Psychopharmacology (2005) 180: 539���547 DOI 10.1007/s00213-005-2191-9 ORIGINAL INVESTIGATION Hiske van Duinen . Monicque M. Lorist . Inge Zijdewind The effect of caffeine on cognitive task performance and motor fatigue Received: 19 October 2004 / Accepted: 12 January 2005 / Published online: 19 February 2005 # Springer-Verlag 2005 Abstract Rationale: In everyday life, people are usually capable of performing two tasks simultaneously. Howev- er, in a previous study we showed that during a fatiguing motor task, cognitive performance declined progressive- ly. There is extensive literature on the (positive) effects of caffeine on cognitive and motor performance. These effects are most pronounced under suboptimal conditions, for ex- ample during fatigue. However, little is known about the effects of caffeine on cognitive performance during a fatigu- ing motor task. Objective: This study was aimed to inves- tigate whether a moderate dose of caffeine could attenuate the decline in cognitive performance during a fatiguing motor task. Methods: The study consisted of a placebo and a caffeine (3 mg/kg) session. A total of 23 subjects com- pleted these sessions in a semi-randomized and double-blind order. In each session, subjects performed maximal volun- tary contractions of the index finger, a choice reaction time (CRT) task and a dual task consisting of a fatiguing motor task concomitantly with the same CRT task. After the fa- tiguing dual task, the CRT task was repeated. Results: Caf- feine improved cognitive task performance, in both the single and dual task, as shown by decreased reaction times together with unchanged accuracy. Cognitive performance in the dual task deteriorated with increasing fatigue. How- ever, the decrease in cognitive performance in the begin- ning of the dual task, as observed in the placebo condition, was partly prevented by caffeine administration (i.e., no increase in reaction times). We found no effects of caffeine on motor parameters (absolute force, endurance time or electromyographic amplitude). Conclusions: Caffeine im- proved cognitive performance. This effect also extends under demanding situations, as was shown by the perfor- mance during the dual task, even during progressive motor fatigue. Keywords Caffeine . Cognitive performance . Fatigue . Dual task . Maximal voluntary contraction . Submaximal contractions . Reaction time . Accuracy Introduction Coffee is a widely consumed beverage and it is estimated that about 90% of Dutch adults drink coffee regularly (Hameleers et al. 2000). Coffee contains caffeine, which is known to have stimulatory effects on the central nervous system. The stimulating effects of caffeine are predom- inantly caused by an antagonistic action on adenosine re- ceptors. Hence, caffeine increases the levels of several neurotransmitters such as dopamine, acetylcholine and se- rotonine (for review, see Fredholm et al. 1999). In daily life, caffeine is commonly used to suppress feelings of fa- tigue. Furthermore, after caffeine consumption, subjective feelings such as increased alertness, energy and ability to concentrate are often cited. The effects of caffeine are most pronounced when subjects perform under suboptimal con- ditions, characterized by fatigue or tediousness, or in tasks placing high demands on the information processing system (Lieberman et al. 1986 Lorist et al. 1994 Ruijter et al. 1999). The simultaneous execution of two tasks is a condition that places high demands on the processing system. Several studies showed that in a cognitive dual-task condition, caffeine has a positive effect on performance (Brice and Smith 2002 Ruijter et al. 1999). By using a dual-task paradigm, we previously showed a mutual interference between motor and cognitive perfor- mance during a protocol that induced progressive amounts of muscle fatigue (Lorist et al. 2002). In this study, subjects H. van Duinen (*) . M. M. Lorist . I. Zijdewind Department of Medical Physiology, University of Groningen, Groningen, The Netherlands e-mail: h.van.duinen@med.umcg.nl Tel.: +31-50-3638735 Fax: +31-50-3632751 M. M. Lorist Department of Experimental and Work Psychology, University of Groningen, Groningen, The Netherlands

performed submaximal contractions until they could no longer maintain the contraction at the desired force levels. Concomitantly with this fatiguing motor task, the subjects had to execute a choice reaction time (CRT) task. The reaction times and the number of errors served as a measure of cognitive performance. The growing amount of fatigue was accompanied by an increase in both reaction times and number of errors. We attributed the progressive decline in cognitive performance to an increasing demand on central resources by the motor task. Thus, fewer resources were available for the cognitive task performance. As caffeine seems to have an effect on the availability of central re- sources, the investigation of this interference���between motor fatigue and cognitive task performance���seems to be a logical direction for further research. Aside from its beneficial effects on cognitive perfor- mance, caffeine is also known to affect motor performance. The effects of caffeine on short-term exercise, however, are not consistent. Some studies show an increase in endurance time after caffeine consumption (Jackman et al. 1996 Kalmar and Cafarelli 1999 Plaskett and Cafarelli 2001), while others show no effect (Lopes et al. 1983 Williams et al. 1987). In addition, the effects of caffeine consump- tion on muscle force are contradictory. Several studies fail to show significant ergogenic effects (Lopes et al. 1983 Williams et al. 1987) however, in two well-controlled studies, Cafarelli et al. (Kalmar and Cafarelli 1999 Plaskett and Cafarelli 2001) found an increase in muscle force after caffeine administration. The use of caffeine in studies is not very consistent. There are differences in duration of caffeine abstinence, dose, and way of administration. In some studies, subjects have to abstain from coffee for just 1 h before the experiment, while other studies demand abstinence from caffeine for several days. Since it was shown that caffeine has a positive effect in subjects who had only been minimally deprived of caffeine (Christopher et al. 2005 Warburton 1995 Warburton et al. 2001), we decided to allow the participants to drink their morning cup of coffee. In this way, the detrimental effects of caffeine withdrawal symptoms were prevented and the daily ritual of subjects would not be affected (Lane and Phillips- Bute 1998 Rogers et al. 2003). In previous studies, the doses used varied from a single dose of 32 mg (Lieberman et al. 1987) up to 1,400 mg of caffeine (Streufert et al. 1997), or body weight-related doses of 1 (Yeomans et al. 2002) to 13 mg/kg (Pasman et al. 1995). According to Graham (2001), the optimal dose of caffeine is between 3 and 6 mg/kg body weight. To mimic everyday life as much as possible, we used a dose of 3 mg/kg (equivalent to about two cups of coffee) and dissolved the caffeine in decaffeinated coffee. We conducted this study to investigate the effect of a moderate dose of caffeine on performance during a cog- nitive task and on motor parameters, force and endurance time. Furthermore, we used a dual-task paradigm to in- vestigate whether a moderate dose of caffeine could atten- uate the decline in cognitive performance observed during a fatiguing motor task. The preliminary results were pre- sented in abstract form (van Duinen et al. 2003). Materials and methods Participants A total of 24 healthy adults (mean age 24��6 years, 11 males, 13 females) participated in this study. All subjects were right-handed and non-smokers they usually drank three to six cups of coffee per day (or an equivalent of that amount of caffeine from other dietary sources). The sub- jects had normal or corrected-to-normal vision and intact hearing. Each subject signed a written informed consent prior to the study. All procedures were undertaken with the approval of the local ethics committee (METc, Academic Hospital Groningen) and were performed in accordance with the standards set out in the Declaration of Helsinki. Experimental set-up Each participant sat at an experimental table the lower arms rested on the table with the elbows in an angle of 135��. The right hand was positioned halfway between pronation and supination. The hand and lower arm were immobilized with pressure plates and Velcro tape. To mea- sure the abduction force of the first dorsal interosseus muscle (FDI) of the right hand, the proximal interphalan- geal joint of the index finger was inserted (slightly ab- ducted) in a snugly fitting ring that was rigidly connected to an isometric force transducer (for details, see Zijdewind and Kernell 1994). Electromyographic (EMG) recordings were obtained from the FDI of both hands with a surface electrode (diameter, 4 mm) placed over the muscle belly and a reference electrode placed at the metacarpophalan- geal joint of the index finger. A band-shaped earth elec- trode was placed around the right wrist. EMG and force recordings were amplified, filtered (EMG, 10 Hz���1 kHz force, DC-500), and sampled via a PC equipped with a data-acquisition interface (1401+, Cambridge Electronic Design, Cambridge, UK). The sampling rates were 2,000 and 500 Hz for EMG and force recordings, respectively. Off-line analysis was performed with custom-made scripts (Spike2, Cambridge Electronic Design, Cambridge, UK). Tasks The subjects performed three tasks: a motor task, a cog- nitive task, and a dual task. The motor task consisted of voluntary abductions of the right index finger. The contractions could either be maxi- mal (MVC) or submaximal (30% cMVC). At the start of the experiment, subjects performed three maximal volun- tary contractions (MVCs 4 s) of the index finger (ab- duction) at approximately 1-min intervals (Enoka et al. 1989 Fuglevand et al. 1993 Zijdewind and Kernell 1994). If the difference between the peak forces of two consecutive MVCs exceeded 5%, a subsequent trial was performed. The strongest contraction was used as the ���control MVC��� (cMVC). During the cognitive tasks (see 540