The Cogni-Action Project: Physical Activity, Brain, and Cognition

Regarding the study (i.e., crossover-randomized trial), children will be randomized to each protocol session, participating in all of them with two-weeks apart . Thus, they will assist to our gym to perform one of the three protocol and subsequently they will be move to a laboratory where they will undergo EEG and eye-tracking measurements during resting state condition and during two different cognitive tasks More details can be found in following sections.

The design of the project will consist of a counterbalanced cross-over trial where the participants will perform, on different days, three different PA training protocols in a random order. Each child will visit our laboratory three times, every two weeks to ensure a "wash-out" period with the purpose of mitigating the PA effect from each protocol. Participants will assist to the three different sessions at the same weekday (e.g., Monday) and at the same daytime to avoid differences in preceding school activities or circadian rhythms. Participants will undergo a "Sedentary condition" (SC), sitting and watching a documentary on TV as has been used previously, and two different PA protocols consisting in "Moderate-Intensity Continuous Training" (MICT), i.e., continuous outdoor walking, and "Cooperative High-Intensity Interval Training" (C-HIIT), composed of a circuit training with a partner (physical education teacher). These three protocols have been chosen in view of what corresponds to types of activities usually performed in schools. Thus, SC mimics the time spent during any academic lesson (i.e. mathematics or language lessons); MICT represents the typical activity in a PE class; and C-HIIT has been selected since it is postulated as an exercise protocol that could possibly be implemented in the PE class. The three protocols are designed so that energy expenditure is equivalent between them, to eliminate differences in the energy expenditure as a possible confounding factor. The energy expenditure calculation protocol is detailed on supplementary material.

After each PA protocol, the participants will undergo EEG and eye-tracking measurements as follows: 1st, resting state spontaneous brain activity will be assessed during 3 min with EEG only, and subsequently, eye-tracking will be registered in parallel with EEG while participants perform the cognitive tasks (a working memory task and a reading task). The measurements will start between 20 to 25 minutes after the end of each condition. In this way, we will emulate a school recess in the Chilean context, before the start of the next subject. All measures will be carried out in the "Laboratorio Lenguaje \& Cognición ELV", which belongs to the Literature and Language Sciences Institute of the Pontificia Universidad Católica de Valparaíso. To avoid subjective influence of evaluator, when the participants go to the laboratory, the evaluator will be blinded according to which PA protocol participants performed.

Neuroelectric and eye-tracker measurements EEG measurement: A B-Alert X24 device for EEG (Advanced Brain Monitoring, California, United States) will be used, which consists of 24 active electrodes that minimize the noise of electrical devices outside the biological processes of interest, ensuring good signal quality. Two channels will be used to register electrooculographic activity, to better discriminate between true electroencephalographic and electrooculographic activity associated with eye movement and blinking. Recordings will be done at a sampling rate of 256 values per second, with a band-pass filter between 0.1 - 100 Hz, and a notch filter of 50 Hz to eliminate the noise of the surrounding alternating current in the room. It should be noted that this equipment is wireless, which makes it more convenient and faster to mount on participants. For this study the investigators will focus on the channels Fz, Cz, CPz, Pz, POz, Oz, O1 and O2.

Eye-tracker: A Tobii Pro TX300 (Tobii, Stockholm, Sweden) will be used to track eye movements directly through a light sensitive camera near the infrared spectrum. This equipment studies the visuomotor characteristics during the process of reading and the trajectories of the look around each word. In addition, it will serve to define the exact moments at which these words are read and thus be able to synchronize the EEG record to calculate an average signal (N400) that reflects the brain's processing of language. The same device records raw values of pupil size (diameter), saving all data with a sampling rate of 300 Hz.

Cognitive tasks during neuroelectric and eye-tracker measurements Working memory task. A N-back task will be run through E-Prime 2.0 software synchronized with the electroencephalographic recording system and the eye-tracker, to measure neuroelectric activity and pupil size diameter time-locked to the presentation of each stimulus (targets and non-targets). The protocol is very similar to the one previously explained for the 0-back and 2-back conditions during fMRI assessment. The main difference of the present task with respect to the fMRI's N-back task is that children must press in the keyboard the number "1" when they detect a target stimulus or the number "2" for non-targets. This type of task generates an EEG potential related to cognitive processing, characterized by a positive deflection that peaks roughly around 300 ms (P300), which has been widely studied in various types of tasks that require conscious attention, in any perceptual modality. In other words, it is a potential related to events that are consciously perceived, which does not depend on the stimulated perceptual path, but rather on the need to reallocate cognitive resources and develop appropriate responses according to the ever-changing context. To isolate the EEG patterns of P300 in working memory, the investigators will subtract to the result of a 2-back task, the one of a 0-back. The P300 event-related potential is expected to emerge especially in the average EEG signal time-locked to the onset of detected targets stimuli. Artifacts such as electrooculographic signals will be subtracted from the data with an algorithm implemented in EEGLAB, an EEG analysis toolbox running in MATLAB (Mathworks®). Manual rejection of single trials will be considered based on visual inspection of artifacts in data.

Reading task. Language processing will be assessed by a reading comprehension task where short paragraphs will be presented to be read normally. The eye-tracking technique will be employed to record eye position, movement and pupil diameter, simultaneous to the EEG record. Participants will seat in front of a monitor, at approximately 60 cm, placing their chin on an ergonomic stand piece, which helps to decrease head movements and improves eye-tracker signal precision. The reading task will be presented with the latest version of Tobii Pro Lab. After calibrating the system, each child is given a brief contextualization of the topic, before they start to read, knowing that later they will be asked seven comprehension questions with varying degrees of complexity. The participants must indicate the correct answer among three alternatives. It should be noted that three literary texts will be used, presented in a random fashion (1 per exercise condition). These texts will have the same difficulty, the same distribution of areas of interest, similar amount of words, and will be previously validated by a panel of experts. The reading of each word is expected to generate an EEG negative potential with latency close to 400 ms (N400). The amplitude, topographic distribution and latency of this potential are electrophysiological neuromarkers of linguistic ability in general, and the ability to construct meaning from words. The eye-tracker will be used to study the ocular trajectory during reading and establish the moments when words are read. EEG and eye-tracking will be registered simultaneously to record eye movements related to reading, giving measures such as reading speed, fixation durations, number of fixations, among others), and to measure eye fixation related EEG potentials associated with cognitive load while reading words. The reader should note that this potential coincides experimentally with the N400, so they may represent different brain processes superimposed, or be essentially the reflection of the same mechanism.

Primary outcomes in crossover acute counterbalanced study Resting state spontaneous brain activity. After the exercise protocol and before the cognitive tasks, spontaneous brain activity will be measured with EEG while children rest in a seated position, in order to measure the peak frequency and reactivity of the alpha wave. The dynamics of the alpha peak frequency (between 7 and 13 Hz) during a 3 minutes period will be calculated by means of a Hilbert transform. Then, an average value and trend of the alpha peak frequency will be calculated. It has been postulated that the higher the alpha peak frequency, the greater general cognitive capacity and processing speed are. The investigators focus on alpha waves because this brainwave has traditionally associated with different theories of attention and conscious perception. The frequency of the alpha wave has been discovered to increase after a session of cognitive tasks. Additionally, the power and frequency of alpha brainwaves prior to cognitive evaluation have been correlated to performance. If PA exerts a positive effect over cognitive performance, a single session of PA should also cause an increase in alpha peak frequency. In fact, this was found by Gutmann et al. in two consecutive studies. However, a statistical relation between PA-induced increments in alpha peak frequency and cognitive performance is missing.

Working memory performance: Hits, misses, omissions, and accuracy will be calculated for targets and non-targets of the n-back task, as well as response latency.

Reading comprehension: Seven questions with varying degree of comprehension complexity will be given after the texts are read. For every question there will be four alternatives to choose as the correct answer.

Cognitive load: The investigators group several measures on this category because all of them have been related with the concept of cognitive load. The first two come from the EEG, and the other two from the eye-tracker.

P300: The P300 event-related potential associated with the different type of stimulus in the n-back (target and non-target) will be quantified according to amplitude and latency of the peak.

N400: This linguistics-related event-related potential also will be described according to latency and amplitude. Because it will be synchronized with the eye fixation on words during the reading task, it is conceptually a type of fixation-related potential.

Rapid pupil size variations: Pupil reacts to cognitive requirements and it has been associated with increases in cognitive load during various tasks including working memory and reading. The amount of high frequency variations in pupil diameter is estimated by a novel Index of Cognitive Activity, which increases with cognitive load. This measure can be represented as a continuous variable in time thanks to the high sampling rate of the eye-tracker. The Index of Cognitive Activity will be measured during the n-back and reading tasks.

Word Fixation times: The length of visual fixations also has been associated with the processing time that the brain dedicates to elaborate a cognitive interpretation of perception. Average and standard deviation of fixations around words will be calculated for the reading task.