Mwjohns.com

Psychopharmacology (2008) 200:255-260DOI 10.1007/s00213-008-1202-z Effects of caffeine on alertness as measured by infraredreflectance oculography Natalie Michael & Murray Johns & Caroline Owen &John Patterson Received: 15 January 2008 / Accepted: 12 May 2008 / Published online: 9 June 2008 using infrared reflectance oculography and converted into a Rationale Caffeine is a well-known stimulant that can be drowsiness score, Johns Drowsiness Scale (JDS).
used to increase alertness and performance especially in Results Caffeine significantly reduced JDS scores (drowsi- low arousal situations such as monotonous highway driving ness) and reaction times, and these changes persisted for 3 to or after sleep deprivation. The effects of caffeine in rested, 4 h. Self reports of sleepiness were not as sensitive, with alert, participants are less clear, and this may be attributable Karolinska Sleepiness Scale scores only being significantly to difficulties in objectively assessing small changes in lower in the caffeine compared to placebo condition at 30 min Objectives The present study examined the effects of Conclusions The results demonstrated that despite being caffeine in non-sleep-deprived participants with methods well rested, administration of caffeine significantly in- that have previously been shown to be sensitive to changes creased alertness and enhanced performance, and these in alertness. In order to avoid confounding results, low, or changes were able to be detected with the JDS.
non-users of caffeine, were sought as participants.
Materials and methods Twelve subjects participated in a Keywords Alertness. Blinks. Caffeine. Vigilance.
within-subjects double-blind placebo-controlled design study and were administered either a capsule containing 200 mg ofcaffeine or placebo on two separate days. Ten-minute long Caffeine is one of the most commonly used of all psychotropic tests of vigilance were performed at baseline and then at 30, drugs in today's society (Fredholm et al. ). When 60, 120, 180, and 240 min after swallowing the capsule.
administered to sleep-deprived participants, it has been During vigilance tests, eye blink variables were measured shown to increase performance and arousal level. Specifi-cally, caffeine has been found to decrease reaction times,increase sleep latency, and decrease physiological signs ofsleepiness in the electroencephalogram (Beaumont et al.
Patat et al. Wesensten et al. Performance in alert subjects often improves with the administration of caffeine, but the magnitude is usually smaller than in Swinburne University of Technology,P.O. Box 218, Hawthorn Victoria 3122, participants with lowered arousal levels (Lorist et al. ; Difficulties in objectively measuring drowsiness also in alert (non-sleep-deprived) people. The effects have previously been studied with physiological recordings,measures of performance (e.g., reaction times), and self ratings. A new system to measure alertness/drowsiness has Peter MacCallum Cancer Center Research Division,Melbourne, Australia become available and has not yet been studied in combination with caffeine (Optalert™, Johns et al. ).
a minimum of 3 days. Participants were asked to aim for This system utilizes infrared reflectance oculography to between 7 and 8 h of sleep the night proceeding each measure ocular variables, which are incorporated into a experimental day and to refrain from consuming any form drowsiness score, the Johns Drowsiness Scale (JDS; Johns of caffeine from midnight prior to an experimental day, et al. ). The JDS has been shown to indicate decreases in alertness associated with sleep deprivation and correlates Participants were asked a series of questions related to significantly with reaction times, lapses in performance, their habitual caffeine use. This included whether they and lane departures in a driving simulator (Johns et al.
regularly consumed caffeine, and if yes, on average how many caffeinated products they consumed per day. They To the authors' knowledge, no studies have examined were also asked to specify the type and number of each the effects of caffeine on blink duration or velocity or the caffeinated product. Using standard values of caffeine effects of caffeine on any ocular variable in alert partic- content of products, an estimate of caffeine consumption ipants. A limited number of studies have examined the within the participant sample was calculated. Coffee, tea, effects of caffeine on ocular variables in sleep-restricted or and energy drinks were considered to contain 100, 40, and sleep-deprived participants, but generally, these studies fail 80 mg, respectively (Food Standards Australia New to show any effect. Previously, caffeine has not effected Zealand ). On each experimental day, participants also pupil diameter, pupil contraction latency (Minzhong et al.
indicated the last time that they had consumed caffeine and ), or blink rate (Horne and Reyner and only specified what type of caffeinated product it was.
indicated trends towards increased saccade velocities All participants had normal vision requiring no correc- (Minzhong et al. The only other study that could tion and were not taking any medication that could directly be found regarding caffeine and ocular variables was one influence alertness. Two female participants did report use where clonidine was administered in order to lower arousal of oral contraceptives, and one smoked on a regular basis, levels (Smith et al. Interestingly, the caffeine was which could have interacted with the caffeine administered able to counteract the decreases shown in saccade velocity during the study (Abernethy and Todd Parsons and caused by clonidine (Smith et al. ). It is possible that in Neims ). Participants reported to be free of any sleep these previous studies, the ocular variables measured were disorders, and their Epworth Sleepiness Scale scores not sensitive enough to changes in alertness generally in indicated normal levels of daytime sleepiness (mean=6.6, order to demonstrate changes in alertness due to caffeine consumption. For example, variables such as blink rate and Each experimental day involved six testing sessions, saccade velocity have been shown to correlate poorly with during each of which participants performed a 10-min decrements in performance when sleep-deprived (Bocca version of the Johns Test of Vigilance (JTV; Johns et al.
). This simple reaction time task involves a push The aim of the present study was to assess any button response to a brief change in shape of three circles differences in a newly developed drowsiness score based presented on a computer screen that occurred at random on ocular variables (JDS) after the administration of intervals of between 5 and 15 s. There were between 50 and caffeine in non-sleep-deprived individuals. As the JDS has 64 such stimuli presented per session. To ensure equal been shown to be sensitive to changes in drowsiness/ samples per participant and session, only the last 50 alertness due to sleep deprivation, it was expected to be reaction times in each session for each participant were effected by caffeine. Reaction times from a vigilance test and self reports of sleepiness were measured to support the Test sessions were scheduled at baseline (approximately validity of any changes observed in the oculography based 9 am) and then at 30, 60, 120, 180, and 240 min after JDS. These measures were also expected to be effected by administration of either 200 mg of caffeine or a placebo.
caffeine and to change in a similar manor to the JDS scores.
Participants and experimenter were blind to which capsulewas being administered. The order of administration wascounterbalanced, with half the participants taking caffeine on their first experimental day and the other taking theplacebo. Once participants began the experimental day they Twelve participants (M/F=5/7) with a mean age of 22 years were allowed to eat only apples and dry biscuits and to (SD = 4.0, range 18-29 years) were involved in the drink only water. They passed the time between testing experiment. The study was approved by the Swinburne sessions by reading and doing other quiet activities.
University Human Research Ethics Committee, and all Whilst participants performed the JTV, their eye and participants provided written informed consent. They eyelid movements were monitored by infrared reflectance attended two experimental days, which were separated by oculography (Optalert™; Johns et al. ). This system measures the relative velocity of the eyelids opening and ( p=0.07 and 0.15; see Fig. The order effect was not closing and the total duration of blinks through infrared significant [F(1, 557)=0.03, p=0.87].
transducers positioned towards the eye. Previous studies Reaction times were significantly effected by drug have determined these to be important variables, which [F(1, 7181)=138.6, p=0.00], subject [F(11, 7181)=83.6, change from when a participant is alert and performing well p=0.00], and order [F(1, 7181)=5.3, p=0.02]. Post hoc to when they are drowsy and suffering performance dependent t tests showed that reaction times were not decrement (Johns ; Johns et al. ). These variables significantly different between the drug conditions at the are combined to produce a drowsiness score from 0 to 10 baseline session (p=0.73). Reaction times were signifi- (JDS) each minute. The JDS is based on a weighted cantly shorter after caffeine than after placebo at all other combination of the aforementioned ocular variables and is testing times (p=0.00; see Fig. ). The time effect was not also sensitive to drowsiness (Johns et al. ; Johns et al.
significant [F(5, 7181)=1.1, p=0.34].
). In the present study, the Optalert™ system calcu- KSS scores were found to differ significantly between lated JDS values automatically; however, description of the participants [F(11, 125)=16.4, p=0.00] and between drug method used to determine JDS scores is available in Johns conditions [F(1, 125)=12.1, p=0.00]. Post hoc dependent et al. (As this system is self calibrating, it can take t tests showed significantly lower sleepiness 30 min after about 4 or 5 min at the beginning to produce JDS scores. In taking caffeine compared with placebo (p=0.00) but not at this study, the minimum number of JDS scores was four in any other testing times ( p>0.05; see Fig. any one 10-min session. Only the last four JDS values persession were analyzed to ensure equal samples per personand session. Participants also rated their sleepiness after each JTV on the modified Karolinska Sleepiness Scale(KSS; Akerstedt and Gilberg Overall, the results indicate that despite the participants An analysis of variance (ANOVA) was performed being rested, their alertness was increased and their separately for reaction times, JDS and KSS scores, with performance enhanced by the administration of caffeine.
the main effects being drug (caffeine vs placebo), time As expected, the JDS scores were significantly reduced (time after taking capsule), order (first vs second experi- when the participants had consumed caffeine compared mental day), and participants. Post hoc dependent t tests with a placebo. This effect was evident at 30 min and were performed for significant main effects for drug at each persisted for almost 3 h, which is consistent with the known pharmacokinetics of caffeine (Arnaud Liguori et al.
; Van Deventer et al. ).
Also as expected, the changes in JDS were similar to the changes observed in reaction times. Reaction times weresignificantly shorter after caffeine administration compared Only six of the 12 participants rated themselves as regular with placebo at each of the time intervals. Interestingly, users of caffeine. The average daily caffeine intake across reaction times remained rather stable from 60 min onwards all participants was 66 mg per day (SD =72 mg).
in the caffeine condition and did not show any evidence of Participants reported no caffeine consumption in the increasing towards the latter sessions, as would be expected specified time periods. Participants reported sleeping for7.5 h (SD=0.7 h) the night preceding an experimental day,with no differences in sleep duration between caffeine(M=7.3, SD=0.6) and placebo days [M=7.6, SD=0.8;F(1, 10)=2.07, p=0.18]. No differences in sleep durationoccurred with the order participants performed each session[F(1, 10)=1.07, p=0.33].
ANOVA revealed significant main effects on JDS scores of drug [F(1, 557)=26.00, p=0.00], time [F(5, 557)=3.00,p=0.01], and subject [F(11, 557)=22.88, p=0.00]. Post hocdependent t tests showed that JDS scores were notsignificantly different at the baseline sessions ( p=0.27)between the two drug conditions. JDS was significantlylower at 30, 60, and 120 min after taking caffeine comparedto placebo ( p=0.00, 0.01, and 0.00 respectively). These Fig. 1 Mean JDS scores after drug administration. All bars represent differences were no longer significant at 180 and 240 min the current study, the JDS, which is based on a combinationof variables including total blink duration and the velocityof eyelids opening and closing, changed with caffeineadministration. No other studies examining these ocularvariables are available; however, the current results are inline with trends towards, and significant increases, insaccade velocity with caffeine administration that have beenreported by others (Minzhong et al. ; Smith et al. The changes in JDS in the current study, however, are in contrast to most other previous research into ocularvariables, which have demonstrated no difference in pupildiameter, pupil contraction latency, and blink rate withcaffeine administration after sleep restriction or deprivation Fig. 2 Mean reaction times after drug administration. All bars possible explanation as to why others have not observedchanges in ocular variables after caffeine administration if the reaction times paralleled the pharmacokinetic prop- may be due to the sensitivity of the measures they used to detect changes in alertness generally. Ocular variables such The self-reported levels of sleepiness measured with the as blink rate and saccade velocity have not proven very KSS were not as sensitive to caffeine as the other variables.
effective in detecting changes in alertness/drowsiness Self-rated sleepiness only showed significant differences caused by sleep deprivation (Bocca and Denise ; between drug conditions 30 min after capsule administra- Caffier et al. and blink rate has also been shown to tion. This may have been due to the ability of the KSS to be largely effected by individual differences and task detect small changes in arousal in relatively alert partic- (Caffier et al. ; Stern et al. ), so it is not surprising ipants and may be even due to the significant variability that they have demonstrated limited or no changes after caffeine administration. The Optalert™ system and JDS do The current behavioral results are in line with those not use such variables. A possible additional advantage of found in the sleep-deprived state, which have shown the JDS for detecting changes in alertness due to caffeine vigilance and reaction times to improve after the adminis- administration may be attributed to the use of several ocular tration of caffeine, when compared to placebo (Bonnet et al.
; Patat et al. Wesensten et al. The reaction Many researchers have indicated the importance of not just time results here are also comparable with studies of alert relying on one ocular variable to measure drowsiness people showing improvements after caffeine administration (Heitmann et al. ; Morris and Miller Van Orden (Lorist et al. ; Smith et al. and summarized in et al. ). It may be this combination of variables used to generate the JDS that gives the measure more sensitivity to Previous research regarding the effects of caffeine on smaller changes in alertness than would individual measures ocular variables is limited and generally finds no effect. In such as blink rate or pupillary response. This study showsthat even when people are reporting hours of sleep at nightthat are considered to be adequate, their alertness levelsduring the day can still be improved by caffeine. It iscommonly believed that the only known direct effect ofcaffeine is on adenosine receptors (Fredholm ;Fredholm et al. ); however, it is possible that theindirect effects of caffeine also contributed to the effects seenhere on reaction times and JDS scores.
The current study involved participants that usually ingested only small amounts of caffeine, if any. Thus, thedifferences in performance, JDS, and KSS seen betweencaffeine and placebo in this study are accredited to theeffects of caffeine on the CNS rather than reversal ofcaffeine withdrawal suggested to occur in habitual caffeineusers (James and Rogers As a consequence of using Fig. 3 Mean KSS scores after drug administration. All bars representSEM low or non-users of caffeine, the dose given to participants was much higher than they would usually have in daily life.
Akerstedt T, Gilberg M (1990) Subjective and objective sleepiness in For the purpose of the current study, this dose illustrated the the active individual. Int J Neurosci 52:29-37 Arnaud MJ (1987) The pharmacology of caffeine. Prog Drug Res predicted relationships; however, the effects on JDS in particular may be different with a smaller amount of Beaumont M, Batejat D, Pierard C, Coste O, Doireau P, Van Beers P, caffeine or in those that have developed a tolerance to Chauffard F, Chassard D, Enslen M, Denis JB, Lagarde D (2001) caffeine due to regular caffeine consumption. The effects of Slow release caffeine and prolonged (64-h) continuous wakeful-ness: effects on vigilance and cognitive performance. J Sleep Res multiple doses of caffeine may also differ. These points could be a focus of future research. Additionally, the Bocca ML, Denise P (2006) Total sleep deprivation effect on current study only recorded sleep length for the night prior disengagement of spatial attention as assessed by saccadic eye to each experimental day. Thus, it is possible that Bonnet MH, Balkin TJ, Dinges DF, Roehrs T, Rogers N, Wesensten participants had not been obtaining adequate sleep previous NJ (2005) The use of stimulants to modify performance during nights and had some level of sleep debt. This could have sleep loss: a review by the sleep deprivation and stimulant task resulted in participants having some residual adenosine that force of the American Academy of Sleep Medicine. Sleep had not been eliminated with sufficient sleep and must be Caffier PP, Erdman U, Ullsperger P (2003) Experimental evaluation of considered when interpreting results.
eye-blink parameters as a drowsiness measure. Eur J Appl In conclusion, the present study gives insight into the effects of caffeine on JDS scores, a measure of alertness/ Food Standards Australia New Zealand (2006) Caffeine drowsiness based on ocular variables, in alert individuals, Fredholm BB (1995) Astra award lecture: adenosine, adenosine receptors and the actions of caffeine. Pharmacol Toxicol 76:93- most of whom consumed either no caffeine or very little caffeine on a regular basis. The study demonstrated that a Fredholm BB, Battig K, Holmen J, Nehlig A, Zvartau EE (1999) moderate dose of caffeine in such individuals can reduce Actions of caffeine in the brain with special reference to factors JDS scores, which are a measure of drowsiness. Differences that contribute to its widespread use. Pharmacol Rev 51:83-133 Heitmann A, Guttkuhn R, Aguirre A, Trutschel U, Moore-Ede M in findings between this study and previous research (2001) Technologies for the monitoring and prevention of driver utilizing different ocular variables to measure alertness/ fatigue. In: Proceedings of the First International Driving drowsiness may be attributed to the use of ocular variables Symposium on Human Factors in Driver Assessment, Training that are less sensitive to changes in alertness/drowsiness by and Vehicle Design. Aspen, CO, USA, pp 81-86 Horne JA, Reyner LA (1996) Counteracting driver sleepiness: effects others. Additionally, the failure of others to combine ocular of napping, caffeine, and placebo. Psychophysiology 33:306-309 variables into one index may have influenced differences in James JE, Rogers PJ (2005) Effects of caffeine on performance and results. The changes in reaction times caused by caffeine mood: withdrawal reversal is the most plausible explanation.
administration in the present study were also similar to Johns M (1991) A new method for measuring daytime sleepiness: the those observed in JDS. Self-rated sleepiness (KSS scores) Epworth sleepiness scale. Sleep 14:540-545 on the other hand did not appear to be very sensitive to Johns M (2003) Eyelid closure, visual suppression and hypovigilance these changes in alertness. Overall, the results of this study in the drowsy state: lapses in performance with eyes open or demonstrated that administration of caffeine can signifi- Johns M, Tucker A, Chapman R, Michael N, Beale C (2006) A new cantly increase alertness and enhance performance in alert scale of drowsiness based on multiple characteristics of blinks: and well-rested individuals, and these changes can be the Johns drowsiness scale. Sleep 29(suppl):A365 Johns M, Tucker A, Chapman R, Crowley K, Michael N (2007) Monitoring eye and eyelid movements by infrared reflectanceoculography to measure drowsiness in drivers. Somnologie Crowley for critical review of this manuscript, the anonymous reviews Liguori A, Hughes JR, Grass JA (1997) Absorption and subjective for their feedback, and Dr Andy Tucker for performing additional post effects of caffeine from coffee, cola and capsules. Pharmacol Sleep Diagnostics Pty Ltd is a Research and Development Lorist MM, Snel J, Kok A, Mulder G (1994) Influence of caffeine on Company that provided financial support for this research by way of selective attention in well-rested and fatigued subjects. Psycho- salaries for Dr Johns and for Miss Michael during the data collection and analysis phase. Dr Johns is also a share holder of the company.
Morris TL, Miller JC (1996) Electrooculographic and performance Furthermore, this study was performed in compliance with the indices of fatigue during simulated flight. Biol Psychol 42:343- Minzhong Y, Russo M, Johnson D, Kamimori G (2004) Oculomotor measures over 30-hours of continuous wake with and withoutcaffeine. Sleep 27(Suppl):A138 Parsons WD, Neims AH (1978) Effect of smoking on caffeine clearance. Clin Pharmacol Ther 24:40-45 Abernethy DR, Todd EL (1985) Impairment of caffeine clearance by Patat A, Rosenzweig P, Enslen M, Trocherie S, Miget N, Bozon MC, chronic use of low-dose oestrogen-containing oral contraceptives.
Allain H, Gandon JM (2000) Effects of a new slow release formulation of caffeine on EEG, psychomotor and cognitive functions in sleep-deprived subjects. Hum Psychopharmacol and blood gastrin after coffee consumption. Dig Dis Sci 37: Smith A, Brice C, Nash J, Nutt DJ (2003) Caffeine and central Van Orden KF, Jung T-P, Makeig S (2000) Combined eye activity noradrenaline: effects on mood, cognitive performance, eye measures accurately estimate changes in sustained visual task movements and cardiovascular function. J Psychopharmacol Vanderveen JE, Armstrong LE, Butterfield GE, Chenoweth WL, Smith A, Sutherland D, Christopher G (2005) Effects of repeated Dwyer JT, Fernstrom JD, Kanarek RB, Levander OA, Sternberg doses of caffeine on mood and performance of alert and fatigued EM (2001) Caffeine for the sustainment of mental task volunteers. J Psychopharmacol 19:620-626 performance: formulations for military operations. National Stern JA, Walrath LC, Goldstein R (1984) The endogenous eye blink.
Wesensten NJ, Killgore WDS, Balkin TJ (2005) Performance and Van Deventer G, Kamemoto E, Kuznicki JT, Heckert DC, Schulte alertness effects of caffeine, dextroamphetamine, and modafinil MC (1992) Lower esophageal sphincter pressure, acid secretion, during sleep deprivation. J Sleep Res 14:255-266

Source: http://www.mwjohns.com/wp-content/uploads/2008/09/michael_et_al_2008_effects_of_caffeine_on_alertness_as_measured_by_infrared_reflectance_oculography.pdf