The purpose of our study was to assess in a population of healthy subjects the effects of a commercially available guaranà preparation on psychological well-being, anxiety and mood when the product was used according to the labelled dosages and instructions. Results provide no evidence for any major effects of guaranà on the psychological tests employed. The only significant difference in favour of guaranà was found in female subjects and it was limited to just one of the six areas of psychological well-being (namely, self-acceptance), thus it cannot be excluded that it represents just a random finding. Given the null results reported, it is important to consider whether the study was adequately powered to detect significant effects if they existed. Power analysis of the results show that the least hypothetical differences between the effects of guaranà and placebo which could be detected in our study were in the range 2.1-2.8 for PWB scales and 2.2 for the SAS scale, while they were in the range 8.6-11.6 for the Bond–Lader scales. Least hypothetical differences are therefore very small for the PWB scales and for the SA scale. In particular, each of the PWB scales may range 14–84 [13–15], thus the least detectable differences would be just 3.5% of the entire interval, which is highly unlikely to be somehow clinically significant. For the SAS scale total scores range from 20–80 and clinically relevant intervals are: 20–44 normal, 45–59 mild to moderate anxiety, 60–74 marked to severe anxiety, 75–80 extreme anxiety . Intervals are therefore nearly seven-fold higher than the least detectable differences in the present study, thus it is unlikely that any significant effects could have been overlooked. As for the Bond Lader scales, least detectable differences in our study were in the range 8.6-11.6. The Bond Lader scales were used in at least one previous study showing positive results with guaranà : after acute administration of guaranà to 26 healthy subjects (18 females, 8 males) there were statistically significant increases of alert and content ratings, the size being in both cases on average around 6–7 units, a difference which could have been overlooked in our setting. Nonetheless, according to our results the average differences observed between placebo and guaranà in either alert, content and calm were in the range 1.3-3.3, therefore unlikely to be clinically relevant and in any case half or less those found in the study by Haskell et al. . It should be noted that also Kennedy et al.  examined the effects of a single dose of guaranà extract on mood using the Bond-Lader scales. They reported no signifcant effects, however they also provided no figures and it is therefore not possible to compare those results with the present findings.
Studies investigating the neurobehavioural effects of guaranà in humans so far considered mainly the acute effects. Kennedy and co-workers  used 75 mg of a dried ethanolic extract of guaranà and found positive effects on some tasks of the Cognitive Drug Research (CDR) computerised assessment and on the Serial subtraction task, and in particular improvements to speed of attention, secondary memory, serial subtractions and speed of sentence verification, but no effects on the Bond-Lader mood scales. The same group in a subsequent study  showed that 222 mg of guaranà (containing 40 mg caffeine) added to vitamin/mineral effervescent tablets improved task performance, in comparison to placebo, in terms of both increased speed and accuracy of performing the Rapid Visual Information Processing and attenuated mental fatigue associated with extended task performance. Haskell and co-workers  used a standardized guaranà extract containing 11-12% caffeine and found positive effects on secondary memory performance in the CDR computerised assessment with 37.5 and 75 mg, on alert ratings of the Bond-Lader mood scales with 300 mg and on content ratings with 37.5, 75, 150 and 300 mg. Fernandes Galduróz and de Araújo Carlini , who failed to report any effects with guaranà, evaluated 1000 mg guaranà containing 21 mg caffeine on tests of psychomotor speed and accuracy (letter cancellation), working memory (digit span and digit symbol substitution), memory (free recall and learned material) and the Mosaic test (involving visuomotor performance, planning and problem solving). No effect was also found on anxiety and quality of sleep.
Just a few studies addressed the effects of guaranà after prolonged administration. The study by de Oliveira Campos and co-workers  examined the effects of 50 mg guaranà by mouth twice daily for 21 days in 75 breast cancer patients receiving systemic chemotherapy using a double-blinded procedure and found positive results on several fatigue scales with no worsening of sleep quality or anxiety or depression. In another double-blinded, cross-over study, da Costa Miranda et al.  randomized 36 patients with breast cancer undergoing adjuvant radiation therapy to either guaranà 75 mg daily for 28 days or to placebo and found no effect on radiation-induced fatigue and depression. Fernandes Galduróz and de Araújo Carlini  administered 1 g of powdered guaranà (containing 2.1% caffeine) per day for five months to normal, elderly volunteers who were evaluated at baseline and after 3 and 5 months. Evaluations included the Mini-Mental State as screening for dementia, the Digital Span to assess immediate memory, the Free Recall for recent memory, Digital Symbol for psychomotor activity and concentration, Cancellation tests for vigilance and attention, the Mosaic test for visual and spatial organization, the Rave-Progressive Matrices for general abilities, and in addition quality of sleep and anxiety were assessed. No significant effects were detected on any of the functions tested and in addition no subjects reported any rise in libido or improved sexual performance.
Based upon available literature, it appears therefore that guaranà has been investigated mainly as a stimulant, in agreement with its chemical composition, while the possible effects of guaranà on mood have been only occasionally considered. So far only Haskell et al.  reported some positive results on mood, while Kennedy et al.  failed to observe any significant effects. Nonetheless, in both cases single doses of guaranà were studied and short-term effects were reported.
Guaranà-containing products are usually recommended for regular use: this is the main reason why our study was concerned with the effects of guaranà regularly taken according to the labeled dosages and instruction over several days. The results of our study, which found no effect of guaranà on psychological well-being, anxiety and mood, are in line with previous studies reporting no activity of guaranà [6, 9]. We decided to assess the effects of guaranà on psychological well-being since the possible increase of “feelings of well-being” is among the most common claims for guaranà-containing products in particular on the Internet, where guaranà is one of the most popular herbal ingredients sought for the purpose of increasing alertness and fitness , however to our knowledge no scientific studies existed so far which addressed this issue. In our study we used a guaranà extract commercially available and standardized to 2.5% caffeine. Enrolled subjects took a total of 1080 mg/day, according to labelled commercial instructions and corresponding to 27 mg caffeine per day. Likely explanations for lack of effects and apparent discrepancies with other studies (e.g. Haskell and co-workers  who found positive effects on some ratings of the Bond-Lader mood scales) include the possibility that the selected tasks were insensitive to the levels of treatment used, or the testing regimens were not adequate to observe the effects of guaranà. Indeed, caffeine is one of the main chemicals in guaranà seeds, its content being usually 4-8% . The half-life of caffeine however is about 5 h (see e.g. ), therefore any acute effects are likely to disappear in a short time, while the supposed long-term benefits of frequent caffeine consumption (e.g. reduced risk of type 2 diabetes and liver cancer) possibly require years of more or less regular intake (see e.g. [18, 19]): it cannot be excluded that the same applies to the possible effects of caffeine on mood, which so far never received thorough investigation .
In conclusion, the present study failed to observe any significant effect of a 5-day treatment with a commercial preparation of guaranà on psychological well-being, anxiety and mood in healthy subjects. A number of adverse events are reported with the use of herbal food supplements containing guaranà, including irritability, heart palpitations, anxiety and other central nervous system events , especially in children, adolescents, and young adults with seizures, diabetes, cardiac abnormalities, or mood and behavioral disorders or those who take certain medications . Also in view of the weak evidence available regarding the actual effectiveness of commercial products containing guaranà, caution shoud be always used and advised to potential users. Considering the increasing popularity of products containing guaranà – as well as a number of other traditional herbs – and sold as dietary supplements for health and fitness purposes, controlled studies are strongly warranted to assess their benefits in comparison to the labelled claims.