mise à jour du
21 mai 2006
Pharmacol Biochem Behav
 Effects of brief caffeinated-beverage deprivation on mood, symptoms, and psychomotor performance
Lane JD
Department of Psychiatry and Behavioral Sciences,
Duke University Medical Center, Durham, USA.


The effects of short-term deprivation of caffeinated beverages on mood, withdrawal symptoms, and psychomotor performance were studied in habitual coffee drinkers. Twenty-four male and female coffee drinkers were tested at midday (1130-1330 h) under two conditions. On one day they consumed caffeinated beverages ad lib prior to testing, and on the other they remained caffeine abstinent. The order of treatments was counterbalanced. Mood and withdrawal symptom reports were collected by questionnaires. Psychomotor performance was tested with a computerized test battery. Caffeinated-beverage deprivation was associated with decreased vigor and increased fatigue and with symptoms including headache. No changes in psychomotor performance were observed. Even short periods of caffeinated-beverage deprivation, equivalent in length to missing regular morning coffee, can produce noticeable unpleasant caffeine-withdrawal symptoms by the middle of the day. These symptoms may be a common side effect of habitual caffeinated beverage consumption.
CAFFEINE is one of the most commonly used drugs, but it is not without side effects. Perhaps the most frequent is the pattern of physiological withdrawal symptoms that occur when habitual consumers abruptly stop. The characteristics and time course of the withdrawal syndrome appear to be consistent and are characterized by headache and arousal deficits that develop in a day or two and last up to a week with continued abstinence. The syndrome occurs even in people who consume as little as 100 mg of caffeine daily, equivalent to a single cup of coffee. One recent study of caffeine withdrawal in low to moderate consumers found that caffeine deprivation for 2 days produced increases in symptoms of depression and anxiety, decreases in vigor and friendliness, and increases in fatigue and confusion. Deprivation also elicited a variety of specific symptoms related to irritability, sleepiness and fatigue, difficulty with thinking and working, headache, and feeling generally unwell. These effects can be clinically important because the symptoms associated with caffeine withdrawal overlap with medical complaints commonly reported to physicians.
Most experimental investigations of caffeine withdrawal symptomatology have involved several days of caffeine deprivation. Although long periods of caffeine deprivation provide the opportunity to observe the full range and intensity of symptoms as they develop and resolve over time, such extended deprivation is relatively uncommon under normal circumstances, except for the rare individuals who attempt to quit caffeine consumption "cold turkey." Shorter periods of deprivation, for example, when a regular coffee drinker misses his or her normal morning coffee, would be much more common in everyday life. Studies suggest that these short deprivation periods too can lead to cli a y s ificant withdrawal symptoms, such as headach and fatigue.
In an earlier study of caffeine effects on neur endocrine stress reactivity in the work evironment, assessed mood and withdrawal symptom in people who were deprived of caffeine overnight and then eceived either 300 mg of caffeine or placebo at the start of the workday. Participants performed their normal work activities for 4 h and then rated their experience of the morning. When given placebo, participants reported higher levels of sleepiness, lethargy, and headache and a reduced desire to socialize. They also reported casually that it was much harder to work and to pay attention to what they were doing. Simply being deprived of normal morning coffee appeared to have clinically significant effects on these regular coffee drinkers, even after a few hours of deprivation.
The current study was designed to pursue this observation and to investigate the effects of such short-term caffeine deprivation on withdrawal symptoms and psychomotor performance. I sought to explore how regular coffee drinkers would feel during a normal workday morning if deprived of their regular morning coffee and whether such deprivation produced cognitive performance deficits that could affect their work. Regular coffee drinkers were tested at midday after mornings when they either consumed coffee and other caffeinated drinks ad lib or abstained completely from caffeine. Self-report questionnaires assessed mood and caffeine withdrawal symptoms, and a battery of computerized psychomotor tasks assessed psychomotor performance. Based on earlier observations, it was expected that even this brief period of deprivation would be associated with detectable withdrawal symptoms and performance decrements.
Periods of experimental caffeinated-beverage deprivation equivalent to people skipping their normal morning coffee produced detectable symptoms of caffeine withdrawal at midday. These effects were observed both in the POMS measures of self-rated mood and in the appearance of specific symptoms that have been associated with caffeine withdrawal. The pattern of results is similar to that observed for longer periods of deprivation and in my earlier ambulatory study .
The POMS factor for Vigor-Activity represents a mood of vigorousness, ebullience, and high energy associated with feeling cheerful, alert, active, and full of pep. The POMS factor for Fatigue-Inertia represents a m of w mess, inertia, and low energy. The combinatio of reduced vigor and increased fatigue reported on the PO *S was consistent with the pattern of reported withdrawal mptoms, which emphasized decreased levels of arousal as ociated with difficulty in concentrating. However, the shorte period of deprivation in our study was not associated with ircreases in anxiety or depression, as noted in longer periods of caffeine deprivation.
It is noteworthy that even this short period of deprivation produced significant ratings of headache and flu-like symptoms. On average, these differences may appear small in magnitude, but examination of the number of participants who experienced these particular symptoms suggests a different interpretation. Only one person reported headache on the ad lib consumption day, and the headache was given a rating of 1 on the O to 3 scale. In contrast, 10 people reported headache during the morning of deprivation, including 4 who gave their headache the maximum rating available. No one reported any flu-like symptoms during ad lib consumption, but five people did during deprivation, rating the magnitude as a 1. Thus, even short periods of caffeinated-beverage deprivation may produce clinically significant physical symptoms in regular coffee drinkers. In many respects, these observations are similar to reports of headache symptoms during short-term caffeine abstinence associated with religious fasting and surgical procedures.
The observed differences in casual blood pressure (lower when caffeine-deprived) are consistent with laboratory findings that caffeine administration is associated with increases in blood pressure. Our own laboratory studies have found that a single 250 mg dose of caffeine raises resting systolic and diastolic blood pressure by 7-10 mmHg 60 min after administration. The effects seen here are consistent with these earlier observations, given the variability in caffeine dose and timing in the present ad lib study. They confirm that ad lib caffeine consumption is associated with elevated blood pressure compared with caffeine abstinence, even in habitual coffee drinkers who should have developed tolerance to the drug's effects. This finding has implications for epidemiological studies of caffeine and cardiovascular disease risk, which have often collected blood pressure data under fasting (thus, caffeine-deprived) conditions. Casual blood pressure in heavy coffee drinkers is probably underestimated under such conditions, which could lead to false negative results regarding the association of coffee drinking and elevated blood pressure and misleading conclusions about the coronary disease risks associated with coffee or caffeine. Coffee's potential as a hypertension and coronary disease risk factor may need to be reevaluated.
Although anecdotal reports from participants and subjective measures of mood and symptoms suggested the presence of diminished cognitive capacity and functional impairment during caffeinated beverage deprivation, no deficits in psychomotor task performance were found. Similar negative results are common in the decades of research into caffeine's effects on performance, where comparisons of caffeinated and caffeine-deprived conditions yield performance differences that are typically small and capricious. The battery of tasks covered a variety of psychomotor functions from simple to complex. Only the serial memory task yielded possible evidence of impairment, and this was compromised by an order interaction. Given the changes in mood and symptoms, performance deficits caused by functional impairment or decreased motivation would be expected. It is possible that the specific tasks of the present study do not tap the dimensions of cognitive performance affected by caffeine deprivation. Furthermore, these tasks were all of relatively short duration, and participants may have been able to push themselves to overcome any withdrawalrelated deficits. Recently Streufert and colleagues reported that caffeine deprivation produced significant deficits in managerial performance measures collected during long, complex work simulations. Perhaps longer periods of more naturalistic cognitive and work-related tasks will provide a clearer demonstration of performance deficits in future studies.
The attempt to investigate whether heavier consumers experienced stronger withdrawal symptoms yielded some supporting evidence of correlation. This effort was hindered by the relatively small sample size, but relationships were observed for at least some of the symptoms. Other studies have demonstrated that deprivation can produce symptoms of withdrawal even in people who consume light to moderate amounts of caffeine, even as low as 100 mg (one cup of coffee) per day. Although this may be true, our preliminary evidence suggests that the experience of withdrawal symptoms may be more intense in people who habitually consume larger amounts of caffeine.
Compliance with instructions for caffeine abstinence was not confirmed objectively by measures of caffeine level in plasma or saliva. However, the possibility that some participants failed to maintain abstinence in the deprivation condition is not a serious limitation. Participants were asked directly about their compliance with instructions for abstinence or diary record-keeping, and we have no reason to suspect their reports. Moreover, scattered noncompliance with the abstinence condition would not likely yield the significant differences between ad lib and deprived conditions observed here. Rather, it would tend to increase the variability of scores in the deprived condition, making it even more difficult to detect differences between the two.
The present study was intended to simulate natural conditions of caffeinated beverage consumption and deprivation in the real world. This decision had several implications for the outcome. Because participants were asked to consume ad lib, caffeinated beverages, and presumably caffeine dose, varied both in amount and timing. Variations in caffeine dose probably contributed to variability among participants in scores for mood, symptoms, and performance on the ad lib day, which may have prevented detection of differences in some variables. In contrast, expectations about the effects of caffeine deprivation may have contributed to the observed differences in mood and symptoms, which were based on retrospective self-reports. Participants were not blind to treatment condition, because they maintained their own ad lib or abstinent status, and beliefs about caffeine withdrawal symptoms could have colored their reports. Moreover, the disruption of other normal routines that was caused by the experimental demands for caffeinated beverage deprivation may have had a negative effect on mood during the morning. A naturalistic study such as this cannot control these extraneous factors. As a result, observed differences reflect more than the presence or absence of caffeine. However, they do represent the broader experience of caffeinated-beverage deprivation, which naturally includes the expectations and the changes in routine, and which was the subject of the investigation.
In many respects, the present study confirms what most regular coffee drinkers would probably admit: they suffer when they don't get their regular morning coffee. However, investigation of the clinically significant effects of even brief periods of deprivation is worthwhile because these symptoms (e.g., headache, fatigue, etc.) are such common complaints presented to physicians and may be otherwise difficult to explain. Moreover, given the widespread use and increasing popularity of coffee, it is worth noting that habitual caffeine consumption is not without a potential cost to well-being. At the very least, habitual coffee drinkers run the risk of misery when they cannot get their regular cup. 
Aloe F Yawning. Arq Neuropsiquiatr 1994; 52; 2; 27-36`
Beale MD. Murphree TM. Excessive yawning and SSRI therapy. Int. J. Neuropsychopharmacol. 2000; 3: 275-276.
Bertschy G; Vandel S; Sechter D; Bizouard P Bâillements et excitation sexuelle sous clomipramine. Place des mécanismes serotoninergiques. A propos of 2 cases Encephale 1991; 17; 6; 515-7
Cohen AJ. Fluoxetine-induced yawning and anorgasmia reversed by cyproheptadine treatment. J. Clin. Psychiatry 1992: 53: 174.
Goldberg RL Sustained yawning as a side effect of imipramine Int J Psychiatry Med 1983-84; 13; 4; 277-80
Gutierrez-Alvarez AM Do your patients suffer from excessive yawning? Acta Psychiatrica Scandinavica 2007;115(1)-80-81
Harrison W; Stewart J; McGrath PJ; Quitkin F Unusual side effects of clomipramine associated with yawning. Can J Psychiatry 1984; 29; 546
Klein DF. J Repeated observations of yawning; clitoral engorgement; and orgasm associated with fluoxetine administration. Clin Psychopharmacol 1989; 9; 5; 384
Klein DF False suffocation alarms, spontaneous panics, and related conditions. an integrative hypothesis. Arch Gen Psychiatry. 1993; 50; 4; 306-317
Harada KI. Paroxetine-induced excessive yawning Psychiatry Clinical Neurosciences 2006; 60: 260
Holmgren R; Holmgren B; Rodriguez R; Gonzalez RM. Serotonergic modulation of yawning.Pharmacol Biochem Behav 1979; 11; 3; 371-2
Lane JD Effects of brief caffeinated-beverage deprivation on mood, symptoms, and psychomotor performance. Pharmacol Biochem Behav. 1997;58(1):203-208
Lane JD, Phillips-Bute BG Caffeine deprivation affects vigilance performance and mood. Physiol Behav. 1998;65(1):171-175
McLean JD, Forsythe RG, Kapkin IA Unusual side effects of clomipramine associated with yawning Can J Psychiatry 1983; 28; 7; 569-70
Modell JG. Repeated observation of yawning, clitoral engorgement, and orgasm associated with fluoxetine administration. J. Clin. Psychopharmacol. 1989; 9: 63-65.
Observation personnelle
Pae CU, JJ Kim et al Injured temporomandibular joint associated with fluoxetine-monotherapy-induced repeated yawning Gen Hosp Psychiatry 2003; 25; 3; 217-218
Prescrire Bâillements : parfois un effet indésirable d'un médicament La Revue Prescrire 2005; 25 ; 265; 676
Sommet A et coll. "Drug-induced yawning:a review of the French pharmacovigilance database" 9e congrès annuel de la Société française de pharmacologie; 26e journées de pharmacovigilance, Bordeaux : 26-28 avril 2005. Fundamental Clin Pharmacol 2005; 19 (2):227 (abstract P123).
Tesfaye Y, Lal S Hazard of yawning Canadian Med Assoc J 1990; 142; 1; 15 & 1991;1 45;12; 1560
Urba-Holmgren R, Holmgren B, Rodriguez R, Gonzalez RM. Serotonergic modulation of yawning. Pharmacol. Biochem. Behav. 1979; 11: 371-372.
 -Evans SM, Griffiths RR. Caffeine withdrawal: a parametric analysis of caffeine dosing conditions. J Pharmacol Exp Ther. 1999;289(1):285-294.
-Phillips-Bute BG, Lane JD.Caffeine withdrawal symptoms following brief caffeine deprivation. Physiol Behav 199763(1):35-39
-Lane JD Effects of brief caffeinated-beverage deprivation on mood, symptoms, and psychomotor performance. Pharmacol Biochem Behav. 1997;58(1):203-208
-Lane JD, Phillips-Bute BG Caffeine deprivation affects vigilance performance and mood. Physiol Behav.1998;65(1):171-175