- Email: [email protected]
The Relationship Between Sleep Deprivation and the Nocturnal Decline of Blood Pressure
A]H
1996; 9:11.36-1138
The Relationship Between Sleep Deprivation and the Nocturnal Decline of Blood Pressure Steven J. Rosansky, Sudeep J. Menache~y, David Whittman, and Jason C. Rosenbe~g
Blood pressure (BP) has a circadian pattern with a 10% to 15% drop in systolic and diastolic BP comparing nighttime and daytime averages. The mechanism and the “reason” for the decline of nocturnal blood pressure has not been described. If the nocturnal decline is a restorative physiologic process we reasoned that sleep deprivation would increase the nocturnal drop. Thus, we tested the hypothesis that there is a compensatory decline in nocturnal sleep blood pressure after a period of nocturnal sleep deprivation. Twenty-four house staff personnel (normotensive and not taking any BP medications) with a mean age of 29.3 years were recruited for this study (13 men and 11 women). Subjects were randomly assigned to have 24-h BP monitoring after being sleep deprived (on call with 3 t 1.3 h of sleep) or after a normal night’s sleep (7.3 ~ 0.8
B
lood pressure (BP) has a circadian rhythm with a decline at night during sleep. ] Shift workers have a reversal of their circadian blood pressure pattern with a decline during daytime sleep and a rise during nighttime activities.z Activity is probably the main determinant of circadian blood pressure change. The physiologic mechanism responsible for the sleep related decline in blood pressure has not been well described.3
h). There was no significant difference in hours asleep or time to bed or time awake for both 24-h studies. Subjects had similar activities for both monitoring intervals. The percent change in day awake versus night asleep mean values for systolic BP, diastolic BP, and mean arterial pressure for sleep deprived and normal night’s sleep intervals were compared using paired t tests. None of these paired parameters were significantly different despite a subjectively “deeper” sleep postcall. In conclusion, sleep deprivation does not appear to result in a compensatory decline in nocturnal blood pressure. Thus, the “reason” for the normal nocturnal decline in blood pressure remains to be explored. Am J Hypertens 1996; 9:1136-1138 KEY WORDS: Blood pressure, nocturnal decline, circadian rhythms, sleep deprivation.
The current study tests the hypothesis that the decline of blood pressure during sleep is a normal physiologic restorative process. If this is the case then one would expect a greater decline of blood pressure during sleep after a night of sleep deprivation versus after a normal night’s sleep. In order to study this phenomenon we examined the circadian blood pressure pattern of medical house staff after being both on call and sleep deprived and also after a normal day’s work. METHODS
Received June 29, 1995. Accepted July 1, 1996. From the Medical Service, Dorn V.A. Hospital and Department of Medicine, University of South Carolina, School of Medicine, Cm lumbia, South Carolina. Address correspondence and reprint requests to S.J. Rosansky, MD, Chief, Renal Section (lllF), WJB Dorn V.A. Hospital, 6439 Garners Ferry Road, Columbia, SC 29209-1639.
01996 by the Amwica]z ]ouvnal of Hypertension, Published by Elscuier Science, Inc.
Ltd.
After signing informed consent, we enrolled 24 normotensive healthy house staff subjects into the study. SpaceLabs 90207 monitors (Redmond, WA) and A & D TM-2420 monitors (Tokyo, Japan) were used for two 24-h ambulatory blood pressure monitoring stud0895-7061/96/$15,00 PI1 S0895-7061(96)00300-7
A/H-NOVEMBER
1996-VOL.
9, NO. 11, PART 1
SLEEP DEPRIVATIONAND NOCTURNALBLOOD PRESSURE 1137
ies. Subjects use the same BP monitor each time. Monitoring occurred after sleep deprivation (the day following on call and after sleeping 4 h or less) and after a normal night’s sleep (at least 5 l/zh of sleep). Monitoring for each subject began and ended at 08:00. Systolic and diastolic blood pressure and heart rate were monitored every 15 min for the 24-h interval. Additionally, sleep diaries were acquired for each 24-h blood pressure monitoring interval. These diaries were kept to indicate time to bed and time awake and also to evaluate the quality of the night’s sleep as compared to a normal night’s sleep. Difficulty falling asleep (scored from 1 to 3, 1 = less difficult, 3 = more difficult), quality of sleep (scored from 1 to 3, 1 = poor, 3 = excellent), and type of sleep (scored from 1 to 3, 1 = light, 3 = deep) were the questions asked to determine the quality of sleep. Subjects did not smoke any cigarettes, take any prescription medicine, or drink any alcohol 24 h prior to entering the study. Data Analysis The percent change in hemodynamic parameters was calculated as follows: (day – night)/ day x 100, where day was 07:00 to 23:00 and night was 23:00 to 07:00. Sleep-deprived and non–sleep-deprived hemodynamic parameters were compared using twosided paired t tests. RESULTS Out of 34 subjects initially monitored, 24 completed both of the 24-h studies. Five subjects refused to be monitored the second time. Five subjects did not experience adequate sleep deprivation to be included in the study. The mean age of these 24 subjects was 29.3 years; mean height was 1.76 m, and mean weight 74.5 kg. Eight percent of the subjects were black and 92’ZO white. Fifty-four percent were male and 46’%ofemale. Eight subjects were fourth year medical students (subinterns), 12 were first year residents, and four were second year residents. The activities of subjects were similar on both days preceding ambulatory monitoring. Subjects worked an average of 10 h on a non-call day and 21 h on a call day. Prior to initiation of 24-h monitoring, subjects did not nap after returning home from work during each of the 24-h monitoring intervals. Subjects were monitored during a normal work day in the same clinical service for each 24-h monitoring interval. The SpaceLabs 90207 monitor was used both times in 20 of the 24 subjects. Subjects slept an average of 3.0 t 1.3 h during the on-call night and 7.3 t 0.8 h during the normal night’s sleep. As seen in Table 1, there was no significant difference in total hours slept, time to bed at night, time awake in the morning, difficulty falling asleep, quality of sleep, or total sleep score during the sleep interval precall or postcall. A significantly deeper sleep (type of sleep score: 2.59 v 2.05, P = .01) occurred postcall.
TABLE 1.PRECALL VERSUS POSTCALL SLEEP SCORE AND DAYTIME AND NIGHTTIME BLOOD PRESSURE SUMMARIES
Precall
Postcall
P
7.96 (1.57)
8.08 (1.02)
.75
23:05 (1.49)
23:13 (2.10)
.80
06:48 (1.22)
06:31 (1.06)
.34
2.00 (0.14) 1.95 (0.11)
1.73 (0.15) 2.00 (0.11)
.19 .76
morning~ Type of sleep ~
1.95 (0,11) 2.05 (0.16)
1.77 (0.11) 2.59 (0.13)
.26 .O1*
Total sleep score*’ Daytime summarytt Systolic BP, mm Hg Diastolic BP, mm Hg Mean arterial pressure, mm Hg Heart rate, beats/rein Nighttime summary++ Systolic BP, mm Hg Diastolic BP, mm Hg Mean arterial pressure, mm Hg Heart rate, beats/rein Percent changefl Systolic BP Diastolic BP Mean arterial pressure Heart rate
7.95 (0.37)
8.09 (0.23)
.75
123.3 (1.9) 77.1 (1.3)
122.6 (2.1) 76.5 (1.4)
.55 .52
91.9 (1.4) 77.1 (1.3)
91.5 (1.5) 75.8 (1.6)
.63 .31
108.7 (2.0) 61.7 (1.6)
107.4 (1.9) 61.5 (1.4)
.33 .81
77.3 (1.6) 63.5 (1.6)*
76.6 (1.5) 61.1 (1.5)*
.55 .03”
11.9 (1.0) 19.9 (1.6)
12.3 (1.0) 19.7 (1.4)
.73 .88
15.9 (1.2) 17.6 (1.4)
16.2 (1.1) 19.2 (1.4)
.85 .34
Totalhourssleptwhile monitored Timeto bed while monitored* Timeawakewhile monitored’ Difficultyfalling alseept Qualityof sleep~ Howrestedin the
Mean (? SE). N = 24. Precall is defined as the 24-h interval after a normal night’s sleep. Postcall is dejim,d as tlw 24-h itlterval after being on-call and sleep-deprived. Two-tailed P values are ~iven. * Time repotted in hours (00:00-23:59).
t.%ored from
1-3, 1 = less difficult, 3 = more dijficcdt.
+ Scored from 1-3, 1 = j700r,3 = excellent. I Scored from 1-3, 1 = less rested, 3 = morerested. $ Scored from 1-3, 1 = [ight, 3 = deep. *“ Mea~l of sleep diary questionnaire
ttDatJinze
results.
is defined as 0:700-23:00.
~~ Nighttime is defined as 23:00-07:00. f~ calculated
as (day - nigkt)/day X
100.
As seen in Table 1 there was no significant difference in mean 24-h daytime or nighttime blood pressures, monitored precall or postcall. A significantly lower heart rate (although not clinically important) was seen postcall (P = .03). Table 1 also compares the percent change in day/
1138
ROSANSKYETAL
night blood pressure values. No significant difference in day/ night percent change was seen for any of these hemodynamic variables (P >.30 for all). DISCUSSION The current study is the first attempt to evaluate the affect of sleep deprivation on circadian blood pressure in a normal population. Previous measurements of 24h ambulatory BP in medical staff revealed higher 24h BP while on call than during an average nonworking day.45 As in shift workers, the predicted normal circadian pattern for blood pressure was abolished with 24-h on-call responsibility. These studies confirmed the notion that activity is the predominant factor determining blood pressure during a 24-h interval. The effect of sleep deprivation on circadian blood pressure patterns has not been explored. Our study did not find the expected compensatory decline in nocturnal blood pressure after sleep deprivation. This may be due to several factors. Firstly, subjects may not have been adequately sleep deprived to see the hypothesized results. Secondly, without a sleep laboratory, we are unable to be sure if subjects slept norreally after a night on call. Thirdly, if subjects had daytime naps during the 24-h postcall interval, this would decrease the day /night difference postcall. The latter two problems are unlikely since all subjects kept the same sleep /wake interval for the two 24-h monitoring periods. Also, the longer period of sleep deprivation prior to postcall sleep versus precall sleep would make it unlikely that subjects slept less well
A]H-NOVEMBER
1996-VOL.
9, NO. 11, PART 1
postcall. As well, sleep diaries demonstrated a “deeper” sleep postcall (Table 1) versus precall. Before discarding the hypothesis of this study, one must consider the possibility of a type 2 error. This seems unlikely in view of the extremely close values for day /night parameters measured pre- and postcall (Table 1). Much more research is needed to explain the normal physiology of circadian blood pressure and to determine if humans benefit from the normal sleeprelated decline in blood m-essure. , REFERENCES 1, Mann S, Millar Craig MW, et al: Physical activity and the circadian rhythm of blood pressure. 57:291 s-294s.
Clin Sci 1979;
2. Chau NP, Mallion JM, de Gaudemaris R, et al: Twentyfour hour ambulatory
blood pressure in shift workers.
Circulation 1989;80:341–347.
3. Talan MI, Engel BT: Effect of sympathetic blockade on
diurnal variation of hemodynamic patterns. Am J Physiol 1989;256:R778–R785.
4. del Arco-Galan C, Suarez-Fernandez C, Gabriel-Sanchez
R: What happens to blood pressure when on-call? Am J Hypertens 1994;7:396-401.
5 ~,
Bevan EG, Findlay JG, Murray GD, McInnes GT: Twenty-four-hour blood pressure in junior medical staff. J R Coil Physicians Lend 1992;26:367–371. Pickering TG, Harshfield GA, Kleinert HD, et al: Blood pressure during normal daily activities, sleep, and exercise: comparison of values in normal and hypertensive subjects. JAMA 1982;247:992–996.
1996; 9:11.36-1138
The Relationship Between Sleep Deprivation and the Nocturnal Decline of Blood Pressure Steven J. Rosansky, Sudeep J. Menache~y, David Whittman, and Jason C. Rosenbe~g
Blood pressure (BP) has a circadian pattern with a 10% to 15% drop in systolic and diastolic BP comparing nighttime and daytime averages. The mechanism and the “reason” for the decline of nocturnal blood pressure has not been described. If the nocturnal decline is a restorative physiologic process we reasoned that sleep deprivation would increase the nocturnal drop. Thus, we tested the hypothesis that there is a compensatory decline in nocturnal sleep blood pressure after a period of nocturnal sleep deprivation. Twenty-four house staff personnel (normotensive and not taking any BP medications) with a mean age of 29.3 years were recruited for this study (13 men and 11 women). Subjects were randomly assigned to have 24-h BP monitoring after being sleep deprived (on call with 3 t 1.3 h of sleep) or after a normal night’s sleep (7.3 ~ 0.8
B
lood pressure (BP) has a circadian rhythm with a decline at night during sleep. ] Shift workers have a reversal of their circadian blood pressure pattern with a decline during daytime sleep and a rise during nighttime activities.z Activity is probably the main determinant of circadian blood pressure change. The physiologic mechanism responsible for the sleep related decline in blood pressure has not been well described.3
h). There was no significant difference in hours asleep or time to bed or time awake for both 24-h studies. Subjects had similar activities for both monitoring intervals. The percent change in day awake versus night asleep mean values for systolic BP, diastolic BP, and mean arterial pressure for sleep deprived and normal night’s sleep intervals were compared using paired t tests. None of these paired parameters were significantly different despite a subjectively “deeper” sleep postcall. In conclusion, sleep deprivation does not appear to result in a compensatory decline in nocturnal blood pressure. Thus, the “reason” for the normal nocturnal decline in blood pressure remains to be explored. Am J Hypertens 1996; 9:1136-1138 KEY WORDS: Blood pressure, nocturnal decline, circadian rhythms, sleep deprivation.
The current study tests the hypothesis that the decline of blood pressure during sleep is a normal physiologic restorative process. If this is the case then one would expect a greater decline of blood pressure during sleep after a night of sleep deprivation versus after a normal night’s sleep. In order to study this phenomenon we examined the circadian blood pressure pattern of medical house staff after being both on call and sleep deprived and also after a normal day’s work. METHODS
Received June 29, 1995. Accepted July 1, 1996. From the Medical Service, Dorn V.A. Hospital and Department of Medicine, University of South Carolina, School of Medicine, Cm lumbia, South Carolina. Address correspondence and reprint requests to S.J. Rosansky, MD, Chief, Renal Section (lllF), WJB Dorn V.A. Hospital, 6439 Garners Ferry Road, Columbia, SC 29209-1639.
01996 by the Amwica]z ]ouvnal of Hypertension, Published by Elscuier Science, Inc.
Ltd.
After signing informed consent, we enrolled 24 normotensive healthy house staff subjects into the study. SpaceLabs 90207 monitors (Redmond, WA) and A & D TM-2420 monitors (Tokyo, Japan) were used for two 24-h ambulatory blood pressure monitoring stud0895-7061/96/$15,00 PI1 S0895-7061(96)00300-7
A/H-NOVEMBER
1996-VOL.
9, NO. 11, PART 1
SLEEP DEPRIVATIONAND NOCTURNALBLOOD PRESSURE 1137
ies. Subjects use the same BP monitor each time. Monitoring occurred after sleep deprivation (the day following on call and after sleeping 4 h or less) and after a normal night’s sleep (at least 5 l/zh of sleep). Monitoring for each subject began and ended at 08:00. Systolic and diastolic blood pressure and heart rate were monitored every 15 min for the 24-h interval. Additionally, sleep diaries were acquired for each 24-h blood pressure monitoring interval. These diaries were kept to indicate time to bed and time awake and also to evaluate the quality of the night’s sleep as compared to a normal night’s sleep. Difficulty falling asleep (scored from 1 to 3, 1 = less difficult, 3 = more difficult), quality of sleep (scored from 1 to 3, 1 = poor, 3 = excellent), and type of sleep (scored from 1 to 3, 1 = light, 3 = deep) were the questions asked to determine the quality of sleep. Subjects did not smoke any cigarettes, take any prescription medicine, or drink any alcohol 24 h prior to entering the study. Data Analysis The percent change in hemodynamic parameters was calculated as follows: (day – night)/ day x 100, where day was 07:00 to 23:00 and night was 23:00 to 07:00. Sleep-deprived and non–sleep-deprived hemodynamic parameters were compared using twosided paired t tests. RESULTS Out of 34 subjects initially monitored, 24 completed both of the 24-h studies. Five subjects refused to be monitored the second time. Five subjects did not experience adequate sleep deprivation to be included in the study. The mean age of these 24 subjects was 29.3 years; mean height was 1.76 m, and mean weight 74.5 kg. Eight percent of the subjects were black and 92’ZO white. Fifty-four percent were male and 46’%ofemale. Eight subjects were fourth year medical students (subinterns), 12 were first year residents, and four were second year residents. The activities of subjects were similar on both days preceding ambulatory monitoring. Subjects worked an average of 10 h on a non-call day and 21 h on a call day. Prior to initiation of 24-h monitoring, subjects did not nap after returning home from work during each of the 24-h monitoring intervals. Subjects were monitored during a normal work day in the same clinical service for each 24-h monitoring interval. The SpaceLabs 90207 monitor was used both times in 20 of the 24 subjects. Subjects slept an average of 3.0 t 1.3 h during the on-call night and 7.3 t 0.8 h during the normal night’s sleep. As seen in Table 1, there was no significant difference in total hours slept, time to bed at night, time awake in the morning, difficulty falling asleep, quality of sleep, or total sleep score during the sleep interval precall or postcall. A significantly deeper sleep (type of sleep score: 2.59 v 2.05, P = .01) occurred postcall.
TABLE 1.PRECALL VERSUS POSTCALL SLEEP SCORE AND DAYTIME AND NIGHTTIME BLOOD PRESSURE SUMMARIES
Precall
Postcall
P
7.96 (1.57)
8.08 (1.02)
.75
23:05 (1.49)
23:13 (2.10)
.80
06:48 (1.22)
06:31 (1.06)
.34
2.00 (0.14) 1.95 (0.11)
1.73 (0.15) 2.00 (0.11)
.19 .76
morning~ Type of sleep ~
1.95 (0,11) 2.05 (0.16)
1.77 (0.11) 2.59 (0.13)
.26 .O1*
Total sleep score*’ Daytime summarytt Systolic BP, mm Hg Diastolic BP, mm Hg Mean arterial pressure, mm Hg Heart rate, beats/rein Nighttime summary++ Systolic BP, mm Hg Diastolic BP, mm Hg Mean arterial pressure, mm Hg Heart rate, beats/rein Percent changefl Systolic BP Diastolic BP Mean arterial pressure Heart rate
7.95 (0.37)
8.09 (0.23)
.75
123.3 (1.9) 77.1 (1.3)
122.6 (2.1) 76.5 (1.4)
.55 .52
91.9 (1.4) 77.1 (1.3)
91.5 (1.5) 75.8 (1.6)
.63 .31
108.7 (2.0) 61.7 (1.6)
107.4 (1.9) 61.5 (1.4)
.33 .81
77.3 (1.6) 63.5 (1.6)*
76.6 (1.5) 61.1 (1.5)*
.55 .03”
11.9 (1.0) 19.9 (1.6)
12.3 (1.0) 19.7 (1.4)
.73 .88
15.9 (1.2) 17.6 (1.4)
16.2 (1.1) 19.2 (1.4)
.85 .34
Totalhourssleptwhile monitored Timeto bed while monitored* Timeawakewhile monitored’ Difficultyfalling alseept Qualityof sleep~ Howrestedin the
Mean (? SE). N = 24. Precall is defined as the 24-h interval after a normal night’s sleep. Postcall is dejim,d as tlw 24-h itlterval after being on-call and sleep-deprived. Two-tailed P values are ~iven. * Time repotted in hours (00:00-23:59).
t.%ored from
1-3, 1 = less difficult, 3 = more dijficcdt.
+ Scored from 1-3, 1 = j700r,3 = excellent. I Scored from 1-3, 1 = less rested, 3 = morerested. $ Scored from 1-3, 1 = [ight, 3 = deep. *“ Mea~l of sleep diary questionnaire
ttDatJinze
results.
is defined as 0:700-23:00.
~~ Nighttime is defined as 23:00-07:00. f~ calculated
as (day - nigkt)/day X
100.
As seen in Table 1 there was no significant difference in mean 24-h daytime or nighttime blood pressures, monitored precall or postcall. A significantly lower heart rate (although not clinically important) was seen postcall (P = .03). Table 1 also compares the percent change in day/
1138
ROSANSKYETAL
night blood pressure values. No significant difference in day/ night percent change was seen for any of these hemodynamic variables (P >.30 for all). DISCUSSION The current study is the first attempt to evaluate the affect of sleep deprivation on circadian blood pressure in a normal population. Previous measurements of 24h ambulatory BP in medical staff revealed higher 24h BP while on call than during an average nonworking day.45 As in shift workers, the predicted normal circadian pattern for blood pressure was abolished with 24-h on-call responsibility. These studies confirmed the notion that activity is the predominant factor determining blood pressure during a 24-h interval. The effect of sleep deprivation on circadian blood pressure patterns has not been explored. Our study did not find the expected compensatory decline in nocturnal blood pressure after sleep deprivation. This may be due to several factors. Firstly, subjects may not have been adequately sleep deprived to see the hypothesized results. Secondly, without a sleep laboratory, we are unable to be sure if subjects slept norreally after a night on call. Thirdly, if subjects had daytime naps during the 24-h postcall interval, this would decrease the day /night difference postcall. The latter two problems are unlikely since all subjects kept the same sleep /wake interval for the two 24-h monitoring periods. Also, the longer period of sleep deprivation prior to postcall sleep versus precall sleep would make it unlikely that subjects slept less well
A]H-NOVEMBER
1996-VOL.
9, NO. 11, PART 1
postcall. As well, sleep diaries demonstrated a “deeper” sleep postcall (Table 1) versus precall. Before discarding the hypothesis of this study, one must consider the possibility of a type 2 error. This seems unlikely in view of the extremely close values for day /night parameters measured pre- and postcall (Table 1). Much more research is needed to explain the normal physiology of circadian blood pressure and to determine if humans benefit from the normal sleeprelated decline in blood m-essure. , REFERENCES 1, Mann S, Millar Craig MW, et al: Physical activity and the circadian rhythm of blood pressure. 57:291 s-294s.
Clin Sci 1979;
2. Chau NP, Mallion JM, de Gaudemaris R, et al: Twentyfour hour ambulatory
blood pressure in shift workers.
Circulation 1989;80:341–347.
3. Talan MI, Engel BT: Effect of sympathetic blockade on
diurnal variation of hemodynamic patterns. Am J Physiol 1989;256:R778–R785.
4. del Arco-Galan C, Suarez-Fernandez C, Gabriel-Sanchez
R: What happens to blood pressure when on-call? Am J Hypertens 1994;7:396-401.
5 ~,
Bevan EG, Findlay JG, Murray GD, McInnes GT: Twenty-four-hour blood pressure in junior medical staff. J R Coil Physicians Lend 1992;26:367–371. Pickering TG, Harshfield GA, Kleinert HD, et al: Blood pressure during normal daily activities, sleep, and exercise: comparison of values in normal and hypertensive subjects. JAMA 1982;247:992–996.