Male Sprague-Dawley rats (Daehan bio, Chungcheongbuk-do, Republic of Korea) weighing 80±10 g (postnatal day 21) were assigned to three different groups by a nonparticipating researcher. Animals were housed under conditions of controlled temperature (25°C±2°C), humidity (50%-55%), and light (12-hour light-dark cycles), with access to food and water ad libitum. In the laboratory experimental design, the animals were divided at random into three groups: vehicle (n=8), Day 1 (n=8), and Day 14 (n=8). This study was approved by the K University Animal Care and Use Committee. All the experimental procedures were performed in accordance with the animal care guidelines of the National Institute of Health and the Korean Academy of Medical Sciences.

Moderate to high doses of caffeine (approximately 100-400 mg, 2.5-10 mg/kg/day) are known to produce symptoms of nervousness, jitteriness, and fidgetiness in children and adolescents [15]. However, experiments using an animal model should consider the pharmacokinetic differences between animals and humans. In our study, the dose we administered was higher than what one would administer to a human because the half-life of caffeine in the human body is approximately 2.5-4.5 hr before it is decomposed by enzymes within the liver, whereas the half-life of caffeine in rats is known to be 0.7-1.2 hr [16]. The most active and damaged cognitive function was observed in rats that received 30 mg/kg of caffeine via intraperitoneal injection [17,18]. Pharmacokinetic studies of subjects who consumed caffeine either through oral ingestion or intravenous administration have confirmed that similar blood levels of caffeine are present in both cases [19]. So, we administered the 30 mg/kg of caffeine (Enzo Life Sciences, Inc., Farmingdale, NY, USA) via intraperitoneal injection at the same time (14:00). The vehicle group was injected with 1 ml of phosphate-buffered saline (PBS). In the other groups—Day 1 and Day 14— we administered the caffeine (30 mg/kg) dissolved in 1 ml of PBS over different periods of time, depending on the group, 2 h prior to the behavioral test. Caffeine administration for 13 days in rats was regarded as equivalent to one year in humans [20]. Thus, a period of 14 days can be interpreted as long-term administration [21,22]. To prevent the rats from experiencing caffeine withdrawal symptoms, we administered the caffeine until the final day of the experiment. We carried out three kinds of behavioral tests over 2 days: an open-field test (OFT), a Y-maze test, and a passive avoidance test. There were 3 h intervals between the behavior tests for rest and to allow for adaptation.

The OFT was employed to assess the rats’ sensorimotor function, including the general activity level and the gross locomotor level. In rodents, the activity of grooming is particularly sensitive to various kinds of stress and stimulation; consequently, we used it to represent the animals’ anxiety level [23]. To observe the mobility of the rats, a wooden box (77×77×25 cm) was divided into 16 equal squares delineated by lines. All the animals were placed in the experimental apparatus and given 1 h to adapt to their environment. A movement was recorded every time all four paws of a rat occupied a new square; the incidences of rearing and grooming for a period of 5 min were also recorded. The observer recording the data was blind to the groups and random order. After each trial, the equipment was cleaned with an ethanol solution to erase scent clues that could affect the results.

Spontaneous alternation behavior was evaluated in the Y-maze utilizing the rats’ natural curiosity about new environments to assess the performance of the animals’ spatial working memory. The apparatus was made from wood formed into a Y shape, with three arms, designated as A, B, and C. Each arm measured 5×25×14 cm, and the fixed angle between the arms was 120°. All the rats were placed in the experimental apparatus and were allowed 1 h to adapt to the environment. Initially, the animals were placed in the C arm. Each time an animal moved all four paws into another arm for 5 min, the name of the new arm was recorded. Alternation was calculated as one point, when the animals followed a sequence of moving into different consecutive arms, for example, A→B→C or B→A→C. If an animal’s spatial working memory was intact, it would enter new arms, driven by its nature to explore novel environments. The scoring formula was as follows:

The passive avoidance test is used to evaluate the fear-motivated learning and memory of rats based on their habituation of liking dark places over light places. This test used the automatic system device (Gemini, San Diego Instruments, USA). The apparatus, which measured 66×33×43 cm, was made of acryl and aluminum and consisted of two compartments, a light chamber and a dark chamber, connected via a sliding door that closed automatically when a rat moved from one chamber to another. The light chamber was an environment disliked by rats, whereas the dark chamber served as an avoidance chamber. Exactly 3 h before the test, the rats were moved into a lightless testing room to be habituated. The test was conducted over 2 days and consisted of two sessions. On the first day, in the training session, the rats were placed in the light chamber, which was then illuminated, prompting the rats to enter the other chamber. As soon as a rat entered the dark chamber, the door was closed, and an electric shock of 0.3 mA was administered for 2 s through the floor. On the second day, the test ended when the rat moved into the dark chamber from the light chamber, and no electric shock was administered. The latency time was when the rat moved into the dark chamber from the light chamber. The maximum latency time was set at 300 sec. If a rat went over this limit, the session was finished immediately. If the rat’s learning and memory were intact, it remained in the light chamber longer, rather than entering the dark chamber, because it could remember the electric shock from the training session.

All the data was expressed as a standard error of the mean and was evaluated for normal distributions via the Shapiro-Wilk test. As the variables did not follow normal distribution, the Kruskal-Wallis test was used to obtain overall group comparisons. In addition, Mann-Whitney U tests were used as post-hoc tests to compare two groups at a time. Results which had normality were performed by one-way analysis of variance test, followed by Tukey’s post-hoc test. All analyses were conducted using SPSS Ver. 23.0 (IBM-SPSS Inc., Chicago, Il, USA). The standard of p<0.05 was used as the criterion for statistical significance.

To study the effect of high-dose caffeine on body weight in adolescents, we measure body weight on the postnatal days (PNDs) 28, 35, and 42 (Fig. 1). There was no significant difference between the groups.

We counted the number of movements and the incidence of rearing to determine how high-dose caffeine affects general activities. In the Day 1 and Day 14 groups, the number of movements increased significantly as compared to the number counted in the vehicle group (p <.001; Fig. 2A). In both the Day 1 and Day 14 groups, the incidence of rearing also increased significantly as compared to the incidence observed in the vehicle group (p<.05; Fig. 2B). These results showed that the locomotor activity was affected by high-dose caffeine and not the duration of administration.

In both the Day 1 and Day 14 groups, the incidence of grooming increased significantly as compared to the number counted in the vehicle group (p=.028; Fig. 2C). The rats treated with caffeine exhibited more frequent self-grooming, which indicates anxiety-related behavior.

There was no significant difference between the groups in the Y-maze test, as indicated by the statistics of the vehicle group, Day 1 group, and Day 14 group (p=.802) in Fig. 3A. These results suggest that high-dose caffeine had no effect on spatial working memory, which is a type of short-term memory. However, in both Day 1 and Day 14 groups, the level of escape latency in the passive avoidance test decreased significantly as compared to that observed in the vehicle group (p=.014; Fig. 3B). The learning and memory function was also only influenced by high-dose caffeine.