Sleep, 6(1):29-35
© 1983 Raven Press, New York
Postprandial Sleepiness: Objective
Documentation via Polysomnography
Monte L. Stahl, William C. Orr, and Cynthia Bollinger
Sleep Disorders Center, Presbyterian Hospital, Oklahoma City, Oklahoma, U.S.A.
Summary: Fifteen normal volunteers were evaluated to assess the effect of a
meal on sleep onset latency. The meal was administered in a counterbalanced
design on 1 of 2 successive days. Subjects napped 20 min subsequent to the
meal (or at the corresponding time on the no-meal day) and 1 h after the
initiation of the first nap. Ten subjects completed the Stanford Sleepiness Scale
(SSS) on arriving at the laboratory, and just prior to nap 1 and nap 2. Sleep
onset latency after the meal was not significantly different from that obtained
under the no-meal condition, but was significantly less on nap 1 as compared
with nap 2 irrespective of day or meal. SSS did not reveal subjective differences in sleepiness between the initial estimate and the postmeal estimate.
Only five subjects showed a decrease in sleep onset latency postprandially
(1-11 min). Although group differences in postprandial sleepiness were not
documented, the phenomenon was clearly exhibited by certain individuals.
Thus, postprandial sleepiness is not an invariable consequence of meal ingestion; rather, it appears to be affected by numerous variables such as hunger,
volume of the meal, and meal constituents. Key Words: Postprandial
sleepiness-Sleep latency-Subjective sleepiness.
Postprandial sleepiness is a commonly recognized and familiar experience. Although
eating is anecdotally recognized as a soporific experience, there is a paucity of objective
documentation concerning the effect of food ingestion on subjective and objective
parameters of sleep. It has, however, been noted as a serendipitous finding. Colquhoun
et al. (1) documented that measures of performance (signal detection, response latency,
and mathematical calculations) are positively correlated with body temperature. In
similar investigations they documented that when performance on a simple arithmetic
calculation test was plotted with a rise in body temperature across the day, the output,
or number of sums attempted, improved as temperature rose across the morning. There
was, however, a consistent drop in performance during the postlunch period despite a
further rise in body temperature (2). This postprandial fall in performance was also
demonstrated with detection rate on a vigilance task (3). Horne et al. (4) also noted a
Accepted for publication October 1982.
Address correspondence and reprint requests to Monte L. Stahl, Sleep Disorders Center, Presbyterian
Hospital, Northeast 13th at Lincoln Boulevard, Oklahoma City, OK 73104, U.S.A.
29
30
M. L. STAHL ET AL.
postlunch dip in performance for subjects whose performance was generally better in
the morning as opposed to the evening.
Recent research suggests that food constituents can affect sleep behavior. In 1979
Hartman and Spinweber (5) noted that 1 g of L-tryptophan significantly reduced sleep
onset latency, whereas 0.5 g did not. Since 1 g of L-tryptophan is within daily dietary
intake, the authors suggested that this amino acid might contribute to the physiology of
sleepiness after ingestion of a high protein meal that may contain up to 1 g of tryptophan.
The effect of tryptophan in the brain is mediated by a complex set of factors. It has
been well documented in rats that a single protein-free, high carbohydrate meal will
elevate brain tryptophan levels with a consequent increase in levels of brain serotonin
and its major metabolite, 5-hydroxyindoleacetic acid (6). The addition of protein to a
carbohydrate meal may actually suppress the increase in brain tryptophan and serotonin. The protein contributes neutral amino acids to the plasma that compete with
tryptophan for uptake in the brain. Fernstrom et al. (7) documented a decrease in the
ratio of plasma tryptophan to large neutral amino acids in normal human subjects as
protein content in the diet was increased. They suggested further that dietary intake of
carbohydrate and protein may also affect brain monoamine synthesis in humans. One
may also infer that these alterations in brain metabolism may, in turn, affect sleep
and/or the subjective sensation of sleepiness.
In 1978 Richardson et al. (8) established the sleep onset latency and the Multiple
Sleep Latency Test as a reliable method of objectively measuring daytime sleepiness.
Patients with pathological daytime sleepiness have significantly shorter sleep onset
latencies than controls. Since documentation of daytime postprandial sleepiness has
been anecdotal in nature and has not been investigated via polysomnography, this
investigation sought to document sleepiness in a randomly selected group via polygraphic criteria (sleep onset latency) within the first 100 min after eating a typical meal.
METHODS
Subjects were 12 male and 11 female volunteers between the ages of 21 and 30 years.
N one was obese and all were entrained to a routine of regular nighttime sleeping hours.
None had complaints of excessive daytime sleepiness or disturbed nocturnal sleep.
Subjects were instructed to maintain a normal bedtime and awakening time and to
refrain from alcohol consumption prior to each laboratory session. Written informed
consent was obtained. Subjects were informed that the purpose ofthe investigation was
"to document the occurrence of postprandial sleepiness following a meal ... or sleepiness which may occur without having just eaten a meal."
Subjects were evaluated according to the following protocol. On day 1, subjects were
asked to awaken at 8:00 a.m. and to eat a predetermined breakfast of 1 piece of toast
with butter and jelly, 1 cup of orange juice, and 1 cup of coffee, if coffee was normally
consumed. Subjects were instructed to eat "a normal amount for you, but not more
than is listed." They arrived at the laboratory at 9:30 a.m., at which time electrodes
were attached for polygraphic monitoring of sleep according to standard criteria (9). At
10:00 a.m. subjects comsumed a hamburger consisting of a buttered bun and 3 oz.
(cooked) of ground beef with mayonnaise, tomato, and lettuce. Subjects also consumed
Sleep. Vol. 6. No. I, 1983
POSTPRANDIAL SLEEPINESS
31
4 oz. of french fries and 2 cups of ginger ale. This meal contained 31.4 g of protein, 113.4
g of carbohydrate, 46.9 g of fat, and approximately 357.3 mg of tryptophan. Subjects
were instructed to eat the entire meal. Twenty minutes after completing the meal,
subjects were placed in bed and asked to go to sleep (postprandial nap O. Subjects were
awakened within 5 min of the first spindle in the polygraphic tracing, to avoid accumulating sleep. One hour after the initiation of the first nap, the second nap (postprandial nap 2) was conducted. At the completion of the second nap the subject was
dismissed from the laboratory.
On day 2, the same protocol was followed as on day 1, except that the 10:00 a.m.
meal was omitted. The meal consumption was counterbalanced across days 1 and 2 to
obviate order effects. Ten subjects completed the Stanford Sleepiness Scale (SSS) on
arriving at the laboratory, and just prior to the first and second naps on both days.
A mealtime of 10:00 a.m. was selected to avoid the possibility that hunger would
disrupt or delay sleep in the no-meal condition. Therefore, time of day was kept constan"t, whereas the variable of hunger as a possible deterrent to sleep was eliminated. In
addition, the time of the evaluation would not overlap with the afternoon dip in sleep
onset latency documented by Richardson et al. (8).
RESULTS
The data are presented in Tables 1 and 2. Eight of 23 subjects did not sleep during
one of the naps within 50 min. Only those subjects (nine males and six females) who
slept during all four naps were included in the data analysis. The latencies to stages 1
and 2 were scored by standard criteria (9). Data were analyzed by means of an analysis
of variance for nested variables (0). Analyses revealed no significant differences between males and females, days 1 and 2, or meal and no meal (latency to stage 2).
Therefore, the consumption of the meal did not significantly alter sleep latency. When
data were analyzed for latency to stage 1, nap 1 latency was significantly less (p < 0.02)
than the latency during nap 2 (nap 1 x = 6.65 min, SD = 4.64 min versus nap 2 x = 11.43
min, SD = 9.85 min). This difference was not statistically significant when latency to
stage 2 was analyzed (nap 1 x = 13.22 min, SD = 7.36 min versus nap 2 x = 17.84 min,
SD = 11.50 min). The SSS scores did not vary significantly with any main effects. Of
the 15 subjects, only 5 showed a decrease in sleep onset latency to stage 1 during the
first nap following the meal as compared with the first nap in the no-meal condition; this
decrease ranged from 1 to 11 min. The other subjects actually showed an increased
sleep onset (to stage 1) during the first nap following the meal.
DISCUSSION
Group differences in postprandial sleepiness were not documented either objectively
(sleep onset latency) or subjectively (SSS). The intervening meal did produce a decreased sleep onset latency in certain individuals; however, the results of this study
strongly suggest that postprandial sleepiness is affected by a myriad of variables that
should be addressed in future studies, including emptying and digestion time, volume of
the meal, protein and carbohydrate content, time of day, and level of daytime sleepiness (previous night's sleep).
Sleep, Vol. 6, No. I, 1983
Vv
~
"'"
~
~
5>'
.-~
-
TABLE 1. Nap sleep onset latency (min)
~
Day 2
Day 1
Nap 1
Subject
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Mean
(SD)
Stg 1 Stg 2
4.15
9.54
Nap 2
7.25
11.05
2.62
9.85
18.05
15.97
16.57
11.70
6.54 16.97
13.40 20.12
Nap 2
Nap 1
5.90 12.24
11.57 13.67
9.00 10.98
3.25
Nap 2
Stg 1 Stg 2 Stg 1 Stg 2 Stg 1 Stg 2 Stg 1 Stg 2
~
9.05
t'-<
4.09 13.35
1.54 4.69
7.35 12.82
4.30 6.92
4.45 7.49
8.55 32.47
18.72
13.24 39.65
9.74 11.10
11.41
28.40 33.35
19.47 32.19
0.07
4.14
2.55
9.39
10.52 11.92
7.54 11.42
3.30
12.06
6.87
14.30
4.35
7.27
7.52
18.69
12.54
17.30 5.98 13.07
5.13
11.39 25.73
10.89 15.39
8.22 8.28
7.81
6.27
14.78
9.03
6.75
3.31
18.00
13.19
15.07
1.04
Vl
~
3.49
2.82
11.47 11.09
5.74 15.89
8.54
21.20
15.22
17.20
::t:
t'-<
t"rl
~
16.37 43.42
~
t'-<
5.04 13.24
7.57
15.82 12.33 13.42
3.29 6.69 9.14
12.74 45.17 45.71
2.60 5.84 4.15
7.17 12.49 16.62
5.19 8.43 13.33
Nap 1
5.25
7.55
12.59
3.37
17.19
Nap 2
-----
Stg 1 Stg 2 Stg 1 Stg 2 Stg 1 Stg 2
17.39 29.39 22.60 24.79
5.50 6.10 21.30 23.02
4.22
Nap 1
No-meal naps
Postprandial naps
No-meal naps
Postprandial naps
9.82
2.44
2.12
10.85 3.74
1.21
4.25
8.57 31.05
8.02
1.56
32.24
7.04
5.19
2.96
26.12
12.91
7.55
6.29
12.21
10.17
12.65
16.69
13.80
POSTPRANDIAL SLEEPINESS
33
TABLE 2. SSS score
Postprandial naps
Mean
(SD)
No-meal naps
Mean
(SD)
Initial
Nap 1 Nap 2
2.8
(0.7)
2.8
(0.4)
3.0
(0.4)
4.0
(1.6)
4.0
( 1.6)
2.5
(0.5)
Our data seem consistent with previously reported temperature and performance
data if it is assumed that rising temperature and good performance are associated with
increased alertness and therefore resistance to sleep (i.e., longer sleep onset latency).
Thus, the relatively shorter nap 1 sleep onset latencies might be explained by the
occurrence of the ascending limb of the circadian temperature cycle, resulting in lower
core body temperatures associated with the earlier nap. When the latency to stage 2 was
compared, nap 1 was not found to be significantly different from nap 2. Therefore, stage
1 may prove to be more sensitive to subtle changes in sleepiness.
The volume of the meal given to volunteers in this experiment was held constant.
Therefore, it represented a relatively light lunch for some of the male volunteers,
whereas some of the female volunteers reported that they rarely ate this volume at any
one meal. If afferent stimulation owing to gastric distention was a factor, the effect
would have been variable across the group.
Of primary importance in assessing the effect of food constituents on sleep onset
latency is the metabolism of tryptophan. Beazell (11) has shown that amino acids are
not absorbed in the stomach, and Adibi and Gray (12) noted that tryptophan had the
second lowest rate of intestinal absorption in man. This rate may be retarded further by
competition from other amino acids for the transport site. In addition, tryptophan slows
gastric emptying by way of cholecystokinin release (13). Beazell (11) showed that after
1 h, less than 1% of the protein and only 9.9% of the carbohydrate from a 16.3-g protein
and 30-g carbohydrate meal had been digested to a soluble form. In the present study it
is unlikely that gastric, emptying occurred rapidly enough to allow digestion of tryptophan within the time of evaluation (1 h, 20 min).
The meal used in this experiment was selected because it represented a typical lunch
or light evening meal. It contained both protein (31.4 g) and carbohydrate (113.4 g),
which Frenstrom et al. (7) have shown may actually lower the ratio of plasma tryptophan to other neutral amino acids. The work of Hartmann and Spinweber (5) and Porter
and Horne (14) indicates that in order to affect sleep (either sleep onset latency or sleep
architecture), relatively high doses of tryptophan (> 1 g) or high carbohydrate (glucose)
levels (which increase insulin release and subsequent brain tryptophan uptake) are required. The amount of tryptophan contained in this meal was approximately Y:3 g and
therefore probably was insufficient to produce sleepiness.
A hypnogenic effect of gastrointestinal origin has also been noted by investigators of
alimentary physiology. In 1929, Alvarez (15) noted that distention of a jejunal balloon
caused human subjects to fall asleep. Kukorelli and Juhasz (16) induced cortical synchronization in cats by balloon distention and electrical stimulation of the splanchnic
Sleep, Vol. 6, No. /, /983
34
M. L. STAHL ET AL.
nerve. A similar effect was noted by Fara et al. (17). The introduction of milk into the
duodenum resulted in the cats' curling up and going to sieep (EEG spindling). They
postulated that sedation induced by intraduodenal fat (milk) may be mediated by either
a release of a gastrointestinal hormone (probably cholecystokinin-pancreozymin) or an
indirect neurogenic mechanism of stimulation of duodenal receptors. In man, pancreatic enzyme secretion (including cholecystokinin) is stimulated by tryptophan (18).
However, tryptophan itself, as noted earlier, is not in high concentrations within the
first postprandial hour and cholecystokinin does not reach peak serum levels for 45-60
min after eating (19). Therefore, it is unlikely that cholecystokinin, tryptophan, or
duodenal afferents significantly influence sleepiness in the first postprandial hour.
These factors more likely contribute to the sleepiness noted in most individuals 2 - 3 h
after lunch.
Horne et al. (4) have clearly identified subjects as either" morning or evening types,"
based on self-assessment questionnaires and performance efficiency. A postlunch dip in
performance was evident for morning types but not for evening types. Therefore, an
individual's innate tendency to perform efficiently in the morning versus evening may
actually be linked to the tendency to fall asleep after eating. Subjects in this experiment
were screened to eliminate individuals with abnormal levels of daytime sleepiness, but
they were not carefully questioned on the level of alertness in the morning versus
evening or the presence of postprandial sleepiness.
The purpose of this evaluation was to determine the presence or absence of postprandial sleepiness in a randomly selected group. It appears that postprandial sleepiness is not the inevitable consequence of food intake for all individuals. It is a multifactorial phenomenon involving both dietary and other psychobiological factors.
Acknowledgment: The authors gratefully acknowledge Dr. J. Catesby Ware for his critical
review of the manuscript and James Duke for his assistance with statistical analysis of the data.
REFERENCES
1. Colquhoun WP, Blake MJF, Edwards RS. Experimental studies of shift work. I: A comparison of
rotating and stabilized four-hour shift systems. Ergonomics 1968;11:437-53.
2. Colquhoun WP, Blake MJF, Edwards RS. Experimental studies of shift work. II: Stabilized 8-hour shift
systems. Ergonomics 1968; 11 :527 -46.
3. Colquhoun WP, Blake MJF, Edwards RS. Experimental studies of shift work. III: Stabilized 12-hour
shift systems. Ergonomics 1969; 12:865-82.
4. Horne JA, Brass CG, Pettitt AN. Circadian performance differences between morning and evening
"types." Ergonomics 1980;23:29- 36.
5. Hartmann E, Spinweber CL. Sleep induced by L-tryptophan: Effect of dosages within the normal dietary
intake. J Nerve Ment Dis 1979;167:497-9.
6. Wurtman RJ, Fernstrom JD. Control of brain monoamine synthesis by diet and plasma amino acids. Am
J c/in Nutr 1975;28:638-47.
7. Fernstrom JD, Wurtman RJ, Hammarstrom-Wiklund B, Rand WM, Munro HN, Davidson CS. Diurnal
variations in plasma concentrations of tryptophan, tyrosine, and other neutral amino acids: Effects of
dietary protein intake. Am J CUn Nutr 1979;32: 1912-22.
8. Richardson GS, Carskadon MA, Flagg W, Hoed JVD, Dement WC, Mitler MM. Excessive daytime
sleepiness in man: Multiple sleep latency measurement in narcoleptic and control subjects. Electroencephalogr Clin Neurophysiol 1978;45:621-7.
9. Rechtschaffen A, Kales A, eds. A manual of standardized terminology, techniques and scoring system
for sleep stages of human subjects. Brain Information Service/Brain Research Institute, University of
California at Los Angeles, 1968.
Sleep, Vol. 6, No.1, 1983
POSTPRANDIAL SLEEPINESS
35
10. Boneau CA. The effects of violations of assumptions underlying the t test. Psychol Bull 1960;57:49-64.
11. Beazell JM. A reexamination of the role of the stomach in the digestion of carbohydrate and protein. Am
J Physiol 1941;132:42-50.
12. Adibi SA, Gray SJ. Intestinal absorption of essential amino acids in man. Gastroenterology
1967;52:837-45.
13. Fisher M, Hunt IN. Effects of hydrochlorides of amino acids in test meals on gastric emptying. Digestion
1977; 16: 18- 22.
14. Porter JM, Horne JA. Bedtime food supplements and sleep: Effects of different carbohydrate levels.
Electroencephalogr Clin Neurophysiol 1981;51:426-33.
15. Alvarez we. Physiologic studies on the motor activities of the stomach and bowel in man. Am J Physiol
1929;88:658-60.
16. Kukorelli T, Juhasz G. Sleep induced by intestinal stimulation in cats. Physiol Behav 1976;19:355-8.
17. Fara JW, Rubinstein EH, Sonnenschein RR. Visceral and behavioral responses to intraduodenal fat.
Science 1969;166: 110-1.
18. Go VLW, Hofmann AF, Summerskill WHJ. Pancreozymin bioassay in man based on pancreatic enzyme
secretion: Potency of specific amino acids and other digestive products. J Clin Invest 1970;49: 1558-64.
19. Rayford PL, Fender HR, Ramus NI, Reeder DD, Thompson JC. Release and half-life of CCK in man. In:
Thompson JC, ed. Symposium gastrointestinal hormones. Austin: University of Texas Press,
1975:301-18.
Sleep, Vol. 6, No.1, 1983