Chewing gum selectively improves aspects of memory in healthy volunteers.

Many people chew gum partly due to the belief that it increases aspects of mental performance, including concentration. To the best of our knowledge no empirical evidence exists to support this contention. The present experiment, therefore, examined the effects of chewing gum using a comprehensive and sensitive cognitive assessment battery and two tasks manipulating cognitive load. Heart rate responses were also measured. Seventy-five healthy adult participants (mean age 24 6 years) were randomly assigned to one of three experimental conditions (Nˆ 25 per group): `̀ chewing'' ± a piece of sugarfree chewing gum (Wrigley's Extra Spearmint) was chewed naturally and constantly throughout the procedure; `̀ sham chewing'' ± participants mimicked chewing movements in the absence of gum; `̀ quiet control'' ± no chewing behaviour was performed. Aspects of attention, working memory and long-term memory were assessed using the Cognitive Drug Research (CDR) computerised battery. Stimuli were presented on a colour monitor and, except for two written word recall tasks, responses were collected automatically using a `̀ Yes''/`̀ No'' response module. The tasks were presented in the order: Word presentation, Immediate Word Recall, Picture Presentation, SimpleReactionTime,DigitVigilance,ChoiceReactionTime, Spatial Working Memory, Numeric Working Memory, Delayed Word Recall, Word Recognition and Picture Recognition (for details, e.g. see Kennedy et al. 2000). Following the CDR battery, participants performed computerised Serial Subtractions tasks. These assess concentration and working memory and allow manipulation of cognitive load (see Scholey et al., 2001 for details). In the present study Serial Threes (involving the repeated subtraction of three from a randomly generated starting number using the computer's numeric keypad) thenSerial Sevens (subtraction of seven)were used, each for 2min. Each cognitive task outcome measure was analysed by oneway analysis of variance, with Dunnett comparisons to isolate between-group effects where appropriate. The most striking