L-theanine

[Lorri Dent]

chamomile, rosebuds, hibiscus, lavender, licorice root, butterfly pea flower, tulsi, l-theanine, passionflower, orange blossom, natural rose flavor with other natural flavors, rosehips, apple pieces, elderberries, raspberries

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Attention deficit hyperactivity disorder (ADHD) is characterized by impaired attention, hyperactivity and increased impulsivity1. Impaired sustained attention, operationally defined as the inability to maintain voluntary focus on events that occur infrequently for a sustained period of time, is a major and extensively studied cognitive deficit associated with ADHD2,3,4. Similarly, multiple studies and meta-analyses have shown that impulse control is impaired in ADHD5. This repeatedly observed phenomenon led to the theory that lack of inhibitory control may be the primary cognitive deficit in ADHD6. In addition to impairments in sustained attention and impulse control, children with ADHD may also experience deficits in multiple executive functions including planning, visual and verbal working memory, and task-switching2. Thus, ADHD seems to affect multiple cognitive domains and these deficits may be contributing to academic underachievement that is associated with ADHD7.

Impaired sustained attention and inhibitory control in patients with ADHD is associated with increased task-associated mind wandering8,9,10,11,12. Increased mind wandering is negatively correlated with ADHD symptom severity8,9. Multiple functional MRI (fMRI) studies have established that a functional brain network, known as the default mode network (DMN), which includes the medial prefrontal cortex (anterior node of DMN), posterior cingulate cortex and precuneus (posterior node of DMN), shows increased reactivity during mind wandering even in neurotypical individuals10. Comparison of fMRI brain reactivity neurotypical vs. ADHD children while engaged in a sustained attention task revealed that children with ADHD show increased task-related reactivity of the posterior DMN11. In another study, high level of impulsivity was associated with decreased reactivity of both anterior and posterior DMN12. Thus, increased reactivity of DMN is an established phenomenon contributing to impaired sustained attention and inhibitory control among individuals with ADHD. In addition, functional neuroimaging surrogates of impaired inhibitory control in ADHD include decreased reactivity of several regions of a network known as the central executive network (CEN) (e.g. bilateral dorsolateral prefrontal cortex and inferior frontal gyrus)13,14,15 Accordingly, treatment of individuals with ADHD with stimulant medications (e.g. methylphenidate) acutely decreased the task-related hyperactivity of DMN while enhancing the reactivity of CEN16,17.

Even though several classes of medications are available to manage ADHD (e.g. stimulants such as methylphenidates and amphetamines; and non-stimulant medications such as atomoxetine, guanfacine and clonidine), intake of these medications are commonly associated with adverse effects ranging from loss of appetite, and irritability to suicidal ideation18,19 and therefore are not well-tolerated20. Moreover, poor responsiveness, pharmacological tolerance, risk of poisoning, high risk of misuse and improper use (e.g. despite contraindications) are common concerns of ADHD medications21,22,23. Therefore, there is an unmet need for safe, effective alternatives with known mechanisms of action to manage ADHD.

Taken together, l-theanine, caffeine and their combination have been observed to improve sustained attention in healthy adults. Yet, the risk of development of adverse effects with low-moderate doses l-theanine or caffeine is minimal. Thus, l-theanine, caffeine and particularly their combination have the potential to be translated to manage cognitive deficits associated with ADHD. We aimed to examine the effects of l-theanine, caffeine and their combination on sustained attention, impulse control and overall cognition in male children with ADHD in a randomized placebo-controlled four-way repeated measures crossover study. We further aimed to concurrently explore the neurophysiological mechanisms of action of l-theanine, caffeine and their combination using fMRI, which provides a means of deducing neurophysiological processes that are associated with changes in cognition and behavior46. We hypothesized that l-theanine, caffeine and their combination would improve sustained attention and impulse control by decreasing task-related activity of DMN in the brain, which is thought to be associated with mind wandering. We further hypothesized that l-theanine, caffeine and their combination would improve impulse control by increasing reactivity of inferior frontal and dorsolateral prefrontal cortical regions of the brain that are thought to be the neural substrates of impulse control.

Potential participants/families expressing interest were subjected to a pre-screening telephone interview. Children with gross visual, hearing, intellectual, neurological or psychiatric impairments (except ADHD) that could affect performance in cognitive/neuropsychological test batteries were excluded. Similarly, children who were on medications except for stimulants that may affect cognitive functions and children who were on medications that may interact with caffeine were excluded. Given that the study involved a neuroimaging component, children with absolute contraindications to undergo magnetic resonance imaging were also excluded.

The participants were allowed to practice the Go/NoGo task and the Stop-signal task for 10-min following administration of the substances. Then, the participants relaxed until 55 min had elapsed from the time of administration of the substance. Next, a scanning session was performed on a 3 T Siemens Skyra scanner equipped with a 20-channel head coil (see Supplementary Information 1 for the detailed scanning protocol including acquisition parameters). During the scanning session, first a 12.5-min functional scan was administered while the participants performed the Go/NoGo task. Second, another 12.5-min functional scan administered while the participants performed the Stop-signal task. The Go/NoGo task51,52 and the Stop-signal task53 were programed and presented using PsychoPy 2.0 (University of Nottingham, UK, see Supplementary Information 1 for the task description). The stimuli for the tasks were presented on an LCD screen and projected onto a mirror attached to the head coil of the scanner. The participants used a two-button fiber-optic hand-held device to respond to stimuli. The scanning session also included a 4.5-min T1-weighted structural scan.

All behavioral data were analyzed via two-level models (i.e. testing visits nested within subjects) constructed using the lmerTest package in R statistical software (version 3.4.2). Each behavioral outcome was regressed on a dummy variable coded for the administered substance. Placebo condition was included in all models as the reference category. Testing visit number was included in all models as a covariate to account for the practice effect. In the Go/NoGo task, rate of correctly-responded Go trials (i.e. hit rate), rate of responded NoGo trials (i.e. false alarm rate), sensitivity to the Go signal (i.e. d-prime) computed based on the signal detection theory62,63, and reaction time to Go trials were regressed on the administered substance. In the Stop-signal task, mean reaction time to correctly-responded Go trials within a testing visit, rate of correctly-inhibited Stop trials (i.e. inhibition rate), mean Stop-signal delay and Stop-signal reaction time (SSRT) computed based on the horse-race model64 were considered as outcome variables. Age-adjusted total cognition composite scores obtained from NIH Cognition Toolbox Test Battery were similarly regressed on the administered substance.

Raw structural and functional MRI data were converted to NIfTI format using dcm2nii converter65 and were pre-processed using tools in Freesurfer66,67 and FSL (version 6.0, Oxford, UK) software (see Supplementary Information 1 for details on pre-processing). Functional MRI data of the Go/NoGo and Stop-signal tasks were analyzed via two-level models constructed using the FEAT tool in FSL (see Supplementary Information 1 for the detailed analysis protocol). In level 1 analyses, contrasts were modeled to examine the brain reactivity to correctly-responded Go trials and correctly-inhibited NoGo trials in the Go/NoGo task and inhibitory control in the Stop-signal task. In level 2 analyses, level 1 contrasts for each treatment was compared against the placebo within DMN and CEN, maintaining FWER at 0.05 via a permutation-based cluster thresholding approach.

Four participants were not on stimulant medications during the period in which the testing visits were scheduled (i.e. summer vacation), yet all participants reported symptomatic benefits of taking methylphenidate hydrochloride during the school year. One participant was on a methylphenidate hydrochloride daily (PRN) dose during the period of testing and reported having treatment free days during school vacation as directed by his physician. As such, the participant or his parents did not complain of adverse effects of withholding methylphenidate prior to testing sessions. Even though the participants were allowed to take stimulant medications upon completion of testing sessions, participants or parents did not complain of any symptoms suggestive of adverse effects of l-theanine/caffeine treatment or potential interactions between l-theanine or caffeine with methylphenidate.

Even though we observed improvements in reaction times (i.e. faster reaction times) of sustained attention tasks with the administration of l-theanine, caffeine and their combination among healthy adults in our previous studies, we did not observe improvements in hit rate or the error rate26,33. However, in the present study, l-theanine, caffeine and their combination improved the hit rate, but not reaction times in the Go/NoGo task. These contrasting findings are not surprising given that the performance of any cognitive task is subjected to a speed-accuracy tradeoff69. That is, a certain task may have a ceiling or flooring effect when only the accuracy rates are concerned, but may reflect the differences of cognitive function of interest among participants as differences of speed (i.e. reaction times). Conversely, a task with ceiling or flooring effects on reaction times may differentiate participants based on accuracy rates. The Go/NoGo task presented in the current study51,52 is a perfect example of the latter, given the task by design limited the time the participants had to react to each stimulus, thereby reflecting the differences in sustained attention among participants as differences in accuracy rates. Others have also observed significant improvements in accuracy rates or hit rates, but not reaction times with the administration of l-theanine and caffeine to healthy adults (e.g. Ref.25,70. Therefore, the improvements in hit rates observed with l-theanine, caffeine and their combination in the Go/NoGo task and more importantly the improved d-prime seen with the combination could be considered to reflect underlying improvements of sustained attention.