressure in midlife: the CARDIA sleep study. Arch Intern Med. 2009; 169: 1055– 1061. CrossrefMedlineGoogle Scholar
  • 9. Gottlieb DJ, Redline S, Nieto FJ, Baldwin CM, Newman AB, Resnick HE, Punjabi NM. Association of usual sleep duration with hypertension: the Sleep Heart Health Study. Sleep. 2006; 29: 1009– 1014. CrossrefMedlineGoogle Scholar
  • 10. Stranges S, Dorn JM, Cappuccio FP, Donahue RP, Rafalson LB, Hovey KM, Freudenheim JL, Kandala NB, Miller MA, Trevisan M. A population-based study of reduced sleep duration and hypertension: the strongest association may be in premenopausal women. J Hypertens. 2010; 28: 896– 902. CrossrefMedlineGoogle Scholar
  • 11. Friedman O, Bradley TD, Ruttanaumpawan P, Logan AG. Independent association of drug-resistant hypertension to reduced sleep duration and efficiency. Am J Hypertens. 2010; 23: 174– 179. CrossrefMedlineGoogle Scholar
  • 12. McCubbin JA, Pilcher JJ, Moore DD. Blood pressure increases during a simulated night shift in persons at risk for hypertension. Int J Behav Med. 2010; 17: 314– 320. CrossrefMedlineGoogle Scholar
  • 13. Silber MH, Ancoli-Israel S, Bonnet MH, Chokroverty S, Grigg-Damberger MM, Hirshkowitz M, Kapen S, Keenan SA, Kryger MH, Penzel T, Pressman MR, Iber C. The visual scoring of sleep in adults. J Clin Sleep Med. 2007; 3: 121– 131. CrossrefMedlineGoogle Scholar
  • 14. Marshall L, Helgadottir H, Molle M, Born J. Boosting slow oscillations during sleep potentiates memory. Nature. 2006; 444: 610– 613. CrossrefMedlineGoogle Scholar
  • 15. Huber R, Ghilardi MF, Massimini M, Tononi G. Local sleep and learning. Nature. 2004; 430: 78– 81. CrossrefMedlineGoogle Scholar
  • 16. Landsness EC, Crupi D, Hulse BK, Peterson MJ, Huber R, Ansari H, Coen M, Cirelli C, Benca RM, Ghilardi MF, Tononi G. Sleep-dependent improvement in visuomotor learning: a causal role for slow waves. Sleep. 2009; 32: 1273– 1284. CrossrefMedlineGoogle Scholar
  • 17. Tasali E, Leproult R, Ehrmann DA, Van Cauter E. Slow-wave sleep and the risk of type 2 diabetes in humans. Proc Natl Acad Sci U S A. 2008; 105: 1044– 1049. CrossrefMedlineGoogle Scholar
  • 18. Fagard RH, Celis H, Thijs L, Staessen JA, Clement DL, De Buyzere ML, De Bacquer DA. Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension. Hypertension. 2008; 51: 55– 61. LinkGoogle Scholar
  • 19. Rao MN, Blackwell T, Redline S, Stefanick ML, Ancoli-Israel S, Stone KL. Association between sleep architecture and measures of body composition. Sleep. 2009; 32: 483– 490. CrossrefMedlineGoogle Scholar
  • 20. Narkiewicz K. Obesity and hypertension: the issue is more complex than we thought. Nephrol Dial Transplant. 2006; 21: 264– 267. CrossrefMedlineGoogle Scholar
  • 21. Orwoll E, Blank JB, Barrett-Connor E, Cauley J, Cummings S, Ensrud K, Lewis C, Cawthon PM, Marcus R, Marshall LM, McGowan J, Phipps K, Sherman S, Stefanick ML, Stone K. Design and baseline characteristics of the Osteoporotic Fractures in Men (MrOS) Study: a large observational study of the determinants of fracture in older men. Contemp Clin Trials. 2005; 26: 569– 585. CrossrefMedlineGoogle Scholar
  • 22. Blank JB, Cawthon PM, Carrion-Petersen ML, Harper L, Johnson JP, Mitson E, Delay RR. Overview of recruitment for the Osteoporotic Fractures in Men Study (MrOS). Contemp Clin Trials. 2005; 26: 557– 568. CrossrefMedlineGoogle Scholar
  • 23. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL, Jones DW, Materson BJ, Oparil S, Wright JT, Roccella EJ. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension. 2003; 42: 1206– 1252. LinkGoogle Scholar
  • 24. Callaway C, Chumlea W, Bouchard C. In , Lohman T, Roche A, Martorell R. eds. Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics; 1988: 39– 54. Google Scholar
  • 25. Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Los Angeles, CA: Brain Information Service, Brain Research Institute, University of California Los Angeles; 1973. Google Scholar
  • 26. Quan SF, Howard BV, Iber C, Kiley JP, Nieto FJ, O'Connor GT, Rapoport DM, Redline S, Robbins J, Samet JM, Wahl PW. The Sleep Heart Health Study: design, rationale, and methods. Sleep. 1997; 20: 1077– 1085. MedlineGoogle Scholar
  • 27. Sankoh AJ, Huque MF, Dubey SD. Some comments on frequently used multiple endpoint adjustment methods in clinical trials. Stat Med. 1997; 16: 2529– 2542. CrossrefMedlineGoogle Scholar
  • 28. Froy O. The circadian clock and metabolism. Clin Sci (Lond). 2011; 120: 65– 72. CrossrefMedlineGoogle Scholar
  • 29. Pallayova M, Donic V, Gresova S, Peregrim I, Tomori Z. Do differences in sleep architecture exist between persons with type 2 diabetes and nondiabetic controls?J Diabetes Sci Technol. 2010; 4: 344– 352. CrossrefMedlineGoogle Scholar
  • 30. Trinder J, Kleiman J, Carrington M, Smith S, Breen S, Tan N, Kim Y. Autonomic activity during human sleep as a function of time and sleep stage. J Sleep Res. 2001; 10: 253– 264. CrossrefMedlineGoogle Scholar
  • 31. Drager LF, Ueno LM, Lessa PS, Negrao CE, Lorenzi-Filho G, Krieger EM. Sleep-related changes in hemodynamic and autonomic regulation in human hypertension. J Hypertens. 2009; 27: 1655– 1663. CrossrefMedlineGoogle Scholar
  • 32. Sayk F, Teckentrup C, Becker C, Heutling D, Wellhoner P, Lehnert H, Dodt C. Effects of selective slow-wave sleep deprivation on nocturnal blood pressure dipping and daytime blood pressure regulation. Am J Physiol Regul Integr Comp Physiol. 2010; 298: R191– R197. CrossrefMedlineGoogle Scholar
  • 33. Redline S, Kirchner HL, Quan SF, Gottlieb DJ, Kapu r V, Newman A. The effects of age, sex, ethnicity, and sleep-disordered breathing on sleep architecture. Arch Intern Med. 2004; 164: 406– 418. CrossrefMedlineGoogle Scholar
  • 34. van den Berg JF, Tulen JH, Neven AK, Hofman A, Miedema HM, Witteman JC, Tiemeier H. Sleep duration and hypertension are not associated in the elderly. Hypertension. 2007; 50: 585– 589. LinkGoogle Scholar
  • 35. Lopez-Garcia E, Faubel R, Guallar-Castillon P, Leon-Munoz L, Banegas JR, Rodriguez-Artalejo F. Self-reported sleep duration and hypertension in older Spanish adults. J Am Geriatr Soc. 2009; 57: 663– 668. CrossrefMedlineGoogle Scholar
  • 36. Gangwisch JE, Heymsfield SB, Boden-Albala B, Buijs RM, Kreier F, Pickering TG, Rundle AG, Zammit GK, Malaspina D. Short sleep duration as a risk factor for hypertension: analyses of the first National Health and Nutrition Examination Survey. Hypertension. 2006; 47: 833– 839. LinkGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.