Dietary Caffeine Consumption Modulates fMRI Measures Paul J. Laurienti,* Aaron S. Field,* Jonathan H. Burdette,* Joseph A. Maldjian,* Yi-Fen Yen,† and Dixon M. Moody* *Department of Radiology and †Department of Medical Engineering, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157-1022 Received March 7, 2002 Caffeine is the most widely used stimulant in the world. The stimulant effects of caffeine are mediated through its antagonistic properties on neuronal aden- osine receptors. In addition, caffeine blocks neurovas- cular adenosine receptors and decreases cerebral per- fusion. Although the effects of caffeine on blood oxygenation level-dependent (BOLD) functional mag- netic resonance imaging measures are extremely im- portant, there are few studies addressing this issue in the literature. Because chronic caffeine use causes an upregulation of adenosine receptors, the differential effects of caffeine in low and high users is of particular interest. The present study was designed to test the hypothesis that caffeine has differential effects on the BOLD signal in high and low caffeine users. We dem- onstrated that the BOLD signal change in visual cor- tex was significantly greater in high users than in low users in the presence of caffeine. In addition, the mag- nitude of the BOLD signal was significantly correlated with caffeine consumption. We propose that the out- come observed here was due to an upregulation of adenosine receptors in high users, resulting in differ- ential contributions of the neural and vascular effects of adenosine in the two study populations. © 2002 Elsevier Science (USA) INTRODUCTION Caffeine is the most widely used neurostimulant in the world and is found in a variety of foods and bever- ages, including coffee, tea, soft drinks, and chocolate. Caffeine consumption has been estimated at 76 mg/ person per day worldwide, as high as 238 mg/person/ day in the United States, and more than 400 mg/ person/day in Sweden and Finland (Nehlig, 1999). Caffeine is a nonselective antagonist of adenosine re- ceptors (Ralevic and Burnstock, 1998) and acts not only as a neurostimulant but also as a vasoconstrictor. Re- cently it has been reported that caffeine enhances the blood oxygenation level-dependent (BOLD) signal in functional magnetic resonance (fMRI) studies (Mul- derink et al., 2002; Parrish et al., 2001). However, it is well known that chronic caffeine use causes an upregu- lation of adenosine receptors (Johansson et al., 1993; Ralevic and Burnstock, 1998), and no study has eval- uated the effects of caffeine on BOLD responses while controlling for dietary caffeine consumption. The neurostimulant effects of caffeine are due to the antagonism of neural adenosine receptors. Adenosine has a general inhibitory effect on neural activity me- diated by both A 1 and A 2A receptors (Dunwiddie and Masino, 2001). However, neural A 1 receptors are widely distributed throughout the brain, whereas A 2A receptors are localized primarily in the striatum, nu- cleus accumbens, and olfactory tubercle (Dunwiddie and Masino, 2001; Moreau and Huber, 1999). There- fore, adenosine-induced neural inhibition is thought to be primarily due to the activation of A 1 receptors (Dun- widdie and Masino, 2001; Haas and Selbach, 2000). Adenosine-induced vasodilatation and subsequent in- creased cerebral perfusion is mediated primarily by A 2A on the neurovasculature (Coney and Marshall, 1998; Feoktistov and Biaggioni, 1997; Ngai et al., 2001; Shin et al., 2000). Although each of these receptor subtypes can be upregulated following chronic caffeine consump- tion, it has been demonstrated that A 1 receptors are more prone to upregulation following chronic blockade than A 2A receptors (Johansson et al., 1993; Ralevic and Burnstock, 1998; Shi and Daly, 1999). In addition, it has been suggested that the effect of caffeine on BOLD signal changes may be dependent on chronic caffeine usage and withdrawal due to the fact that chronic caffeine use results in an upregulation of adenosine receptors (Dager and Friedman, 2000; Johansson et al., 1993; Mulderink et al., 2002; Ralevic and Burnstock, 1998). In fact, recent studies have demonstrated that chronic use and withdrawal modulate the stimulant effects of caffeine, associated electroencephalography changes, and resting cerebral perfusion (Field et al., 2002; Jones et al., 2000; Mathew et al., 1983; Mathew and Wilson, 1985; Swerdlow et al., 2000). However, no study has directly examined the effect of chronic caffeine use and/or withdrawal on stimula- tion-induced changes in the BOLD signal. In the cur- NeuroImage 17, 751–757 (2002) doi:10.1006/nimg.2002.1237 751 1053-8119/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.