Physiology&Behavior.Vol. 52, pp. 397-400, 1992
0031-9384/92 $5.00 + .00 Copyright© 1992PergamonPressLtd.
Printedin the USA.
BRIEF COMMUNICATION
Relationships Between Epinephrine, Waking Time, and Self-Estimates for Naps at Night K Y O K O N I S H I H A R A *l A N D K A Z U K O M O R I t
*Department of Psychophysiology, Tokyo Institute of Psychiatry, Tokyo, Japan, and ¢Division of Work Physiology and Psychology, The Institute for Science of Labour, Kawasaki, Japan Received 16 September 1991 NISHIHARA, K. AND K. MORI. Relationshipsbetweenepinephrine, wakingtime, and self-estimatesfor napsat night. PHYSIOL BEHAV 52(2) 397-400, 1992.--!n a previous study, we established a basal regression line between epinephrine excretion and percent of waking time in bed-rested humans under conditions involving minimal stress. We suggested that this regression line might be clinically useful for observing insomnia. The present study was undertaken to examine whether epinephrine excretion during short-duration sleep, such as naps, is influenced by a transient insomniac state. Polysomnograms during a 2-hour nap from 0000 to 0200 were recorded three times for six healthy male subjects. Epinephrine excretion levelsduring sleep were found to increase along the basal regressionline when percent of waking time went up. The subjects with over 30% waking time during this 2-hour period reported that their sleep was light and not satisfactoryat all. These resultssupport our position that the regression line can be useful for observing a transient insomniac state. Epinephrine excretion
Waking time
Self-estimates
Naps
THE present study is part of a series of studies investigating whether or not epinephrine levels in urine during sleep can be used as an effective and objective indicator of insomnia. In 1985, we found a strong positive correlation between epinephrine excretion and percent of waking time during time allotted for sleep (3). In 1988, we established a basal regression line between these two parameters (the E-W regression line) in bed-rested humans under conditions involving minimal stress (4). We then confirmed that epinephrine excretion levels during 8 hours of usual nocturnal sleep were distributed along the E-W regression line in younger and older volunteers (5). In the present study, levels of epinephrine excreted during nighttime naps were examined to see whether the E-W regression line could be used to observe a transient insomniac state even in short-duration sleep. In addition, we checked subjective sleep estimates. METHOD Six healthy male university students (mean age: 22.3 years) volunteered for the experiment. All had previously participated in one of the authors' studies. The experiment was conducted three times for each subject. The subjects were instructed not to engage in any strenuous physical exercise on the days of the experiment. They came to the laboratory at 2000, and urinated.
Transient insomniac state
After electrodes were attached, the subjects relaxed by listening to music, reading magazines, and talking to each other. Urine samples were collected before going to bed at 0000, and after getting up at 0200. Before going to bed, they were told they would be awakened at 0200. After getting up, they were asked to estimate their sleep on a four-item six-point bipolar scale. For this study, we recorded negative feelings when the subjects reported long awake time, frequent awakenings, light sleep, and not enough sleep. Electroencephalograms (F3-A~, C3-At, Oi-Al), electromyograms, electro-oculograms, and electrocardiograms were continuously recorded using a Medilog 9000 cassette tape recorder. Polysomnograms were replayed to Mingograph. Sleep stages were scored by an epoch of I min according to the criteria of Rechtschaffen and Kales (6). Percent of waking time was calculated as the amount of total waking time divided by the amount of time in bed (2 h). The amount of free epinephrine in the urine samples was measured by high performance liquid chromatography with fluorometric detection using the trihydroxyindole method (2). RESULTS Figure 1 shows epinephrine levels and percent of waking time during the 2-hour nap. All subjects' data were distributed along
t Requests for reprints should be addressed to K. Nishihara, Department of Psychophysiology,Tokyo Institute of Psychiatry, 2-1-8, Kamikitazawa, Setagaya-ku, Tokyo 156, Japan.
397
398
NISHIHARA AND MaRl
8 n~i n 6 =
4
B. uJ
2
0
i0 4"o g0 8'o(%)160 Percent of waking time
FIG. 1. Relationship between epinephrine excretion and percent of waking time during nighttime naps. The broken line represents the E-W regression line in bed-rested humans calculated under conditions involving minimal stress (4). The equation of the E-W regression line was Y = (0.042 _+0.016) X + (0.688 m 0.273) where X = percent of waking time and Y = epinephrine excretion. The shaded area represents the standard deviation of the regression line.
and those lbr state Ill (Fig. 3). Table 1 shows means and standard deviations of sleep variables and epinephrine excretion during the 2-hour nap tbr each state. For state I, mean percent of waking time was 8. I%, and other polysomnographic findings were good: a short period of sleep latency, a small number of awakenings and a large percent of stage 3 ~ 4. Mean epinephrine excretion was low. The subjective estimates for state I were better than for state I11. For state II1, all subjects felt their sleep had been light and not satisfactory at all. Indeed• mean percent of waking time for slate 111 was high at 44.Y'~ and all the data indicated over 30c~ waking time. Polysomnographic findings sho~cd a long period of sleep latency, a large number of awakenings, and a lower percent of stage 3 + 4. Epinephrine excretion increased along the E - W regression line. For slate I1, polysomnographic findings were similar to those for state 1. Epinephrine excretion was distributed around the basal regression line. Most subjective estimates for state I1 were only slightly positive or negative on three items. On the fourth, sleep depth, three subjects on live occasions estimated their sleep depth to be m o d e r a t e b negative, I)ISCUSSION
the E - W regression line calculated in our previous study (4). The mean and standard deviation of epinephrine excretion during sleep was 1.92 + 1.09 ng/min, while waking time was 19.7 + 9.2%. The epinephrine level before sleep was 7.79 _+ 4.21 ng/ min. We divided the total of 18 times into three states, that is, state 1: percent of waking time less than mean - SD, state II: percent of waking time ranged from mean - SD to mean + SD, and state III: percent of waking time more than mean + SD. Figures 2 and 3 show a four-item, six-point bipolar scale: four sleep diagrams and subjective sleep estimates for state I (Fig. 2),
Sub. ~1 N2
w'
B
According to our previous studies (3-5), when percent of waking time during time allotted for sleep increased, the levels of epinephrine excretion also increased along the E - W regression line between the two parameters. In this study, the same results were obtained for short-duration sleep. We compared the self-estimates for sleep between state 1 (good sleep state, less than mean SD percent of waking time) and state III (poor sleep state, more than mean + SD percent of waking time). The subjective estimates of state I were better than those for state !1I. On the whole, the self-estimates tbr state I might have been influenced by the shortness of the 2-hour nap, though polygraphically, they slept very well. State lll's subjective
1
'
Fell asleep quickly,
1
¢
,
Uninterrupted sleep,
1
N1
~
¢
Deep sleep,
Sub. # 3
,
time awake
. Frequently awakened
,Light sleep
¢
2 Adequate amount of sleep,
•
w
. Not
enough
sleep
,Long time awoke
R
Fell
1 2
Uninterrupted s l e e ~
3
Deep sleep.
~'
Adequate amount of sleep,
,.Not enough steep
Fell asleep quickly,
,Long time awoke
B
1
asleep quicklyl
Uninterrupted sleep.
, Frequently awakened
@
~
sob. , 4 : ~ ,
.Light sleep
Adequate amount of sleep,
~
Felt asleep quick/D
i
,Not enough sleep ¢
,
,Long time awake
~
,Light sleep
, Frequently awakened
Uninterrupted skeep~
Deep sleep, Adequale amount of sleepL
,Light sleep
. Frequently awakened
Deep sleep,
N2
,Long
~
.Not enough sleep
FIG. 2. Four sleep diagrams and subjective sleep-estimates fbr state I in order of least percent of waking time. Percent of waking time and epinephrine excretion for each subject were, from top to bottom (4.9%, 1.45 ng/min), (7.5%, 0.34 ng/min), (9.9%, 1.09 ng/min), and (10.0%, 1.14 ng/min).
EPINEPHRINE, WAKING TIME, SELF-ESTIMATES
399
sub. #s N2 "l
Fell asteep quicktyt
2
Uninterrupted sleep.
a
i
•
~Long time awoke ¢
Deep steep
4 2
l
Sub.#6 w '1 li l N1 1 2 ]
-Light steep
Adequate amount of sleep.
R
~ [ 1
¢Not enough sleep
Fell asleep quicktyl
t
Uninterrupted sleep.
A
¢
,Long time awoke . Frequently awakened
Deep sleep,
4
~Light steep
Adequate amount of sleep, ]
Fell asleep quickly, Uninterrupted steep.
,
Not enough sleep
¢
,Long time awoke
@
. Frequently awakened
Deep steep, 0
¢
2
R
1
1 2 3
L ~
4
/
¢
Adequate amount of sleep.
Sub.#5N3w~ -jj-j~lm
. Frequently awakened
.Lighl steep ~Not enough steep
Uninterrupted sleep,
-Frequently awakened
Deepsteep,
~
Adequate amount of sleep.
.Lighl sleep ~Not enough sleep
FIG. 3. Four sleep diagrams and subjective sleep estimates for state II1 in order of greatest percent of waking time. Percent of waking time and epinephrine excretion for each subject were, from top to bottom (59.2%, 3.70 ng/min), (50.8%, 5.21 ng/min), (36.6%, 2.51 ng/min), and (32.5%, 3.45 ng/min).
In Japan, m a n y night-shift workers regularly take a nighttime nap during a working period (1). For health reasons, it is necessary that we investigate objectively whether or not they sleep during such a nap. The results for state 11I m e a n that levels of urinary epinephrine could be of help in objectively investigating a transient i n s o m n i a c state during nighttime naps.
reports were very or moderately negative. T h e i r polysomnographic findings showed a large percent of waking time, a large n u m b e r of awakenings, a n d a long period o f sleep latency. Epin e p h r i n e excretion was found to have increased in this state. The subjective a n d objective findings for state III suggest the existence of a transient i n s o m n i a c state.
TABLE 1 MEANS AND STANDARD DEVIATIONS OF SLEEP VARIABLESAND EPINEPHRINE EXCRETION DURING A 2-h NAP State I
0031-9384/92 $5.00 + .00 Copyright© 1992PergamonPressLtd.
Printedin the USA.
BRIEF COMMUNICATION
Relationships Between Epinephrine, Waking Time, and Self-Estimates for Naps at Night K Y O K O N I S H I H A R A *l A N D K A Z U K O M O R I t
*Department of Psychophysiology, Tokyo Institute of Psychiatry, Tokyo, Japan, and ¢Division of Work Physiology and Psychology, The Institute for Science of Labour, Kawasaki, Japan Received 16 September 1991 NISHIHARA, K. AND K. MORI. Relationshipsbetweenepinephrine, wakingtime, and self-estimatesfor napsat night. PHYSIOL BEHAV 52(2) 397-400, 1992.--!n a previous study, we established a basal regression line between epinephrine excretion and percent of waking time in bed-rested humans under conditions involving minimal stress. We suggested that this regression line might be clinically useful for observing insomnia. The present study was undertaken to examine whether epinephrine excretion during short-duration sleep, such as naps, is influenced by a transient insomniac state. Polysomnograms during a 2-hour nap from 0000 to 0200 were recorded three times for six healthy male subjects. Epinephrine excretion levelsduring sleep were found to increase along the basal regressionline when percent of waking time went up. The subjects with over 30% waking time during this 2-hour period reported that their sleep was light and not satisfactoryat all. These resultssupport our position that the regression line can be useful for observing a transient insomniac state. Epinephrine excretion
Waking time
Self-estimates
Naps
THE present study is part of a series of studies investigating whether or not epinephrine levels in urine during sleep can be used as an effective and objective indicator of insomnia. In 1985, we found a strong positive correlation between epinephrine excretion and percent of waking time during time allotted for sleep (3). In 1988, we established a basal regression line between these two parameters (the E-W regression line) in bed-rested humans under conditions involving minimal stress (4). We then confirmed that epinephrine excretion levels during 8 hours of usual nocturnal sleep were distributed along the E-W regression line in younger and older volunteers (5). In the present study, levels of epinephrine excreted during nighttime naps were examined to see whether the E-W regression line could be used to observe a transient insomniac state even in short-duration sleep. In addition, we checked subjective sleep estimates. METHOD Six healthy male university students (mean age: 22.3 years) volunteered for the experiment. All had previously participated in one of the authors' studies. The experiment was conducted three times for each subject. The subjects were instructed not to engage in any strenuous physical exercise on the days of the experiment. They came to the laboratory at 2000, and urinated.
Transient insomniac state
After electrodes were attached, the subjects relaxed by listening to music, reading magazines, and talking to each other. Urine samples were collected before going to bed at 0000, and after getting up at 0200. Before going to bed, they were told they would be awakened at 0200. After getting up, they were asked to estimate their sleep on a four-item six-point bipolar scale. For this study, we recorded negative feelings when the subjects reported long awake time, frequent awakenings, light sleep, and not enough sleep. Electroencephalograms (F3-A~, C3-At, Oi-Al), electromyograms, electro-oculograms, and electrocardiograms were continuously recorded using a Medilog 9000 cassette tape recorder. Polysomnograms were replayed to Mingograph. Sleep stages were scored by an epoch of I min according to the criteria of Rechtschaffen and Kales (6). Percent of waking time was calculated as the amount of total waking time divided by the amount of time in bed (2 h). The amount of free epinephrine in the urine samples was measured by high performance liquid chromatography with fluorometric detection using the trihydroxyindole method (2). RESULTS Figure 1 shows epinephrine levels and percent of waking time during the 2-hour nap. All subjects' data were distributed along
t Requests for reprints should be addressed to K. Nishihara, Department of Psychophysiology,Tokyo Institute of Psychiatry, 2-1-8, Kamikitazawa, Setagaya-ku, Tokyo 156, Japan.
397
398
NISHIHARA AND MaRl
8 n~i n 6 =
4
B. uJ
2
0
i0 4"o g0 8'o(%)160 Percent of waking time
FIG. 1. Relationship between epinephrine excretion and percent of waking time during nighttime naps. The broken line represents the E-W regression line in bed-rested humans calculated under conditions involving minimal stress (4). The equation of the E-W regression line was Y = (0.042 _+0.016) X + (0.688 m 0.273) where X = percent of waking time and Y = epinephrine excretion. The shaded area represents the standard deviation of the regression line.
and those lbr state Ill (Fig. 3). Table 1 shows means and standard deviations of sleep variables and epinephrine excretion during the 2-hour nap tbr each state. For state I, mean percent of waking time was 8. I%, and other polysomnographic findings were good: a short period of sleep latency, a small number of awakenings and a large percent of stage 3 ~ 4. Mean epinephrine excretion was low. The subjective estimates for state I were better than for state I11. For state II1, all subjects felt their sleep had been light and not satisfactory at all. Indeed• mean percent of waking time for slate 111 was high at 44.Y'~ and all the data indicated over 30c~ waking time. Polysomnographic findings sho~cd a long period of sleep latency, a large number of awakenings, and a lower percent of stage 3 + 4. Epinephrine excretion increased along the E - W regression line. For slate I1, polysomnographic findings were similar to those for state 1. Epinephrine excretion was distributed around the basal regression line. Most subjective estimates for state I1 were only slightly positive or negative on three items. On the fourth, sleep depth, three subjects on live occasions estimated their sleep depth to be m o d e r a t e b negative, I)ISCUSSION
the E - W regression line calculated in our previous study (4). The mean and standard deviation of epinephrine excretion during sleep was 1.92 + 1.09 ng/min, while waking time was 19.7 + 9.2%. The epinephrine level before sleep was 7.79 _+ 4.21 ng/ min. We divided the total of 18 times into three states, that is, state 1: percent of waking time less than mean - SD, state II: percent of waking time ranged from mean - SD to mean + SD, and state III: percent of waking time more than mean + SD. Figures 2 and 3 show a four-item, six-point bipolar scale: four sleep diagrams and subjective sleep estimates for state I (Fig. 2),
Sub. ~1 N2
w'
B
According to our previous studies (3-5), when percent of waking time during time allotted for sleep increased, the levels of epinephrine excretion also increased along the E - W regression line between the two parameters. In this study, the same results were obtained for short-duration sleep. We compared the self-estimates for sleep between state 1 (good sleep state, less than mean SD percent of waking time) and state III (poor sleep state, more than mean + SD percent of waking time). The subjective estimates of state I were better than those for state !1I. On the whole, the self-estimates tbr state I might have been influenced by the shortness of the 2-hour nap, though polygraphically, they slept very well. State lll's subjective
1
'
Fell asleep quickly,
1
¢
,
Uninterrupted sleep,
1
N1
~
¢
Deep sleep,
Sub. # 3
,
time awake
. Frequently awakened
,Light sleep
¢
2 Adequate amount of sleep,
•
w
. Not
enough
sleep
,Long time awoke
R
Fell
1 2
Uninterrupted s l e e ~
3
Deep sleep.
~'
Adequate amount of sleep,
,.Not enough steep
Fell asleep quickly,
,Long time awoke
B
1
asleep quicklyl
Uninterrupted sleep.
, Frequently awakened
@
~
sob. , 4 : ~ ,
.Light sleep
Adequate amount of sleep,
~
Felt asleep quick/D
i
,Not enough sleep ¢
,
,Long time awake
~
,Light sleep
, Frequently awakened
Uninterrupted skeep~
Deep sleep, Adequale amount of sleepL
,Light sleep
. Frequently awakened
Deep sleep,
N2
,Long
~
.Not enough sleep
FIG. 2. Four sleep diagrams and subjective sleep-estimates fbr state I in order of least percent of waking time. Percent of waking time and epinephrine excretion for each subject were, from top to bottom (4.9%, 1.45 ng/min), (7.5%, 0.34 ng/min), (9.9%, 1.09 ng/min), and (10.0%, 1.14 ng/min).
EPINEPHRINE, WAKING TIME, SELF-ESTIMATES
399
sub. #s N2 "l
Fell asteep quicktyt
2
Uninterrupted sleep.
a
i
•
~Long time awoke ¢
Deep steep
4 2
l
Sub.#6 w '1 li l N1 1 2 ]
-Light steep
Adequate amount of sleep.
R
~ [ 1
¢Not enough sleep
Fell asleep quicktyl
t
Uninterrupted sleep.
A
¢
,Long time awoke . Frequently awakened
Deep sleep,
4
~Light steep
Adequate amount of sleep, ]
Fell asleep quickly, Uninterrupted steep.
,
Not enough sleep
¢
,Long time awoke
@
. Frequently awakened
Deep steep, 0
¢
2
R
1
1 2 3
L ~
4
/
¢
Adequate amount of sleep.
Sub.#5N3w~ -jj-j~lm
. Frequently awakened
.Lighl steep ~Not enough steep
Uninterrupted sleep,
-Frequently awakened
Deepsteep,
~
Adequate amount of sleep.
.Lighl sleep ~Not enough sleep
FIG. 3. Four sleep diagrams and subjective sleep estimates for state II1 in order of greatest percent of waking time. Percent of waking time and epinephrine excretion for each subject were, from top to bottom (59.2%, 3.70 ng/min), (50.8%, 5.21 ng/min), (36.6%, 2.51 ng/min), and (32.5%, 3.45 ng/min).
In Japan, m a n y night-shift workers regularly take a nighttime nap during a working period (1). For health reasons, it is necessary that we investigate objectively whether or not they sleep during such a nap. The results for state 11I m e a n that levels of urinary epinephrine could be of help in objectively investigating a transient i n s o m n i a c state during nighttime naps.
reports were very or moderately negative. T h e i r polysomnographic findings showed a large percent of waking time, a large n u m b e r of awakenings, a n d a long period o f sleep latency. Epin e p h r i n e excretion was found to have increased in this state. The subjective a n d objective findings for state III suggest the existence of a transient i n s o m n i a c state.
TABLE 1 MEANS AND STANDARD DEVIATIONS OF SLEEP VARIABLESAND EPINEPHRINE EXCRETION DURING A 2-h NAP State I
