68 healthy high school students (27 females, 41 males), of two different 6th grades from the same school (that shared the same teacher) volunteered for the study. Their ages ranged from 17 to 18 years. None of the participants reported having any history of neuropsychiatric disorders, use of drugs, being sick during the experiment, and did not take any medication at the time of the experiment. Their parents or legal guardians signed a written informed consent approved by the Biomedical Research Ethics Committee of the Alberto C. Taquini Institute, previously to their children's participation in the study.
Data from 23 participants were excluded from the study because they did not participate in the testing session on day 8 (16), and did not match the learning criteria of 30% of correct responses on day 1 (7). Thus, the final sample was 45 participants. Unlike Neumann et al (2020) who present the odor all night long14, our odor was present for 1.5 h to guarantee the cueing to be present during the first cycle of NREM sleep2.
The study was conducted in 2 days, separated by a seven-day interval. On day 1, the classrooms were odorized using spray dispensers before the students entered the room. Both courses received a history lesson (training session) of 20 minutes provided by their own teacher (the same for the two courses) in the presence of coconut odor. To perfume the classroom we used 1 spray dispenser per 3 students. The history lesson was about the city of Petra and the Nabatean culture and commerce. When the teacher finished the lesson the spray dispensers were turned off. Afterward, the students resolved a multiple-choice exam of 13 questions about the lecture (short-term testing session). After finishing the exam, they completed the State Anxiety Inventory and the Pittsburgh Sleep Quality Index.
In each classroom, students were randomly assigned to the Reactivation or the No reactivation groups, and each one of them was provided with an automatic spray dispenser according to the group they were assigned: the Reactivation group was given the same odor that was presented in the classroom during the lesson (coconut), and the No reactivation group was provided with a different odor (violets). They were instructed to use the spray dispensers in their houses the night of the first experimental day (reactivation session). They were told to set the initiation time of the dispenser at the time they went to sleep, and the finishing time 1.30 h after. They returned the dispensers the following day. On day 8, they resolved another multiple-choice exam in the classroom (long-term testing session). After finishing the exam (without a time limit) they completed the State-Trait Anxiety Inventory and the Pittsburgh Sleep Quality Index.
Training (day 1). The training consisted of a history lesson about the ancient city of Petra and the Nabatean culture and commerce, it was presented by their teacher of Political Economics. She used a PowerPoint presentation as a tool to provide the lecture, and it was shown to the students in the classroom using a projector. Each slide was presented for 1.5 minutes as she commented on them orally.
Short-term testing session (day 1). Immediately after training, students resolved a multiple-choice exam consisting of 13 questions with 4 possible answers to each question and one correct answer. The answers to all of the exam questions were among the information written in the slides. They had no time limit to resolve the exam. Half the students of each course received the same set of 13 questions and the other half a different set of questions.
Reactivation (day 1). Memory cueing during sleep was carried out the first night of the experiment. The students received a dispenser that was programmed to spray the congruent (coconut) or incongruent (violet) odor every 15 minutes depending on the group that they were assigned. Volunteers were instructed to activate the dispenser at the time they went to sleep at night of day 1 and to set the finishing time 1.5 h after that. It is important to highlight that the participants did not know which odor they received.
Long-term testing (day 8). One week later in the classroom and without any odor, each subject resolved a different multiple-choice exam from the one that they answered on day 1. Again, they had no time limit to resolve the exam. In each classroom, half the students received the questions that their partners had resolved the week before and vice-versa.
As this study was carried out during COVID-19 pandemic, when the levels of anxiety in the population were increased and a worse quality of sleep was reported19–21, we controlled that both groups had a similar level in both variables. For that, we evaluated the State-Trait Anxiety Inventory (STAI)22 and the Pittsburgh sleep quality index (PSQ)23.
Subjects were randomly assigned to one of 2 conditions: “Reactivation” and, “No reactivation” groups.
“Reactivation” group (R, n = 22). Participants were trained on day 1 on the presence of coconut odor. They received the reactivation using the same odor cue during the first night of sleep and they were finally tested on day 8.
“No reactivation” group (NR n = 23). Participants were trained on day 1. They received the reactivation session using a different odor (violets), as a control condition, during the first night of sleep and they were finally tested on day 8.
The statistical analysis was made with SPSS version 25 (IBM Corporation). The scores reached at short and long-term testing sessions were calculated as the percentage of correct answers for the short and long-term tests, respectively. We also calculated the normalized memory change for each subject as [(# of correct responses at long-term testing session - # of correct responses at short-term testing session)*100/ # of correct responses at short-term testing session].
The scores at short and long-term were analyzed with repeated measures ANOVA with “group” as between-subjects factor with two levels (Reactivation and No reactivation) and “session” as within-subjects factor with two levels (short and long term testing). Furthermore, we compared the memory change between groups with a two-tailed t-test and performed two separately one-sample t-tests to compare both groups’ memory change to the value zero. Alpha was set at 0.05.
We also analyzed the State Anxiety Inventory, Trait Anxiety Inventory and PSQI scores with One-way ANOVAs with “group” as between-subjects factor with two levels (Reactivation and No reactivation). It is important to highlight that 3 students of the No reactivation group did not complete STAI at day 1, 6 the STAI at day 8, 6 the PSQI and 6 did not complete the Trait Anxiety Inventory. 2 students of the Reactivation group did not complete the STAI at day 8, 3 the PSQI and 3 the Trait Anxiety Inventory. We reported partial eta square (η2p) and Cohen’s d as effect size estimates.